new geo-oculus: a mission for real-time monitoring through high...

Geo-Oculus: A Mission for Real-Time Monitoring through High-Resolution Imaging from Geostationary Orbit

All the space you need

ESA STUDY CONTRACT REPORT

ESA CONTRACT No. 21096/07/NL/HE

SUBJECT Geo-Oculus: A Mission for Real-Time Monitoring through High Resolution Imaging from Geostationary Orbit

CONTRACTOR Astrium GmbH

* ESA CR( )No

*STAR CODE

No of volumes: 1 This is volume no: 1

GOC-ASG-RP-003 Issue 1-0

ABSTRACT: This Final Report summarises the results of the study “Geo-Oculus: A Mission for Real-Time Monitoring through High Resolution Imaging from Geostationary Orbit” performed from October 2007 to April 2009 and lead by Astrium GmbH in Friedrichshafen. In the frame of the study a comprehensive survey of the potential user needs has been performed with the result that a demand has been defined within the existing political and institutional framework for a high resolution and high revisit mission from geostationary orbit. Out of the identified applications four have been selected as primary mission objectives used to size a system and confirm principle feasibility on a Phase-0 level. Those primary objectives have been: Disaster monitoring, fire monitoring, algal bloom detection and monitoring and water quality monitoring. Taking the user requirements for these applications a set of mission and system requirements has been derived and a first iteration of the payload, spacecraft and ground segment design has been elaborated. On the payload side the design lead to a telescope with an aperture of 1,5 m diameter and five focal planes. The feasibility of implementing the envisaged GSD of around 10 m (at the equator) has been confirmed. The feasibility of all selected applications has also been confirmed. To tackle the stringent LoS requirements various techniques from disturbance suppression over image processing and active LoS control have been studied. Also the application of image post processing on ground with landmark detection (INR) has been considered. On the spacecraft design emphasis has been placed on the AOCS. It has been confirmed that the required agility of the system can be realized. It has been demonstrated that the allocated manoeuvre time including tranquilisation is feasible which leads to an imaging capability of around 42 images per hour for Geo-Oculus. Finally a first iteration of the ground segment architecture has been elaborated investigating the main challenges fast data dissemination and flexible mission planning taking into account on-demand imaging (emergency missions) but also cloud dynamics. The work described in this report was done under ESA Contract. Responsibility for the contents resides in the author or organisation that prepared it. Names of authors: Astrium study team, lead by Astrium Study Manager Ulrich Schull ** NAME OF ESA STUDY MANAGER: Jean-Loup Bézy (EOP-PIO) Earth Observation Programmes Directorate

** ESA BUDGET HEADING: OUTPUT: 60 GSP SUB-HEADING: 510 Special Studies

* Sections to be completed by ESA ** Information to be provided by ESA Study Manager

DL Final Report

Doc. No: GOC-ASG-RP-002 Page iIssue: 2 Date: 13.05.2009 Astrium GmbH

Distribution List

Quantity Type* Name Company / Department 1 PDF J.-L. Bezy ESA 1 PDF M. Aguirre ESA 1 PDF F. Gascon ESA 1 Word & PDF U. Schull Astrium GmbH 1 Word & PDF U. Schäfer Astrium GmbH 1 Word & PDF T. Knigge Astrium GmbH 1 Word & PDF X. Sembely Astrium SAS 1 Word & PDF L. Vaillon Astrium SAS 1 Word & PDF N. Leveque Astrium Ltd

* Type: Paper Copy or Electronic Copy (e.g. PDF or WORD file etc.)

CR Final Report

Doc. No: GOC-ASG-RP-002 Page iiiIssue: 2 Date: 13.05.2009 Astrium GmbH

Change Record

Issue Revision Date Sheet Description of Change 1 28.01.2009 All First version 2 13.05.2009 Consideration of ESA comments: 2-1, 5-87 Reference to GMES 2-2 Reference to S-2 and S-3 3-20 SSD units changed from km to m 3-23 Description of cloud cover change 4-27 Figure 4.1-1 repaired 4-29 Explanation of versions a and b of MWIR/TIR channels 4-32 Explanation of columns 6 and 7 of Figure 4.3-3 4-53 Additional information on data rate 4-53 Information on TRL of downlink system 4-56 Impact of small mobile stations on downlink system 4-58 Comment on active damping of flexible modes 4-60 to 62 Probability level of pointing performance 4-63, 4-65 Units for manoeuvre times added in tables 4-80 Information on alternative decentralised PDGS 5-87 New section added

TOC Final Report

Doc. No: GOC-ASG-RP-002 Page vIssue: 2 Date: 13.05.2009 Astrium GmbH

Table of Content

Distribution List..........................................................................................................i Change Record.........................................................................................................iii Table of Content ........................................................................................................v

1 Introduction ...................................................................................................1-1 1.1 Scope of the document........................................................................................................... 1-1 1.2 References ............................................................................................................................... 1-2 1.2.1 Applicable Documents ........................................................................................................... 1-2 1.2.2 Reference Documents ........................................................................................................... 1-2 2 Product and Mission Baseline Description.................................................2-1 2.1 Overview................................................................................................................................... 2-1 2.2 Survey for Mission Objectives ............................................................................................... 2-1 2.3 Mission Objectives .................................................................................................................. 2-3 3 System Requirements and Mission Scenarios .........................................3-12 3.1 System Requirements........................................................................................................... 3-12 3.2 Major System Trade-Offs ...................................................................................................... 3-17 3.3 Mission Scenarios ................................................................................................................. 3-18 3.3.1 Mission Scenario Baseline................................................................................................... 3-19 3.4 Cloud Coverage Analysis ..................................................................................................... 3-22 4 Mission and System Level Analyses .........................................................4-27 4.1 Mission Architecture ............................................................................................................. 4-27 4.2 Mission Analysis ................................................................................................................... 4-27 4.3 Payload................................................................................................................................... 4-29 4.3.1 Imaging capability ................................................................................................................ 4-29 4.3.2 Radiometric & image quality performances......................................................................... 4-30 4.3.3 Instrument design ................................................................................................................ 4-33 4.3.4 PLM budgets........................................................................................................................ 4-42 4.4 Line of Sight (LoS) Stabilisation Concepts......................................................................... 4-43 4.4.1 LoS stabilisation main issues: microvibrations and post-integration ................................... 4-43 4.4.2 Microvibrations..................................................................................................................... 4-44 4.4.3 Post-integration.................................................................................................................... 4-46 4.5 Satellite ................................................................................................................................... 4-48 4.5.1 Configuration........................................................................................................................ 4-48 4.5.2 Electrical Architecture .......................................................................................................... 4-51 4.5.3 Power Subsystem................................................................................................................ 4-52 4.5.4 Payload Data Handling and Transmission........................................................................... 4-53 4.5.5 Telemetry and Telecommand .............................................................................................. 4-56 4.5.6 Attitude and Orbit Control .................................................................................................... 4-58 4.5.7 Propulsion System............................................................................................................... 4-66 4.5.8 Structure and Thermal Concept........................................................................................... 4-72 4.5.9 Satellite Budgets .................................................................................................................. 4-77 4.6 Ground Segment ................................................................................................................... 4-77 4.6.1 Ground Segment Architecture ............................................................................................. 4-77 4.6.2 Geo-Oculus dedicated Ground Segment issues ................................................................. 4-80

TOC Final Report

vi Page Doc. No: GOC-ASG-RP-002 Issue: 2 Astrium GmbH Date: 13.05.2009

5 Recommendations on further Analysis .................................................... 5-87 5.1 System Analysis ....................................................................................................................5-87 5.2 Mission Objectives and Data Processing ...........................................................................5-87 6 Conclusion .................................................................................................... 6-1

Annex A Abbreviations .................................................................................. A-1

1 Final Report

Doc. No: GOC-ASG-RP-002 Page 1-1Issue: 2 Date: 13.05.2009 Astrium GmbH

1 Introduction

1.1 Scope of the document This document provides the main findings of the study "Geo-Oculus - A Mission for Real-Time Monitoring through High Resolution Imaging from Geostationary Orbit".

The study teaming is as follows:

• Astrium GmbH Study Prime • Astrium SAS Instrumentation and LoS • Astrium Limited Mission Analysis • DLR Institute for Optical Systems Focal Plane Analysis

The following consultants have contributed to the definition of the suitable applications and the collection of the product requirements:

• ACRI-ST Traffic & Security at Sea, Earth Science Applications • Brockmann Consult Marine Applications • Infoterra GmbH Land Applications

The first task of the study has been an open minded survey for applications for a mission that combines fast-response, high-revisit, near-real-time and high-resolution capabilities to introduce a new class of Earth observation missions. Figure 1.1-1 illustrates the uniqueness of Geo-Oculus to provide high resolution images in the scale of Sentinel 2 at the revisit time and the timeliness of MTG. Based on the survey for applications a set of mission objectives has been selected in consultation with ESA for sizing of the system during this study.

Chapter 2 gives a short overview on the starting point for the survey and describes the approach for the selection of the mission objectives. The mission objectives are briefly described in 2.2.

The second task of the study has been to derive and analyse the system requirements and to establish the preliminary candidate mission concepts.

In chapter 3 the driving system requirements are summed up. The major system trade-offs "Field of View vs. resolution", " Magnetic Bearing Wheels vs. Electric Propulsion for manoeuvres", "Manoeuvre time vs. image post-integration effort" and "Image post integration and Inter-channel co-registration" are described in 3.2

1 Final Report

Page 1-2 Doc. No: GOC-ASG-RP-002 Issue: 2 Astrium GmbH Date: 13.05.2009

Revisit vs. Resolution

1

10

100

1000

10000

100000

0,1 1 10 100 1000 10000

Resolution @ SSP [m]

Rev

isit

Cap

abili

ty [m

in]

Geo-OculusDisasterGeo-OculusFireGeo-OculusMarineMTG (HRFI)

MSG

Metop +NOAAEnvisat

EOS

Terra & Aqua

Landsat 7

SPOT 5

Ikonos

GeoEye-1

Pleiades

RapidEye

Sentinel 2(2Sats)Sentinel 3

Bubblesize indicates Timeliness (small=fast)

Figure 1.1-1 Geo-Oculus compared to other EO-missions in terms of revisit against resolution

1.2 References

1.2.1 Applicable Documents [AD 1] ESTEC Contract No. 21096/07/NL/HE

[AD 2] Statement of Work 'Geo-Oculus: A Mission for Real-Time Monitoring through High Resolution Imaging from Geostationary Orbit'; TEC-EEP/2006.93/FG Iss.: 01, Rev.: 01 Date: 03.04.2007

1.2.2 Reference Documents [RD 1] Geo-Oculus: A Mission for Real-Time Monitoring through High Resolution Imaging

from Geostationary Orbit; EADS Astrium GmbH Proposal No., A.2007-4200-0-1: Author: Dr. Ralf Münzenmayer; Friedrichshafen; July 2007

[RD 2] Rapport de l’étude « Contraintes induites par une instrumentation d’observation HR sur une plateforme Géostationaire », PFGEO.ASTR.TN.001.06, Edition 2.1, 08.06.2007

[RD 3] System Requirements Report, GOC-ASG-TN-002, Iss.: 01, Rev.: 00, Date: 02.06.2008

[RD 4] Product Requirements Report, GOC-ASG-TN-001, Iss.: 01, Rev.: 02, Date: 30.05.2008

[RD 5] LoS Stabilisation Concepts, GOC-ASF-IN-002, Iss.: 01, Rev.: 00, Date: 30.05.2008

[RD 6] Candidate Instrument Concepts Report, GOC-ASF-IN-001, Iss.: 01, Rev.: 00, Date: 30.05.2008

1 Final Report


[RD 7] Candidate Mission Concepts Report, GOC-ASG-TN-003, Iss.: 02, Rev.: 00, Date: 19.11.2008

[RD 8] Instrument Analysis Report, GOC-ASF-IN-003, Iss.: 02, Rev.: 00, Date: 16.01.2009

[RD 9] LoS Stabilisation Analysis Report, GOC-ASF-IN-004, Iss.: 02, Rev.: 00, Date: 16.01.2009

[RD 10] Preliminary Mission Analysis Report, GOC-ASG-TN-004, Iss.: 01, Rev.: 00, Date: 28.01.2009

2 Final Report


2 Product and Mission Baseline Description

2.1 Overview The first task of the study has been the identification and selection of applications for Geo-Oculus in order to define preliminary mission objectives including the technical requirements.

Geo-Oculus is set as an independent mission with the objective to enable observations of the Earth so far not feasible with current or planned systems or missions. Among the wealth of EO-activities on European and international level, ESA has identified the lack of the capability for a combination of fast-response, high-revisit, near-real-time and high-resolution observations. Therefore, the starting point for the study is a geo-synchronous satellite mission with high resolution optical imaging instrumentation, real-time control and agile platform.

A survey for mission objectives covering a very wide field of Earth observation applications has been performed to identify applications that require or profit from the basic characteristics of a mission like Geo-Oculus. This survey is briefly described in 2.2 and in more detail in [RD 4]. Although, the survey identified a number of applications that benefit substantially from Geo-Oculus, one major finding is that this mission lays the foundation for new kinds of Earth observation applications which yet have to be recognized.

For the selection process a ranking scheme has been chosen that rates all applications, that where identified in the survey, in terms of "Political Importance", "Institutional / Non-Profit Importance" "Commercial Importance" and "Suitability of Geo-Oculus". Based on this ranking, a final selection of the preliminary mission objectives has been conducted in consultation with ESA.

2.2 Survey for Mission Objectives For the survey for mission objectives, a comprehensive review for user requirements, potential applications and the related product requirements has been conducted. The scope of this survey covers the political framework in terms of ongoing or future European initiatives, especially the GMES initiative, as well as international treaties and European and national directives, policies and protocols. Synergies with European and international Earth observation systems and missions, like the Sentinels, GEOSS and EPS were identified and considered for the identification of suitable applications for Geo-Oculus.

Mission of Choice The analysis of user requirements and potential applications is conducted with an open mind for user demands that will especially benefit from the mission characteristics in fields of e.g.:

• Ecological, economical and humanitarian incidents • Rapidly evolving events • Local to regional monitoring • Instantaneous situation awareness • Regions regularly covered with clouds

The survey points out that Geo-Oculus is the 'mission of choice' for the above mentioned fields of applications. Yet, another finding is that only few applications already exist that require the specific features of Geo-Oculus to become possible. This is not due to missing interest but due to missing

2 Final Report


capability of existing missions. Nevertheless many existing applications are identified that can profit from Geo-Oculus, some can significantly profit or first become possible in an operational manner.

The range of applications covers a large variety of remote sensing applications of the Earth's surface. For the structuring of the documents and to account for the expertise of and task of the consultant it has been chosen to categorize the applications into four fields of services, the:

• Land Applications, covering all services related to the Earth's solid surface, including the land part of the coastal zones and disaster monitoring;

• Marine Applications, covering the services related to the marine ecosystem, hydrology, oceanography, sustainable exploitation of marine resources and anthropogenic forcing and threat to the environment;

• Traffic and Security at Sea Applications, covering marine operations, natural threat to the citizen and law enforcement related to the oceans;

• Earth Science Applications, covering climate research (esp. role of the ocean), the Earth's radiation budget, monitoring of rapid events and data assimilation.

For all applications the necessary and optional products have been identified and defined to a sound level of detail to provide the required technical parameters for the definition of the mission and the instruments. A threshold, a breakthrough and a goal value have been given for most of the parameters, if available through generally accepted literature.

Important synergies Geo-Oculus provides strong assets for synergies with current and planned European EO-missions. The optimisation for cloud cover, which is considered as a central benefit of Geo-Oculus, is only possible with support data from Meteosat and EPS. On the other hand, Geo-Oculus can support other missions to improve quality of service. Some synergies, receiving and supportive, are listed below:

Receiving synergies:

• Real-time cloud cover information from Meteosat and EPS • Fire presence by any means • Highest resolution support data e.g. for disaster monitoring from SPOT, Pleiades, Ikonos etc.

Supportive synergies:

• Oil slick verification • Gap filling due to cloud cover for Sentinel 2 & 3 • Fine scale and real-time spotlight support to meteorology, e.g. for severe weather events

These synergies are considered as a prerequisite for the selection of mission objectives.

Selection process In a preliminary selection process based on the criteria "Political Importance", "Institutional / Non-Profit Importance", "Commercial Importance" and "Suitability of Geo-Oculus" a set of applications was proposed to ESA and in consultation with ESA the preliminary mission objectives were set. These were used for sizing of the system. For that reason, only sufficiently elaborated applications with available technical requirements could be taken into consideration. Nevertheless, these mission objectives represent a realistic case, demanding a challenging system without overtightening the requirements.

2 Final Report


2.3 Mission Objectives The mission objectives for Geo-Oculus that have been selected in consultation with ESA based on the survey described in 2.2 are:

Primary Mission Objectives:

• Disaster Monitoring • Fire Monitoring • Algal Bloom Detection / Monitoring • Water Quality Monitoring with respect to European Regulation

Secondary Mission Objectives:

• Oil Slick Environmental Information • Erosion / Sediment Transport on the European Shoreline Monitoring

An overview of each of the mission objectives is given in the following:

Disaster Monitoring Service - Primary Objective The disaster monitoring service is aimed at providing overview information in case of natural hazards with significant geographic extend. Based on the findings of PREVIEW (2006) on the priorities adopted for Civil Protection activities of the Member States concerning the risk management and the suitability of high-resolution imaging, the following hazards are considered for the Geo-Oculus disaster monitoring service:

• Large landslides • Floods • Windstorms

It is the goal of the disaster monitoring service to deliver geospatial information with short acquisition delay and timeliness of less than an hour on demand of civil protection organisations. This service shall be tailored for early warning, crisis, and post crisis support to the users.

2 Final Report


Figure 2.3-1: Flood Waters surrounding Yangon City, Myanmar, Cyclone Nargis

Source: MODIS on Terra and Aqua, 28.5m/pixel resolution. Acquired: 05/05/2008 and 18/03/2008 Satellite-detected flood waters over Yangon, as of 5 May 2008. Red areas shown in the map represent standing flood waters identified from Landsat 7 satellite imagery acquired on 5 May 2008 at a spatial resolution of 28.5m. Blue areas represent pre-flood waters identified from Landsat 7 acquired on 18 March 2008. Preliminary analysis not yet verified in the field. Credit: Credit NASA/USGS 2008 Image processing, map created 05/05/2008 by UNOSAT.

Disaster monitoring services will benefit substantially of the significant advantages of the GeoOculus mission in terms of acquisition delay, observation cycle and timeliness, all of which in the range of an hour or less.

Fire Monitoring Service - Primary Objective The fire monitoring service is an on occasion service that becomes active once a fire in the service region is present, which has been detected and reported by other means. Therefore it is the objective of the fire monitoring service is to provide timely fire observations on demand of fire fighting and mitigation organisations. The data products shall provide accurate information on fire location, extend, temperature and development over time to allow for optimized planning of mitigation efforts.

This requires a very agile and responsive overall system to assure acquisition delays and observation cycles (revisit time) shorter than about 10 min and data delivery after acquisition in less than about

2 Final Report


15 min. These requirements are not possible to be fulfilled by LEO missions and underline the strengths of the GeoOculus mission.

The primary region of service provision is the European mainland, nevertheless can it be considered to provide this service for the whole visible regions (e.g. Africa) with reduced priority to avoid interruption of other services.

Figure 2.3-2: Forest fires near Los Angeles in October 2003 imaged by the DLR BIRD satellite in

the course of 24 hours. The GSD is 185 m squared. The scale is as follows: yellow = 0,1 MW / Pixel; orange = 1 MW / Pixel; red = 10 MW / Pixel. © DLR

Additional observations of manifold high temperature events, like volcanic activity, tropical peat land fires and coal seams, mainly for scientific purposes, shall also be covered by this service; hence not drive the system requirements.

Algal Bloom Detection Service - Primary Objective An algal bloom refers to a quick and local increase in the abundance of a phytoplankton species. The so-called Harmful Algal Blooms (HAB) are special cases of the former, with deleterious effects on human health or marine resources (natural or cultured). Therefore the algal bloom detection service is aimed to detect and locate an algal bloom in European waters. This information shall be incorporated into the respective GSE MARCOAST service line.

Users include national environment agencies and fisheries industries (aquaculture, shellfish). All coastal and offshore European waters are concerned (region enclosed between latitudes 35°N and 70° N and longitudes 12°W and 30°E).

Although blooms are relatively well detected by remote sensing technique, the identification of their possible toxicity from space is far from being achieved. Satellites play a crucial role for forecasting (combination of EO data with models) and visualising the extent.

2 Final Report


Figure 2.3-3: Algal bloom at Cape Rodney, NZ; Photo by Miriam Godfrey.

Attention shall be paid to the definition of an Algal bloom. For some cases it is an anarchic increase of biological material compared to a climatology (seasonal evolution), for other application it may just be the sudden increase of chlorophyll concentration (thus covering the seasonal trends) as it is presently done in the Algal Bloom service line of GSE-Marcoast.

Algal Bloom Monitoring Service - Primary Objective Algal blooms in European waters, either detected by Geo-Oculus or reported by local authorities, shall be monitored within the algal bloom monitoring service. The service is aimed to provide detailed information on position, extend and persistence to the users (cp. chapter 0). The data can be derived either by dedicated observations or within the required routine scanning of the algal bloom detection service, if applicable.

Figure 2.3-4: Algal bloom east of Scotland, May 7th, 2008, Envisat-MERIS image, unusually strong

algal bloom degrades bathing water quality and threatens the local ecosystem.

2 Final Report


Figure 2.3-5: Algal bloom in the North Sea and west of France, © Y. Park,

MUMM (MarCoast)

Water Quality Monitoring Service with respect to European Regulation - Primary Objective The water quality monitoring service with respect to European regulation addresses mainly the user needs of the WFD and conventions like the Bathing Water Directive. For the later it is required to examine the bathing-water quality with concern to public health criteria. The objective of this service is to provide timely status reports of the water quality of the European waters.

2 Final Report


Figure 2.3-6: Chlorophyll concentration on 31.03.2007 (Envisat MERIS). Processed for Marcoast.

© Brockmann Consult / LANU (MarCoast)

2 Final Report


Figure 2.3-7: Turbidity of the lake Lohjanjärvi on 20.5.2002 (Landsat 7 ETM+).

© Finish Environment Institute (SYKE)

Oil Slick Environmental Monitoring Service - Secondary Objective The detection of oil spills is coordinated at European level through the CleanSeaNet initiative. All information about oil detection is directed towards EMSA who has the mandate to track and survey illegal discharges at sea in order to intercept polluters. The system of oil observation is run in parallel with Automatic Identification Systems (AIS) and Vessel Monitoring System (VMS) that allows identification of polluters. The system is operational since 2007.

Figure 2.3-8: Left: Fresh oil slick spread widely into a thin film.

Right: Partly dispersed oil slick as seen by an airplane. © Cedre [RD T8]

The oil spill detection today relies especially on SAR images but should be complemented with ancillary information (such as SST and Ocean colour) in order to improve the level of confidence of the detection and corresponding reporting – this statement is especially valid in the Baltic where biogenic spills due to biological material may lead to misdetection of oil spills (see EMSA documentation and GSE-Marcoast phase 2 recommendation). This is the objective of the oil slick environmental monitoring service.

2 Final Report


Erosion / Sediment Transport on the European Shoreline Monitoring Service - Secondary Objective The erosion / sediment transport on the European shoreline monitoring service is dedicated to assess the impacts of natural events like floods and storms, as well as increased river discharge on the European shoreline. Even though that the formulated user needs (see TMAP [RD M4]) demand for a spatial resolution in the range of 1-5 m, which is not achievable by Geo-Oculus, the short delay of image acquisition and delivery combined with a still high spatial resolution, denote Geo-Oculus as an indispensable mission for this service; hence the focus of the service is to timely provide data for damage assessment on demand of the authorities.

Figure 2.3-9: Sediment classification of the Wadden Sea, © K. Stelzer, Brockmann Consult

2 Final Report


Figure 2.3-10: Wadden Sea largest intertidal area worldwide is suspect to erosion due to storms and

flooding

Table 2.3-1 gives an overview of the mission objectives including the basic requirements for each application/mission.

Table 2.3-1 Mission Objectives for Geo-Oculus

Application

Mission 1: Disaster

MonitoringMission 2: Fire

Monitoring

Mission 3: Algal Bloom Detection /

Monitoring

Mission 4: Water Quality Monitoring

wrt. European Regulation

Mission 5: Oil Slick

Environmental Information

Mission 6: Erosion / Sediment

Transport on the European Shoreline

Monitoring

Type of service on demand on demandroutine / on

demand routine on demand on demand

Service regions Europe

all European fire endangered areas

up to 45° N all European waters all European waters all European watersEuropean coastal

watersProduct Image Size 150 x 150 km² 100 x 100 km² 100 x 100 km² 100 x 100 km² 100 x 100 km² 100 x 100 km²

Service period all year summer-early fall all year all year all year all year

Daily service period (solar zenith angle, time span) <80°

24 hours (sun avoidance ok) <60° <60° <60° <60°

Effective Revisit Time 1 hour 10 minutes 1 day 1 day 1 hour on requestTimeliness 1 hour 15 minutes 1 hour 1 day 1 hour 1 hour

Acquisition Delay 1 hour 10 minutes 3 hours 3 hours 1 hour 1 hourSpatial Sampling Distance

(products) 10m 250 m 100 m 100 m 25 m as high as possible

Primary Mission Objectives Secondary Mission Objectives

Geo-Oculus Mission Objectives

3 Final Report


3 System Requirements and Mission Scenarios

3.1 System Requirements The system requirements for Geo-Oculus have been derived from user and product requirements, iterated during the first part of this study and documented within the System Requirements Report, RD [3]. These requirements have been further evolved after MTR. Within this chapter, a summary of the driving requirements is given.

Major Challenges for the Geo-Oculus Mission The unprecedented high resolution combined with large areas to be covered within a short period of time drives the system concept of Geo-Oculus. The high resolution requires a large telescope diameter and high pointing stability, whereas the coverage drives the detector Field of View and short repeat cycles ask for short manoeuvre times. A large numbers of required channels drives the instrument optics and focal plane assembly (number of detectors and filter wheel), whereas the MTF and SNR requirements ask for image post-integration techniques.

Definition of Missions Six mission objectives are defined as follows:

Primary objectives:

• Mission objective 1: Disaster Monitoring; • Mission objective 2: Fire Monitoring; • Mission objective 3: Algal Bloom Detection / Monitoring; • Mission objective 4: Water Quality Monitoring with respect to European Regulation.

Secondary objectives:

• Mission objective 5: Oil Slick Environmental Information; • Mission objective 6: Erosion / Sediment Transport on the European Shoreline Monitoring.

Both primary and secondary mission objectives are considered for the system requirements definition.

In order to derive system observation requirements, these six defined missions objectives are compared in terms of system driving requirements (observation cycle, coverage requirements, etc.), see following table.

3 Final Report


Table 3.1-1: Geo-Oculus mission objectives Mission objective

Observation cycle Coverage Observation time / period

Goal Threshold Goal Threshold 1 - Disaster Monitoring

1 hour 2 days Land areas (on demand)

SZA <75° All year

SZA <80° All year

2 - Fire Monitoring

10 min 1 hour Land areas (on demand)

24 hours; from summer period until early fall

3 - Algal Bloom Detection / Monitoring

1 day 3 days Full coverage of European coastlines

SZA <60° All year

SZA <80° All year

4 - Water Quality Monitoring

1 day 3 days Full coverage of European coastlines

SZA <60°

All year

SZA <80° All year

5 - Oil Slick 1 hour 6 hour Specific areas of European coastlines (on demand)

SZA <60° All year

SZA <80° All year

6 - Erosion / Sediment Transport

1 hour 6 hour Specific areas of European coastlines (on demand)

SZA <60°

All year

SZA <80° All year

Comparing observation cycle and coverage requirements of the different mission objectives, some mission objectives can be grouped and some are partly covered by other missions.

Mission 3 (Algal Bloom Detection / Monitoring) and mission 4 (Water Quality Monitoring) will be combined to one mission, called marine application mission. This combined mission can also (at least partly) cover mission 5 (Oil Slick) and mission 6 (Erosion / Sediment Transport), as the specific areas to be covered (on demand) are within the same area specified for the marine application mission. But the shorter observation cycle requires a dedicated consideration of mission 5 and 6.

Observation Requirements The following observation requirements have been established and summarised hereafter:

• Observation Area • Sun Zenith Angle • Observation Time and Periods • Effective coverage • Field of View • Spatial Sampling Distance • Absolute Geolocation Knowledge • Inter-channel Co-registration

A dedicated observation area is defined for each mission. Disaster and fire monitoring is required for the European land masses, whereas the marine applications, Oil Slick and Erosion missions are required to cover the European coastlines. The area is limited towards the north by a view zenith angle of 80°. The whole area to be considered is about 3.6 deg (N/S) by 7.6 deg (E/W). This drives the manoeuvre size to be considered for the mission scenarios and the number of images for the marine application mission. Additionally, this area has to be considered for the design of the PDH concept.

3 Final Report


Furthermore, the mission shall be capable of providing images over the whole Earth disk, with relaxed geometric, revisit, sensing, etc. requirements.

Following figure shows the area to be covered by the marine application mission. The whole area can be covered by 65-70 images with a FoV of 285km x 285km:

Figure 3.1-1: Coverage for marine application mission with 65-70 285km x 285km FoV images

The Sun Zenith Angle (SZA) shall be < 80 deg for images acquired for the disaster monitoring, oil slick, erosion and marine application mission, whereas for the fire monitoring mission, no SZA has to be specified. The definition of the SZA determines the time window, during which the area can be observed. The operational concept shall consider an optimisation of the SZA for the marine applications. The average SZA of all acquired images within the extended observation area of one observation cycle shall be minimised. The radiometric requirements (definition of minimum radiance) shall assume a SZA < 75 deg for the disaster monitoring and < 60 deg for oil slick, erosion and marine application mission.

The observation times and periods for all missions, except the fire monitoring mission, depend on the specification of the maximum allowed sun zenith angle and on the season. For summer solstice, the observation periods are longest. For this case, a mean observation time for the marine application mission of 9 hours has been considered for the mission scenarios presented within the next chapter.

Due to cloud coverage, the effective coverage (=cloudfree coverage) will differ from the nominal coverage. From system side, the impact of cloud coverage can only be minimised by optimising the observation strategy of the marine applications mission. As it is assumed that emergency missions (fire and disaster monitoring) are conducted in parallel to the marine applications mission, the 2° manoeuvres which have already to be considered for the emergency missions, allow the optimisation of the image acquisition cycle for the marine application mission.

3 Final Report


fire

disaster

marine

fire

disaster

marine

Figure 3.1-2: Optimisation of marine application is possible, since parallel fire / disaster monitoring missions assumed. As for these missions 2°-manoeuvres are considered, an optimisation without any additional manoeuvres is possible.

The choice of the Field of View of one image is driven by the trade-off between resolution and the size of image FoV (constrained by detector technology).

The choice of image FoV size directly influences observation scenarios by the number of required manoeuvres and image takes to cover the observation area (for marine application). Actually, the observation scenario (i.e. number of parallel fire / disaster monitoring missions per time, see below) is not a fixed user requirement. In general, the marine application mission asks for a large FoV and medium resolution, whereas high resolution is first priority for the disaster monitoring mission.

For fire, disaster monitoring and oil slick / erosion missions following product FoVs are specified:

Table 3.1-2: Product FoV requirements at SSP [in km x km at SSP] Threshold Goal Disaster monitoring 150 x 100 300 x 200 Fire monitoring 50 x 33 100 x 66 Oil slick 100 x 66 500 x 333 Erosion 100 x 66 500 x 333

An effective FoV of 285² km² has been implemented for all missions. Only for the panchro channel of the disaster mission, mosaic imaging with smaller single FoV sizes is considered.

These FoV sizes consider pointing errors, which reduce the effective FoV compared to the implemented detector FoV.

The spatial sampling distance (SSD) in N/S direction defined in the mission requirements refers to a certain latitude on Earth and do not consider the degradation of the N/S-SSD from nadir to higher latitudes. Depending on the maximum latitude in which the SSD requirement shall be fulfilled, N/S-SSD at sub-satellite point (SSP) can be derived. For the product requirements, a max. latitude of 52.5 deg, leading to a degradation factor of two has been considered.

The most challenging SSD requirement is for the disaster panchro channel with an SSD in N/S direction of 5 m (goal) to 50 m (threshold) at SSP. The SSD requirements for the other channels are more relaxed (between 50 m goal to 500 m threshold).

The absolute geolocation knowledge of each sample of an image observed at one instance shall be

3 Final Report


better than 0.5 SSD at SSP (threshold) / 0.25 SSD at SSP (goal), with a confidence level of 99.73% over each image. This requirement shall be fulfilled for any images, where land areas are included (presence of GCP). For images which contain only water areas (no GCP), the requirement can be relaxed. For these cases, additional AOCS requirements have to be established, assuring a certain knowledge drift stability between two marine images (see pointing requirements table below).

The absolute geolocation knowledge requirement assures also the image-to-image registration (knowledge) performance, which is twice (worst case) the absolute geolocation knowledge (1 SSD at SSP (threshold) and 0.5 SSD at SSP (goal) for a sample of two consecutive images).

The inter-channel co-registration requirement (knowledge accuracy) is 0.3 px (99.73%) between each two channels (referring to pixel size of channel with worse resolution).

Pointing Requirements Following pointing requirements have been derived and established for Geo-Oculus:

• Pointing coverage – European area shall be covered nominally, with the potential to cover the whole Earth disc;

• APE – to acquire a coverage without gaps for the marine applications; • RPE over integration time – to assure high resolution for the panchro channel; • PDE over integration time – to limit image post-integration effort; • PDE for mosaic imaging – to have products without gaps; • PDE knowledge (reference to images with landmarks) – for marine images without coastline.

Timing Requirements The timing requirements can be found in detail in RD [3]. Following requirements have been defined:

• Acquisition Delay; • Timeliness; • Product Acquisition Time; • Temporal Co-registration.

Sensing and Instrument Requirements A set of sensing and instrument requirements have been established for Geo-Oculus, documented within RD [3]. They are also discussed within the Instrument section of this document. Most of these requirements are purely instrument related, therefore they are not discussed in detail within this system chapter. What should be mentioned, is that several system pointing requirements can be derived from these instrument requirements, which has then be traded on system level (see next chapter). Following requirements have been established:

• Radiometric Requirements; • Spectral Accuracy; • Modulation Transfer Function (MTF); • Polarisation.

3 Final Report


Ground Segment Requirements The ground segment requirements can be found in detail in RD [3]. It is based on a centralised Flight Operations Segment (FOS). Requirements have been specified for:

• Functionalities of FOS; • S-Band TM/TC Ground Station and X-Band Ground Station for PDT; • Centralised Payload Data Ground Segment (PDGS); • Standardised User Portal; • Centralised processing facilities; • High-speed communication connections to the PDT receiving stations; • Interfaces to meteorological service providers for the provision of nowcasting and very short

range forecasting information of cloud coverage.

3.2 Major System Trade-Offs In this chapter the major system trade-offs performed after the MTR, leading to the proposed Geo-Oculus baseline, are summarised. They are discussed in detail within the dedicated chapters / documents.

• Field of View vs. resolution The combination of instrument FoV and resolution is limited by the detector technology (number of pixels). Additionally, both the maximum FoV size and the resolution are limited by the telescope size. Due to the coverage requirements for the marine applications, the choice is to go for a maximum possible FoV size (300km x 300 km with the proposed telescope concept), with medium resolution. For disaster monitoring, the best resolution possible with the proposed telescope concept has been chosen. This leads to a smaller FoV size, which requires mosaic imaging for disaster monitoring.

• Magnetic Bearing Wheels vs. Electric Propulsion for manoeuvres MBWs allow high torques and therefore short manoeuvre times. The drawback are the microvibrations, which are much lower than with ball bearing wheels, but still impact the image quality. EPS would create no microvibrations, but increase the manoeuvre time due to the relative small torque, which leads then to a small number of missions. Furthermore, the propellant demand is significant, especially for EPS with high thrust. As a consequence, it has been decided to go for the MBW solution.

• Manoeuvre time vs. image post-integration effort The pointing instability impacts the image quality. With a good pointing stability, no post-integration (including image motion compensation) is needed, as long as the MTF requirements are met. The major contributors to pointing instability are microvibrations (which are not time varying at a short timescale) and solar array oscillations. The solar array oscillations are a direct function of the waiting time after a manoeuvre. The trade-off is therefore between a long manoeuvre time, needing no post-integration and shorter manoeuvre times, requiring post-integration.

3 Final Report


• Image post integration and Inter-channel co-registration If motion compensation on-board is required, several options are feasible: − 1. Post-integration with respect to the first frame of each channel on-board and inter-

channel co-registration based on landmarks on ground. Major drawback are successive processing steps involving resampling.

− 2. Post-integration with respect to the first frame of the all channels on-board. Advantageous is the avoidance of multiple resampling, however it is necessary to perform image matching between different spectral channels which might lead to a somewhat reduced matching performance. No inter-channel co-registration processing by landmarks has to be performed on-ground.

− 3. Read-out of the panchro channel simultaneous to the first frame of each channel. Post-integration will be performed as for option 1. The panchro channels will provide due to their high resolution high matching accuracy regarding image motion compensation.

− 4. Read-out of the panchro channel simultaneous to each frame of each channel. Post-integration and inter-channel co-registration are performed on-board with respect to the first panchro image acquired. Resampling is therefore applied only once. This would lead to the best performance for both image motion compensation and inter-channel co-registration.

The choice for one of these options depend highly on the processing capabilities available on-board. The currently proposed baseline is a motion compensation on-board using only integer pixel shifts (due to lower processing power), which would not meet the inter-channel co-registration requirements. Therefore, the inter-channel co-registration processing is based on landmark processing on-ground.

3.3 Mission Scenarios The main focus for the analysis of the Geo-Oculus mission scenarios is the combination of a back ground marine application mission with high coverage needs (European coastlines) and fast revisit emergency missions. Both kind of missions have to some extend contrary mission requirements (e.g. need of large FoV for marine and high resolution, combined with high revisit for the emergency missions.

A balancing between effective (cloudfree) coverage for the marine application mission and number of emergency missions has to be performed. This depends on:

• cloud coverage statistics; • importance of emergency missions.

The following diagram shows the correlation between effective coverage and number of emergency missions schematically. For precise numbers, a detailed cloud coverage analysis would be needed.

3 Final Report


Effective Coverage

100 %

TBD%

Theroetical maximum coverageduring one observation cycle(cloudfree min. once per observation cycle)

Eff. cov. after 65 imagesOptimised pattern

TBD%

Number of marine Images

65 130 195Improvement of effectivecoverage by Geo-Oculus

Eff. cov. LEO missionnon-optimised pattern

TBD%

Number ofEmergency Missions

910

8

Figure 3.3-1: Correlation between effective coverage for marine application mission and number of

emergency missions

For the system baseline, a mission scenario with 2.5 times coverage of the European coastlines (= 165 marine images) has been chosen.

3.3.1 Mission Scenario Baseline The key parameters for sizing the proposed mission scenario baseline are:

• Manoeuvre time (based on the proposed magnetic bearing reaction wheel baseline); • Image acquisition time; • Product FoV for marine applications; • Number of images for marine applications.

The number of marine missions and parallel emergency missions have to be traded and balanced against each other. The minimum requirements for Geo-Oculus are:

• Full coverage of European coastlines (about 65 images within 9 hours); • At least one fire monitoring mission in parallel (10 min revisit time); • At least one disaster and one oil slick mission in parallel (60 min revisit time).

The time, which is still left can be used for either increase the effective (cloudless) coverage for marine applications or increase the number of emergency missions. The following table gives an overview on the used baseline parameters:

3 Final Report


Table 3.3-1: Baseline parameters for Geo-Oculus mission scenarios

Missions: Disaster Fire Oil Slick / Erosion Marine

Observation times Daytime 24 h Daytime DaytimeObservation cycle (min) 60 10 60 540

Product FoV (km E/W x km N/S, at SSP) 300 x 141 285 x 285 285 x 285 285 x 285

Image FoV (km E/W x km N/S, at SSP) 157 x 1570.25° x 0.25°

300 x 3000.48° x 0.48°

300 x 3000.48° x 0.48°

300 x 3000.48° x 0.48°

APE (orbit + attitude)PDE (mosaic imaging) 700m - - -

Number of images per product 3 1 1 1

FoV at nadir

Required manoeuvres0.25 deg (70 sec) / 0.4 deg (70 sec)* 2 - - -

2 deg (70 sec) 1 1 1 1Total manoeuvre time (sec) 210 70 70 70

Single image acquisition time (sec) panchro: 0.4others: 7.7 1.2 24.6 24.6

Total image acquisition time (sec) 8.1 1.2 24.6 24.6Total time per one mission (sec) 218 71 95 95

Number of channels 13 5 21 21SSD (m E/W x m N/S, at SSP)

Panchro 21 x 10.5UV-VNIR 40 x 20 40 x 20 80 x 40 80 x 40

MWIR, SWIR -TIR -

Product data amount (Mbits) 3.86E+04 3.95E+03 3.30E+04 3.30E+04

* 1.2 deg with Korsch configuration

+/- 7.5 km

150 x 150375 x 375

-

Based on the mission scenario, the number of manoeuvres (and images) per day have been determined and a schematic mission schedule is shown in Figure 3.3-2. For this missions schedule, the 10 min repeat cycle for the fire monitoring is the sizing parameter.

3 Final Report


Mission Schedule:

Fire

2° Manoeuvre

Fire

2° Manoeuvre

Marine

2° Manoeuvre

Marine

2° Manoeuvre

Disaster panchro

0.25° Manoeuvre

Disaster panchro

0.4° Manoeuvre

Disaster other channels

2° Manoeuvre

MarginFire

2° Manoeuvre

Fire

2° Manoeuvre

Marine

2° Manoeuvre

Marine

2° Manoeuvre

Oil slick

2° Manoeuvre

Marine

2° Manoeuvre

Marine

2° Manoeuvre

Margin

10 m

in10

min

Fire

Fire

Fire

Fire

Disaster panchro

Disaster panchro

Disaster other channels

Margin

Margin

Figure 3.3-2: Mission schedule

3 Final Report


Manoeuvre Times and Image Acquisition Times The manoeuvre times can be separated into two contributors:

• Actual manoeuvre time; • S/A tranquillisation time.

The actual manoeuvre time depends on the size of manoeuvre, chosen technology (wheel size, etc.) and manoeuvre strategy.

The S/A tranquillisation time depends mostly on the pointing drift required for the image acquisition (the more challenging the requirement, the longer the tranquillisation time). An allocation of 70 seconds per manoeuvre has been considered.

3.4 Cloud Coverage Analysis Two of the key features of Geo-Oculus, the possibility for real-time commanding and the capability for short revisit cycles have been found to give an essential asset in order to maximise the mission performance - the optimisation of mission planning for cloud cover. The intention of this analysis has been to identify the potential of Geo-Oculus that can be gained, to validate the system requirements, to identify a possible optimisation strategy and to assess the performance compared to reference missions.

Due to its geostationary orbit, Geo-Oculus has the capability to access every spot within its footprint at the time the spot becomes cloud free. Considering the applied FOV and the possible agility of the system, this capability confined. In result only a certain image acquisition frequency is achieved; hence a selection of the images is required. This leads to the point that the system will have to apply a permanently updated optimisation of the mission planning, to gain maximum possible ground coverage. This optimisation should take into account the current cloud cover situation, possibly supplied by MTG and Metop, the changing illumination conditions throughout the entire day, now-casting and short range forecasting information on the expected cloud cover situation and the constraints placed by the on-demand missions.

In the analysis described in here, a simplified optimisation strategy and mission planning have been used, considered to represent a realistic approach. This strategy accounts for the illumination conditions and maximises the achieved ground coverage.

The entire cloud coverage analysis is based on cloud mask data from MSG with a revisit time of 15 min. The time span, considered in this analysis range from 01/2004 to 05/2007. In a preliminary low level analysis representative days for a detailed evaluation of the cloud coverage are filtered out of the complete dataset. To gain representative results from the analysis, representative days are indicated by analysing every day concerning:

• Cloud amount • Cloud coverage changes • Time of sufficient illumination conditions

By comparing the values of each day with the mean value of the whole data set, several days for detailed analyses have been indicated.

3 Final Report


Referring to these days the detailed cloud coverage analysis is conducted. It comprises four stages:

• Analysis of evolution of geometrical conditions through the day (Illumination situation) • Analysis of cloud coverage amount and dynamics • Evaluation of the performance of Geo-Oculus for different system set ups • Comparison of the performance provided by Geo-Oculus with other planned EO Systems

The analysis of geometrical conditions regards especially the system requirements on the solar zenith angle and the view zenith angle. For Geo-Oculus the view zenith angle of every region is constant all times. By contrast, the solar zenith angle, hence the illumination condition changes through the day and is depended to the season. To consider this in the analysis, the illumination conditions are calculated for each cloud mask file by computing VZA and SZA for each pixel. These information are one necessary input for the simplified mission planning, applied in the performance evaluation of Geo-Oculus.

The second necessary input information are evaluated in the analysis of cloud coverage amount and dynamics. Herein the cloud mask data is evaluated concerning cloud amount and cloud coverage changes through one day.

• Cloud amount is defined as how long a pixel was clouded in the time between 06.00 UTC and 18.00 UTC. It is provided in percent. Analysing the cloud amount allows to point out areas, where observation is possible, in general.

• Cloud cover changes is defined as number of changes of a pixel from clouded to unclouded or vice-a-versa within the considered time-frame (06.00 UTC to 18.00 UTC) in the 15 min time interval of the MSG data. Since 49 cloud mask files are available in this time-frame, a maximum of 48 cloud cover changes can occur.

With the evaluation of the cloud coverage changes, the dynamics of the cloud situation are indicated. With the help of this, it is possible to point out areas where (nearly) cloud free products can be generated, by acquiring the same area several times, as it is possible with Geo-Oculus. Some results of this analysis are to be seen in figure 3.4-1.

3 Final Report


Figure 3.4-1: Cloud amount and Cloud coverage changes during 6.00UTC and 18.00 UTC at

30.09.2005

The plots in figure 3.4-1 shows cloud amount and cloud coverage changes other Europe. It is to be seen, that nearly complete Europe and its coastlines are clouded at least 50% of the day (1st plot). But there are also a lot of areas within Europe or its coastlines, where the cloud situation changes during the day (2nd plot). One can assume, that observations allowing only one acquisition per day for a certain area, as it is provided by a LEO system, will lead to a rather small ground coverage. The

Cloud Amount [%]

Number of Cloud Coverage Changes

3 Final Report


ground coverage can be increased by acquiring the same spot several times, as already mentioned. This is useful especially in areas, where high cloud coverage dynamics occur, hence clouds are moving or dissolving a lot. In these areas, indicated by a high amount of cloud coverage changes, Geo-Oculus can increase the ground coverage through multiple acquisitions of the same areas. In result Geo-Oculus provides a higher performance in means of ground coverage than LEO systems. These results have been correlated with the impacts of changing illumination situation, to indicate the areas where the ground coverage can be increased most by observations with Geo-Oculus for the handled day.

The performance evaluation of Geo-Oculus has been conducted for the Marine Applications mission, which is accomplished as background mission. The results are considered to be representative for these missions and also illustrate the capacity of the system for on Demand missions in the sea areas, like Oil Slick Monitoring. For the Marine Applications, an image pattern has been implemented. The simplified mission planning considers that the acquisition sequence is updated immediately when an update on the cloud coverage information becomes available to the system; hence with every cloud mask file (one new cloud mask file every 15 min) the mission plan is optimised and updated. According to the agility of the system a certain number of acquisitions is possible within 15 min and a selection of the images observed within the next 15 min has to be accomplished. The number of acquisitions within 15 min is also depended to the number of parallel on-demand missions, which have to be accomplished. The selection is based on the results of the analysis of geometrical conditions and of cloud amount and cloud coverage changes. The current baseline foresees 4 images per 15 min. The final product is achieved by combining all the acquired images. In the analysis the combination of the images leads to the total observed area which identifies the performance of Geo-Oculus.

Finally the cloud coverage analysis compares the performance of Geo-Oculus with LEO Systems like Sentinel 2 and Sentinel 3, regarding the total ground coverage which can be achieved at the handled day. For this, different LEO swaths are implemented and superposed with the same cloud mask data, as used for the performance evaluation for Geo-Oculus. Image 3.4-2 shows the performances of Geo-Oculus and LEO systems by highlighting the ground coverage for one day:

Figure 3.4-2: Ground coverage within one day for Geo-Oculus (left) and LEO (Sentinel 3, right) are

highlighted blue

It can be seen that Geo-Oculus provides considerably more ground coverage (~83,3% of the maximum possible coverage, ~55% effective) than a LEO system (Sentinel 3 ~35% of the maximum possible coverage,~23% effective). This is due to the fact that the area for observation is accessible to Geo-Oculus the whole day, whereas a sun-synchronous LEO mission provides commonly ~3 passes over Europe during one day. This provides Geo-Oculus the advantage to benefit already from dynamic

3 Final Report


cloud scenes where the cloud coverage, although the spots might feature high cloud amount. The only restrictions for observations on Geo-Oculus, are areas clouded the whole day with no cloud movement.

Conclusion Geo-Oculus has been found to provide the best possible ground coverage at high resolution with a significant improvement compared to LEO-missions. The achievable ground coverage with Geo-Oculus at ~40 m GSD over Europe is ~2.5 times more than Sentinel 3 at 300 m GSD. This advantage results from the swath width, the orbit geometry of LEO-missions which results in three swaths per day over Europe at fixed local times and on the other side the capability of Geo-Oculus to access whole Europe and to pick the cloud free points in time. The unique feature of multiple acquisitions and near real time mission plan updating is bund to geo-synchronous missions and can not be provided by LEO-missions.

4 Final Report


4 Mission and System Level Analyses

4.1 Mission Architecture A visualisation of all elements contributing to the Geo-Oculus mission architecture is shown in Figure 4.1-1.

Figure 4.1-1: Mission Architecture

4.2 Mission Analysis Mission analyses issues have already been traded in [RD 7] for the following topics:

• type of orbit, • orbit inclination, • orbit determination performance, • orbit transfer and launcher.

The preferred mission parameters which were assumed for the subsequent analyses of technical solutions for the spacecraft are summarised hereafter.

Type and inclination of orbit A geostationary orbit with 0 degree inclination is the suggested baseline. It requires only one spacecraft and provides constant observation conditions (view Zenith angle) but it is linked to about 300 kg of fuel consumption to be assigned for North South station keeping.

4 Final Report


Alternative solutions which have been investigated are:

• geostationary orbit with limited station keeping, • inclined geosynchronous orbit, • Molniya orbit (highly elliptical inclined orbit).

The appealing features of these alternatives are on one hand potential propellant savings for station keeping and on the other hand improved viewing conditions over Europe.

The maximum propellant saving may be achieved for a geostationary orbit with 7.5 deg inclination and 0 deg initial right angle of ascending node. This orbit is rather stable such that no orbit corrections are needed. However, as it is the case for all inclined geosynchronous orbits, the orbit trace projected onto the Earth surface is a figure of eight which the satellite passes through once per orbit. This yields that the satellite is half an orbit over Northern latitudes (with better viewing conditions over Europe) and the other half orbit it is over Southern latitudes. If no corrections of the orbital plane are performed, the local time of the equator crossing will change over the year. This means that the satellite is at the most Northern position e.g. at 12:00 at a certain day of the year but half a year later this position is achieved at midnight. Hence, the good viewing conditions at daytime are achieved for a certain part of the year only and become even worse for the other part of the year. Since this feature is a significant constraint for the mission flexibility, all alternatives with inclined orbits have currently been dropped.

Orbit determination performance The achievable precision for the position of the spacecraft is important since it contributes to the overall pointing budget which is rather stringent. The preferred solution with the best performance is a ranging technique based on 2 or 3 ground stations and using a spread spectrum method for the ranging signal. The orbit determination accuracy is 100 m to 150 m (3σ) along track and 10 m to 20 m (3σ) across track. The interesting feature is that this technology shows the same performance right after a manoeuvre when using 3 ground stations. Moreover, this technique is routinely used by SES Astra which gives strong evidence that it can be successfully applied to Geo-Oculus.

The following alternative technologies have also been investigated:

• single ground station ranging + line of sight measurements, • dual ranging, • DARTS, • short baseline interferometry, • long baseline interferometry, • optical telescope, • use of landmarks, • GPS.

The GPS option is currently being investigated for geostationary orbits. It may provide a similar nominal performance as the ground based spread spectrum method but a significant degradation is expected for a couple of hours after a manoeuvre. All performance data for above listed options are found in [RD 7].

Orbit transfer and launcher A geostationary transfer orbit strategy is suggested as baseline. The initial elliptical orbit provided by the launcher is changed to the final circular orbit with a liquid apogee engine installed in the spacecraft. The achievable maximum mass of the spacecraft in the final orbit is slightly higher

4 Final Report


compared to a strategy where the launcher provides a direct injection. Moreover, this strategy is a common standard with a high level experience whereas the direct injection is offered by few launchers only.

Depending on the final launch mass of the spacecraft, the preferred launchers are Soyuz from Kourou (up to ~3 tons) and Ariane 5 (more than 3 tons). For alternatives of non-European launch service providers see next table.

Table 4.2-1: Launcher Survey: Standard Launch into GTO including Performance Injection into GTO Launcher Name

Launch Service Provider

European LSP

Perf. into GTO [kg]

ΔV to GSO[m/s]

Remark

Ariane 5 ECA Arianespace Y 9000 1500 flight qualified Soyuz Fregat / Kourou Arianespace Y 3000 1480 under development Soyuz Fregat / Baikonur STARSEM Y 1840 1500 flight qualified Atlas 5 ILS N 8670 1804 flight qualified Delta 2 Boeing N 2120 1840 flight qualified Delta 4M Boeing N 6470 1800 not commercially available Delta 4H Boeing N 10819 1800 not commercially available Proton * ILS N 5530 1500 flight qualified Sea Launch Sea Launch N 5850 1500 flight qualified Land Launch * Sea Launch N 3600 1500 flight qualified for direct GEO GSLV Antrix N 2400 1650 flight qualified up to 2t H-2A MHI N 6000 1840 flight qualified up to 5t Long March 3B CGWIC N 5000 1840 flight qualified Falcon 9 Space X N 5070 TBD under development Angara 3 * ILS N 2400 1500 under development Angara 5 * ILS N 5400 1500 under development

* performance for S/C + adapter; all other launchers are S/C separated masses

4.3 Payload

4.3.1 Imaging capability In order to support Fire Monitoring & Marine applications, the instrument provides simultaneous imaging of Earth scenes on four multi-spectral focal planes (UV-blue, Red-NIR, MWIR and TIR) with a ground FoV of 300x300 km (0.48x0.48 deg). The spectral channels are defined in the following figure together with the achieved ground resolution over Europe (worst case given at 52.5 °N corresponding to a viewing zenith angle of 60 deg). For some channels (e.g. for IR ones), subscript "a" refers to Fire Monitoring mission while "b" refers to Marine application and corresponds to different radiometric requirements (e.g. SNR & typical radiance). The VNIR resolution for marine applications is 80 m, twice that of other missions because pixel binning is necessary to meet the challenging SNR requirements of these applications (see next section).

In addition, the Disaster Monitoring applications requires a VIS panchromatic (PAN) focal plane with higher resolution (10.5 m nadir, 21 m over Europe) and reduced FoV (157x157 km, i.e. 0.25x0.25 deg) imposed by the use of the same detector array as the UV-blue & Red-NIR channels. The PAN channel is separated in the field from the other channels.

4 Final Report


The achieved ground resolution & coverage are generally between threshold (T) and goal (G) requirements, with the following exceptions:

• VNIR channels have a resolution (40 m & 80 m for marine) better than the G requirement

• T requirement is not met in TIR for Fire Detection due to sensor size limitation. This is deemed acceptable because the MWIR band is actually used for fire area monitoring, whereas TIR bands are used to monitor fire temperature, for which resolution is not critical.

Channel ID Center wavelength

Bandwidth Focal planes

(nm) (nm)UV1 318 10UV2 350 10

VNIR1 412 10VNIR2 443 10VNIR3 490 10VNIR4 510 10VNIR5 555 10VNIR7 655 155 PANVNIR6 620 10

VNIR8a 665 10VNIR8b 665 10VNIR9 681 8

VNIR10 709 10VNIR11 753 8VNIR12 779 15

VNIR13a 865 20VNIR13b 865 20VNIR14 885 10VNIR15 900 10VNIR16 1040 40SWIR 1375 50

MWIRa 3700 390MWIRb 3700 390TIR1a 10850 900TIR1b 10850 900TIR2a 12000 1000TIR2b 12000 1000

UV-blue

Red-NIR

TIR

SWIR MWIR

Figure 4.3-1: Spectral channels (optional channels are in blue) & imaging capability summary

4.3.2 Radiometric & image quality performances

4.3.2.1 Disaster Monitoring The high resolution PAN channel drives telescope diameter (set to 1.5 m) & pointing stability requirements. The acquisition is performed in 4 successive images, all downloaded for on-ground processing for SNR & MTF recovery. Indeed, the Nyquist MTF requirement for the raw images is relaxed from 10% to 5% to allow best resolution. The required SNR is increased in the same ratio to keep constant the SNRxMTF figure of merit to account for SNR degradation in MTF recovery by ground processing. Thanks to that, a Ground Sampling Distance (GSD) as good as 10.5 m (nadir), i.e. 21 m at 52.5 N is achieved, close to Goal requirement (10 m) for Disaster Monitoring. This resolution is achieved with LoS pointing stability requirements of 5 µrad/s and 0.15 µrad p-p, well within the capability of the selected AOCS design based on magnetic reaction wheels. As shown in §4.4, selecting conventional ball bearing wheels mounted on isolators (jitter increased 0.25 µrad p-p) has however a modest impact, with nadir GSD degraded to 11.5 m (23 m at 52.5 N).

The same CMOS detectors are used for all UV-VNIR focal planes to reduce development cost & risks. The GSD of the UV-blue & Red-NIR channels is then obtained from the PAN GSD in the ratio of the

PAN 1 21.0 157x157UV-blue 4 40Red-NIR 8 40Red-NIR 2 40

SW/MW IR 2 300TIR 2 750

UV-blue 7 80Red-NIR 10 80

SW/MW IR 2 300TIR 2 750

300x300

GSD (m) at 52.5°N

Number of channels

300x300

Disaster Monitoring

FOV (km)

Marine Applications

Channels

Fire Monitoring

300x300

Mission Ground Pixel Size at 52°N [m]

Image ground coverage [square, km]

T G T G Disaster monitoring 100 10 100 200

Fire monitoring 250 100 100 200

Marine applications 1000 100 100 500

Ground resolution & coverage requirements

4 Final Report


FoV (300/157 = 1.9), i.e. 20 m (nadir). Since no SNR requirements were derived from user needs, a value of 150 was selected for all VNIR channels, consistently with PAN channel and Sentinel 2 requirements (around 150 and up to 170 for some channels).

The instrument parameters for each channel are summarised in Figure 4.3-2. The GSD at the maximum latitude where full performance is required is recalled in column 3.The MTF at Nyquist frequency in is given in column 4, showing that the 10% requirement is met for all multispectral channels. For the PAN channel, despite relaxation to 5%, the requirement is met without much margin. This clearly demonstrates that the maximum resolution achievable with a 1.5 m telescope is reached for this channel (10.5 m nadir).

1 2 3 4 5 6 7 8 9Final Nb Post Integ. Time Channel acq.SNR of images Integration of image (s) time (s)

Disaster VNIR2 40 0.117 150 1 No 0.009 0.088Disaster VNIR3 40 0.116 150 1 No 0.011 0.088Disaster VNIR4 40 0.115 150 1 No 0.014 0.088Disaster VNIR5 40 0.111 150 1 No 0.017 0.088Disaster VNIR6 40 0.110 150 1 No 0.021 0.088Disaster VNIR7 21 0.051 300 4 No 0.013 0.352Disaster VNIR10 40 0.111 150 1 No 0.029 0.088Disaster VNIR11 40 0.115 150 1 No 0.005 0.088Disaster VNIR12 40 0.112 150 1 No 0.034 0.088Disaster VNIR13b 40 0.113 150 1 No 0.034 0.088Disaster VNIR14 40 0.098 150 1 No 0.073 0.088Disaster VNIR15 40 0.117 150 1 No 0.011 0.088Disaster VNIR16 40 0.110 150 1 No 0.027 0.088

GSD at 52°N (m)

MTF at NyquistMission Channel

Figure 4.3-2: Performances for Disaster Monitoring

The SNR obtained after accumulation of the number of successive images indicated in column 6 is given in column 5. All multi-spectral channels can be acquired in a single image, i.e. without post-integration, while keeping good image quality (MTF at Nyquist > 10%, see column 4). The integration time of individual raw images is given in column 8. The time to acquire the channel (last column) is obtained by multiplying the number of images by the largest value between the integration time and the array readout time.

4.3.2.2 Marine applications For marine applications with challenging SNR requirements, 2x2 pixel binning is used to increase the collected signal, so the final ground resolution is 40 m nadir and 80 m at 52.5°N, well within the goal requirement of 100 m. Thanks to this lower resolution, the requirements on the pointing stability during imaging periods can be relaxed to 10 µrad/s, allowing to largely reduce the tranquilisation time following a slew manoeuvre. For channels with high SNR requiring long exposure time, the image acquisition is split in several successive images to avoid pixel saturation & image smear due to pointing drift. These images are summed on-board, with for the most critical ones (UV-blue channels) compensation of the image motion (so-called "post-integration").

The instrument parameters for each channel are summarised in Figure 4.3-3. The MTF at Nyquist requirement (10%) is met with good margins for all channels. Post-integration with LoS motion compensation is only required for the bands of the UV-blue focal plane (UV1 to VNIR5). Other bands require several images to avoid saturation at typical flux, these images are simply added in the on-

4 Final Report


board image processing. Column 7 gives the number of images requiring post-integration with LoS motion compensation. For instance, channel VNIR1 requires an accumulation of 18 images (column 6) with image motion compensation between 6 packets (column 7) of 3 simply accumulated images. Channel UV1 requires post-integration with motion compensation of 30 successive images (same value in columns 6 & 7).


Marine UV1 80 0.156 1000 30 30 0.082 2.637Marine UV2 80 0.162 1000 10 10 0.083 0.879Marine VNIR1 80 0.146 1500 18 6 0.035 1.582Marine VNIR2 80 0.161 1300 15 5 0.031 1.318Marine VNIR3 80 0.154 943 6 3 0.050 0.527Marine VNIR4 80 0.175 748 6 3 0.041 0.527Marine VNIR5 80 0.171 557 2 2 0.085 0.176Marine VNIR6 80 0.144 418 2 No 0.059 0.176Marine VNIR8b 80 0.159 376 1 No 0.105 0.105Marine VNIR9 80 0.140 339 1 No 0.118 0.118Marine VNIR10 80 0.165 323 1 No 0.098 0.098Marine VNIR11 80 0.216 478 2 No 0.019 0.176Marine VNIR12 80 0.186 258 1 No 0.073 0.088Marine VNIR13b 80 0.193 213 1 No 0.051 0.088Marine VNIR14 80 0.137 213 1 No 0.108 0.108Marine VNIR15 80 0.201 259 1 No 0.023 0.088Marine VNIR16 80 0.164 250 1 No 0.055 0.088Marine SWIR 300 0.271 250 1 No 4.19E-04 0.025Marine MWIRb 300 0.095 378 1 No 0.017 0.025Marine TIR1b 750 0.097 1818 1 No 5.64E-04 0.016Marine TIR2b 750 0.102 2001 1 No 5.46E-04 0.016

Mission Channel GSD at 52°N (m)

MTF at Nyquist

Figure 4.3-3: Performances for Marine applications

4.3.2.3 Fire Monitoring Fire monitoring is based on three IR channels, a MWIR channel with 300 m resolution to monitor the fire area & location and two TIR channels with 750 m resolution to measure fire temperature. Two VNIR channels with moderate SNR are also required, for which 40m resolution is possible without post-integration if VNIR13a SNR requirement is relaxed from 213 (user requirement) to 150. This value is assumed to simplify the image acquisition scheme.


Fire VNIR8a 40 0.112 80 1 No 0.017 0.088Fire VNIR13a 40 0.103 150 1 No 0.062 0.088Fire MWIRa 300 0.095 233 1 No 2.62E-05 0.025Fire TIR1a 750 0.097 670 1 No 4.72E-05 0.016Fire TIR2a 750 0.102 736 1 No 5.05E-05 0.016

Mission Channel GSD at 52°N (m)

MTF at Nyquist

Figure 4.3-4: Performances for Fire Monitoring

4 Final Report


4.3.3 Instrument design

4.3.3.1 Optical design The instrument is based on a 1.5 m diameter all-SiC monolithic telescope, i.e. the same size as Aeolus/ALADIN, formed by M1 & M2 mirrors in Cassegrain configuration. The PAN channel that requires a long focal length is imaged by a Korsch telescope formed with a third converging mirror following a flat folding mirror placed out of the M1-M2 axis. The other channels are separately imaged by the Cassegrain telescope formed by M1 & M2 mirrors (as shown in Figure 4.3-5), or alternately by the symmetric Korsch configuration. A first dichroïc plate is used to separate the UV-blue & Red-NIR channels from the IR channels, and within each group a second dichroïc plate provides separation between the focal planes. Four filter wheels (one for each focal plane) are used to select the narrow channels in each band. Cold stops are required in front of the IR focal planes which need to controlled at low temperatures (130 K for MWIR and 50 K for TIR).

M3

M2

M1

PAN high resolutiondetector

TIR detector & filter wheel

Field correction & focal length

adjustment

SW/MW IR detector & filter wheel

UV-Blue detector & filter wheel

Red-NIR detector & filter wheel

Figure 4.3-5: Multi-spectral imaging telescope optical architecture

The retained optical combination is based on three Korsch combinations with dioptric correctors (PAN, UV-VNIR, IR) for focal length adjustment & aberration correction in the large FoV. Several folding mirrors are required to accommodate the five channels in the volume below the M1 mirror.

Figure 4.3-6: Optical configuration (after M2)

TIRMWIR PANCHRORED-NIRUV-BLUE

From M2 View from SVM

4 Final Report


4.3.3.2 UV-VNIR focal planes For UV and VNIR spectral bands (from 315 to 1040 nm), silicon semi-conductor is the only material considered, thanks to its large maturity, its lower cost, its ability to build large format arrays and its low dark current at ambient temperature. Monolithic CMOS array is preferred to CCD for its good maturity to build large arrays, its better immunity to GEO harsh radiation environment (as shown by GOCI/COMS CMOS detector qualification) and because smearing during transfer rules out large CCD arrays for Earth observation.

Monolithic CMOS imagers manufactured with processes optimised for imaging applications (so-called “CIS”) look by far as the most promising technology for Geo-Oculus. Indeed, thanks to improved photodiode processes, CIS arrays are featuring excellent electro-optics performances even for small pixel pitches. Thinned backside CMOS technology is considered to improve the fill factor and therefore the detection efficiency. Back-illuminated CMOS are already available in the USA and are being investigated in Europe, so availability in a 5-year frame is very likely. Considering only mandatory channels, the spectral range to be covered by the CMOS detector (0.4 0.9 µm) is compatible with a conventional “broadband” detector. When accounting for optional channels, it is impossible to have a good detection efficiency in the large spectral range to be covered by the CMOS detector (0.315 to 1.04 µm), so two detectors are used, one optimised for UV & short visible wavelengths (“UV-blue” detector) and the other for red & NIR wavelength (“Red-NIR” detector).

The largest space CMOS arrays currently available in Europe are in the range of 1.5k x 1.5k pixels (e.g. COBRA2M 2 Mpixels detector developed by Astrium & ISAE/CIMI for the COMS/GOCI instrument). The next step will be 3k x 3k arrays, expected for 2010-2011. There is therefore a major step to be performed in order to reach the typical 100 Mpixel array size required for Geo-Oculus. Even though there is no strict limitation in CMOS array size, such large arrays set many technical challenges. In particular, the manufacturing process shall be well mastered to guaranty a sufficient yield, i.e. a reasonable cost & development schedule. Moreover, to build a very large format array without dead zone with good electro-optics performances and not to constraint too much the planarity, stitching (i.e. gap-less array assembly during manufacturing process) will be required (see Figure 4.3-7) since the total array will not fit within the stepper field (currently limited to 22x22 mm²).

array

horiscan1

horiscan2

V3

V2

V1

array

horiscan1horiscan1

horiscan2horiscan2

V3

V2

V1

array array array

array array array

array array array

horiscan1 horiscan2

V3

V2

V1 array array array

array array array

array array array

horiscan1 horiscan2

V3

V2

V1

– Sub-blocks are exposed one after another

– Some blocks are used multiple times

– Ultimate limit is given by wafer size

Stepper field

Stitched CMOS Sensor

22mm

Figure 4.3-7: Wafer-level stitching is used to build arrays with size larger than the stepper field

As for the 2 Mpixels detector of the GOCI instrument, it is proposed that the array is divided 4 sub-blocks independently operated, offering a redundancy level in case of failure (see Figure 4.3-8). Each

4 Final Report


sub-block has16 video outputs with 20 Mpixels/s data rate, allowing reading the total array in less than 100 msec.

Column readout circuit 16 outputs 16 outputs

Column decoder

4 stitched 25 Mpix arrays


Column decoder

Figure 4.3-8: Architecture for the CMOS detector for PAN , UV-Blue & Red-NIR focal planes

Each detector is interfaced with a Proximity Electronics Module (PEM) housing all the functions requiring to be located close to the detector, i.e. detector sequencing (e.g. clock generation), bias voltage supply and video signal pre-amplification.

4.3.3.3 MWIR focal plane The selected technological approach is that photo-detectors arrays are manufactured by using the adequate detection material and hybridised on top of a CMOS Read Out Integrated Circuit (ROIC). The ROIC is in charge of providing the reference bias voltage to each photo-detector, injecting the signal at the output of the photo-detector within the corresponding integration capacitance and multiplexing the analogue signals from all the pixels through a reduced number of outputs.

AlGaAs/GaAs or InGaAs QWIP technology was initially preferred to HgCdTe for its better yield, operability and uniformity. QWIP main drawback is its lower sensitivity. However, as MWIR integration time is low with respect to read out time, the sensitivity should not be the driver of the choice, whereas cost, stability, cosmetics and uniformity are important Nevertheless, this choice had to be reconsidered when an additional SWIR channel had to implemented. Indeed, QWIP technology does not allow wide-band detectors with good detection efficiency from 1.3 to 3.7 µm, so a dedicated SWIR focal plane would be required. The right choice is then HgCdTe technology which allows such a combined SWIR/MWIR detector with good detection performances, but also with the yield drawbacks pointed out above.

Driven by the minimum pixel pitch that can be achieved for European indium bump hybridized detectors (i.e. 15 µm), the 30x30 mm2 area of the 2k x 2k photo-detector array assumed for Geo-Oculus is larger than the today European state of the art (20 mm diagonal for QWIP and 25 mm for HgCdTe, but seems reachable within a few years provided the necessary pre-developments are performed. The CMOS ROIC associated to the photo-detector array is another challenge. Stitching will be required, 30x30 mm2 being larger than the stepper field.

The MWIR detector architecture is similar to the UV-VNIR one (see Figure 4.3-9).

4 Final Report



Column decoder

4 stitched 1k x 1k arrays15 µm pixel pitch


Column decoder

Figure 4.3-9: Architecture for the MWIR CMOS hybrid detector

The 2k x 2k array with 15 µm pitch consists in 4 independent 1000x1000 pixels sub-arrays. Each sub-block is read out via 4 video outputs with a 10 Mpixels/s output rate, so the read out period is 25 ms, much larger than the max. integration time (2.5 msec for Marine application).

The operating temperature of the detector is dictated by the level of dark current, which can reduce the useful dynamic range and significantly increase the detection noise. A temperature of 130K is a typical choice for a 3.7 µm MWIR band.

4.3.3.4 TIR focal plane As for MWIR, the two candidate technologies are TIR HgCdTe and AlGaAs/GaAs QWIP. From a qualitative point of view, the trade-off between both materials is identical: better sensitivity for HgCdTe and better yield / uniformity (spatial and spectral) / temporal stability / cosmetics for QWIP. The situation is however worse for TIR HgCdTe as its metallurgy complexity is strongly increasing with cut-off wavelength. A 25 µm pixel pitch is considered as the smallest achievable pixel pitch. The targeted format of 0.8k x 0.8k pixels has a 28 mm diagonal, larger than the HgCdTe state of the art. The development of the GIFTS array made by BAe in the US has shown that, despite important technological and financial efforts, it looks hard to produce with acceptable operability a 20 mm diagonal HgCdTe 2D array with very long cut-off wavelength. On the other hand, 640x480 pixels QWIP arrays with 25 µm pitch are currently produced by few manufacturers. The HgCdTe problem is well known by ESA detection experts, particularly in the framework of MTG studies, justifying the two ways approach proposed by the Agency:

• Improve the weaknesses of HgCdTe, via technological development. This is the object of the running contract "initial design of thermal infrared detector array for MTG"

• Improve performances of alternative ways. This is the object of the contract "Enhanced QWIP/Sb-superlattice Array Detector".

About ROIC, the 800x800 pixels format with 25 µm pitch avoids the need for stitching.

An alternative to quantum detectors that could be figured for Fire Monitoring applications (which have much relaxed noise requirements) is the emerging microbolometer technology.

4 Final Report


Microbolometers measure changes of electrical resistances with the help of pulsed currents. The major advantages are the possibility to operate at room temperature and the monolithic silicon structure which allows cheap production. Microbolometers are sensitive between 7 and 14 µm, but the responsivity is much lower than for quantum detectors. Microbolometers are not retained in the baseline for Fire Monitoring because the performances and the maturity level for GEO applications needs to be consolidated.

The TIR detector architecture is similar to the MWIR one, but simpler thanks to the reduced number of pixels: two 400x800 pixels sub-arrays read out via 2 10 Mpixels/s video outputs, so the read out period is 16 ms, much larger than the max. integration time (<1 msec for all applications).

Figure 4.3-10: Architecture for the TIR CMOS hybrid detector

The operating temperature of the detector is dictated by the level of dark current, which can reduce the useful dynamic range and significantly increase the detection noise. A temperature of 50K is a typical choice for 10 to 12 µm TIR bands.

Column readout circuit

Column decoder

2 stitched 400x800 pix arrays

25 µm pixel pitch

Column readout circuit

Column decoder

2 outputs

2 outputs

4 Final Report


4.3.3.5 Mechanical architecture The mechanical configuration is driven by the Cassegrain telescope with 1.5 m diameter primary mirror (M1), as shown in Figure 4.3-11).

Figure 4.3-11: Overall PLM configuration

The M1 mirror is mounted on the top side of the MIP (Main Interface Plate), whereas the bottom face carries all the focal planes and associated optics. The secondary mirror (M2) is supported by a spider attached to an hexapod structure which also carries the 2.5m long baffle. This configuration minimises the obscuration and provides a high dimensional stability between the M1 & M2, which drives the telescope optical quality.

The instrument is interfaced with the SVM through an hexapod allowing high mechanical and thermal decoupling with respect to the platform.

4.3.3.6 Thermal control The thermal control of the instrument is rather simple because Sun illumination of the interior of the telescope, and in particular the M1 mirror, is avoided by a Sun avoidance manoeuvre when the Sun-to-LoS angle reaches 30 deg, i.e. +/-2h around midnight at equinoxes. This interrupts the imaging sequence (anyway limited to IR bands during night time). The Geo-Oculus telescope thermal architecture is therefore very classical and based on proven concepts & technology, with a combination of passive and active thermal control. The telescope is protected against Sun and cold space by the baffle and the focal plane & external structures are isolated thanks to MLI. Thermal washers are used to decouple the various assemblies and the whole PLM from the SVM. Active thermal control with heaters & thermistors is used to control telescope temperature, though radiative screens on the back of M1 & M2 mirrors and by direct conductive coupling for other elements.

The temperature of the mirror will be very stable during daylight (6h to 18h), when the Sun aspect angle is larger than 90°, i.e. does not illuminate the inner baffle. This corresponds to the phase of UV-

GEO-OCULUS-3

2320 mm

GEO-OCULUS-1

2931 mm

2342 mm

646 mm

4 Final Report


VNIR imaging, where the best accuracy in pointing, defocus and WFE is required.

During night time, the illuminated part of the baffle generates a disturbing flux on the mirror, the resulting thermo-elastic deformations which could generate defocus and wave front error (WFE) are minimised thanks to the high conductivity of SiC material. Moreover, the response of the mirror to this smoothly-varying flux is quick thanks to the combined effect of the high SiC conductivity and the low mass-to-area ratio of the mirror. Thermo-elastic distortions experienced during the night time are therefore not affecting high resolution daytime imaging performances.

The required radiometric performance implies a temperature stabilised environment for each of the detectors, with the following operational temperatures: 50 K for TIR sensor, 130 K for MWIR and 20°C for UV & VNIR detectors. While the obvious solutions are passive cooling for UV & VNIR, and active cooling for TIR, the MWIR sensor could be in principle controlled with one or the other technique, provided that a sufficient radiating area with full view to cold space can be implemented. This is however not possible for the selected dual wing spacecraft configuration, since solar arrays are in view of possible radiating areas on the north & south walls.

The three CMOS detectors and their proximity electronics are cooled by coupling with a small radiating area (0.06 m²) through conventional heat pipes. IR focal planes are housed in cryostats (single stage for MWIR, two stage with intermediate enclosure at 150 K for TIR) and cooled by mechanical cryocoolers. Coolers can be selected among several European products (see Figure 4.3-12), with two candidate technologies, Stirling-cycle coolers or pulse tube coolers. Astrium UK Stirling coolers are proven devices flown on numerous missions. Pulse tube coolers are completing space qualification and should be fully mature for Geo-Oculus. This technology is selected to minimise the number of units (Stirling coolers have to be operated in back-to-back pairs to avoid excessive vibrations) and therefore the mass and complexity. Three redunded coolers are necessary, two Miniature Pulse Tube (one for MWIR and the other for TIR outer enclosure) and a Large Pulse Tube for 50 K TIR enclosure.

Manufacturer ASTRIUM-UK ASTRIUM-UK AIR LIQUIDE AIR LIQUIDE

Model 50-80 K Miniature

Pulse-Tube Large Pulse Tube

Cooler (LPTC) Miniature Pulse Tube

Cooler (MPTC) Type Stirling cooler Pulse Tube Pulse Tube Pulse Tube

Performance 1850 mW à 80 K >> 3 W à 130 K

1400 mW à 80 K >> 2,5 W à 130 K

6 W à 80 K 1300 mW à 80 K

Mass 7,3 kg / cooler 6,4 kg / cooler < 8 kg / cooler < 6 kg / cooler

Figure 4.3-12: Air Liquide Miniature Pulse Tube Cooler (left) and Astrium UK 50-80 K Stirling cooler

(right)

4 Final Report


4.3.3.7 Electrical architecture The electrical architecture is the essentially the same for all five focal planes:

• The focal plane comprising the detector array (for and ) and the Proximity Electronics Module (PEM) housing all the functions requiring to be located close to the detector

• The Remote Electronics Module (REM) hosting the other functions specific to each focal plane and providing the interface with the spacecraft data handling system. In order to minimise the power dissipation on the PLM, the REM units are implemented on the SVM.

All detection chains are connected to a data bus interfacing with the spacecraft central processing unit and the data downloading function. This modular architecture with an independent detection chain for each focal plane allows flexibility in the PLM design and ensures robustness of the mission to a failure. The control electronics (for thermal control and activation of calibration devices and filter wheels) can be hosted in one of the REM as depicted in Figure 4.3-13 or in a dedicated electronics unit.

TM/TC & data bus interface with SVM

Telescope

PAN optics UV-VNIR optics SWIR/MWIR optics TIR optics

Calibration Thermal control

CMOS array

PEM

UV-Blue FPA

UV-blue REM

Memory

Video chain

Power supply

CMOS array

PEM

Red-NIR FPA

Red-NIR REM

Memory

Video chain

Power supply

Hybrid array

PEM

MWIR FPA

MWIR REM

Memory

Video chain

Power supply

Hybrid array

PEM

TIR FPA

TIR REM

Memory

Video chain

Power supply

CMOS array

PEM

PAN FPA

PAN REM

Memory

Video chain

Command/control & data processing

Power supply

Cryocooler Cryocooler





Figure 4.3-13: Electrical architecture of the instrument

4 Final Report


4.3.3.8 Calibration Challenging absolute (Goal: 1%, Threshold: 2%) and relative (0,2% inter-band) radiometric accuracy requirements impose careful calibration of absolute and inter-band offsets and gains. The calibration process shall enable to recover the scene reflectance values from instrument measurements. This implies a complete calibration of radiances/irradiances according to following calibration process: • Corrections of detectors systematic errors (Offset, Dark signal, Dark Signal Non Uniformity

(DSNU), Pixel response non uniformity (PRNU), Dead/bad pixels)

• Correction of absolute value and variation of the overall detection gain (optical transmission, detector efficiency, electronics gain).

• Possibly, stray light correction based on ground characterisation

To reach the above accuracy, in-orbit calibration is required, using a calibration device with properties well characterised on the ground and stable over the mission lifetime.

For PAN & UV-VNIR channels, a retractable Sun diffuser will be used. The preferred candidate diffuser technology are QVD (Quasi Volumic Diffuser) and perforated plates for their low sensitivity to GEO environment. Since sighting the Sun with the instrument is not possible for thermal reasons and full pupil diffuser implementation at telescope entrance is not feasible because of large aperture, two complementary Sun diffusers are proposed:

• A full pupil diffuser implemented near the intermediate focus, sighting the Sun through a window in the baffle. This diffuser does not monitor M1 & M2 possible degradation.

• A small pupil diffuser implemented near the M2 spider to calibrate M1 & M2 transmission (local degradation of the mirrors outside the small pupil area covered by the diffuser is not monitored).

In both cases, calibration is performed during the 4h interruption of measurements around midnight, thus do not interfere with the mission imaging capability. The Sun avoidance manoeuvre performed to keep the Sun-to-LoS angle larger than 30 deg is used to sequentially orient each Sun diffuser towards the Sun. The Sun diffuser LoS (defined by the window in the Sun shield) the has an offset from the instrument LoS equal to the magnitude of the Sun avoidance manoeuvre, and the spacecraft is rotated about the instrument LoS, as illustrated in the following figure. Figure 4.3-14: Geometry of Sun diffuser sighting during Sun avoidance manoeuvre

For IR channels, calibration can be made against two stable radiometric references, a small black body mounted on a flip-flop mechanism or periodic sighting to the cold space

GEO Sun diffuser βsun

θman − βsun

North

Equator

4 Final Report


4.3.4 PLM budgets The PLM budgets are computed for the baseline configuration with the following assumptions:

• A 1.5 m diameter instrument with 5 focal planes.

• Dual wing solar arrays, which is less favourable for thermal control efficiency because radiators on the NS walls have a reduced viewing factor.

• Conventional 2.5 m baffle and Sun avoidance manoeuvres around midnight to keep the LoS-to-Sun angle larger than 30 deg, i.e. preventing that Sun enters the telescope.

• Remote Electronics Modules used for digital processing of the images are implemented in the SVM and accounted for in the SVM budgets.

The PLM budgets are provided in the following figure:

Mass Power

Best estimate 505 kg 423 W

Margins: 20% 101 kg 85 W

TOTAL with margins 606 kg 508 W

Figure 4.3-15: Geo-Oculus instrument interface budgets

2930 mm

2340 mm

650 mm

4 Final Report


4.4 Line of Sight (LoS) Stabilisation Concepts

4.4.1 LoS stabilisation main issues: microvibrations and post-integration The LoS pointing stability requirements are very stringent for the aspects of the mission involving high resolution imaging. This is particularly the case during PAN imaging for disaster monitoring, since the 20 m resolution over Europe corresponds to 0.28 µrad nadir resolution, and to a lower extent when acquiring full resolution VNIR images (40 m resolution over Europe, 0.56 µrad nadir resolution) for disaster or fire monitoring. Image quality is then highly sensitive to LoS motion over the short integration time (up to ~100 msec) required to meet the moderate SNR requirements.

On the contrary, the Marine applications, with a resolution relaxed to 80 m over Europe (i.e. 1.1 µrad nadir), are less sensitive to pointing stability over the image integration time. Nevertheless, since several images need to be post-integrated to reach high SNR requirements, image quality is more sensitive to pointing drifts over the total image acquisition time (several sec).

LoS Stabilisation requirements derived from instrument design are summarised in the following table:

RPE: Relative Pointing Error (stability over the integration time)

0.15-0.2 µrad peak-to-peak for high frequency jitter (>10 Hz)

RME: Relative Measurement Error (over image acquisition time)

0.1 µrad over 5 s max. acquisition time

PDE: Pointing Drift Error (drift over integration time)

Marine applications: 10 µrad/s Fire/Disaster monitoring: 5 µrad/s

Such specifications can not be met without proper management of the LoS pointing stability issue. Depending on the type of disturbances that challenge the LoS stability requirements, and in particular depending on the frequency band affected, different solutions might be proposed for Geo-Oculus :

• High frequency perturbations, with period lower than typical integration time (100 msec), i.e. frequency > 10 Hz, require disturbance reduction techniques. Such disturbances are mainly due to microvibrations generated by moving parts (e.g. reaction wheels and cryocoolers).

• Medium frequency disturbances (with period in the range of a few sec, corresponding to the time to acquire an image based on accumulation of several shots) require image processing techniques to enable post-integration. Such disturbances are mainly due to solar array flexible mode excitation after slew manoeuvres.

• Low frequency disturbances are handled by the AOCS for those observed by the attitude sensors and by ground-based processing (so-called INR, Image Navigation & Registration) for LoS pointing errors due to orbit errors and thermo-elastic distortions between LoS and AOCS reference.

Therefore, the LoS stabilisation issues to be addressed are:

• Microvibrations: level reduction through careful selection of actuators and identification of potential other disturbances.

• Post-integration: Evaluation of the technique to be used to estimate the shift between one image and another one before adding them.

4 Final Report


4.4.2 Microvibrations

4.4.2.1 Candidate mitigation actions The following strategies are defined to limit the impact of microvibration on LoS stability:

• Stopping cryocoolers during PAN imaging. This solution is successfully used in orbit for high accuracy LEO Earth observation. Since switching on and off would not be acceptable in terms of number of electronic cycling, the cryocoolers are in fact kept on, but the amplitude of the engine is simply turned to 0 during imaging and then turned back to full power, without any ageing effect. The drawback of this technique is that thermal control of cold detector is effectively turned-off, and its temperature raises by less than 1 K/s. In the case of GEO-Oculus, since the PAN image acquisition is very short (0.4 s), the temperature raise shall be much less than 1K. These very small temperature cycles are deemed to be acceptable for the detector.

• Elastomeric suspension to isolate the spacecraft from microvibrations generated by cryocoolers

or conventional ball-bearing reaction wheels (BBW). Elastomeric mounts sustaining launch efforts without clamping have been developed and qualified for reaction wheel isolation and will be flight proven with Pleiades in 2009. With suspension frequency around 15 Hz, elastomeric mounts allow efficient attenuation of disturbances above ~50 Hz, where major BBW harmonic disturbances are reported. They are also efficient for high order harmonics which dominate cryocoolers disturbances when the main disturbance at cooler rate (40 to 50 Hz) is cancelled by design (back-to-back Stirling coolers or pulse tube technology). This is therefore the most mature solution to drastically reduce the high-frequency components of BBW & cryocoolers disturbances.

• Magnetic Bearing reaction Wheels (MBW) is a reaction wheel where no mechanical contact

between moving parts is established during normal operation. This is achieved by magnetic levitation and position control of the rotor. The direction of the rotation axis can be actively controlled within certain limits by adjustment of the magnetic fields. This feature allows creating relatively high torques perpendicular to the wheel rotation axis. Hence, a MBW can be used for limited agile slewing manoeuvres. MBW are known to generate much less perturbations than their ball-bearing equivalent. With the availability of such equipments, the resulting high-frequency micro-vibration at instrument level should be reduced.

In the following sections, the two reaction wheel options (MBW and BBW + elastomeric isolator) are compared in terms of microvibration disturbance levels and technology maturity. Cryocooler microvibrations are assumed to be mastered by the combination of elastomeric suspension and cryocooler stop during the most jitter-sensitive phase, PAN imaging.

4.4.2.2 Magnetic bearing wheel (MBW) evaluation for Geo-Oculus The magnetic bearing wheels envisaged for Geo-Oculus corresponds to the new design of Rockwell Collin’s Teldix (RCT) wheels. The currently existing MBW has the status of a technology demonstration, with drive and control electronics located outside of the wheel, only sensor electronics placed inside. In the flight design, the complete electronics equipment will be put inside the wheel.

Table 4-3 compares the technical data of Teldix's 15 Nms BBW to the data of the prototype MBW and the foreseen data of two future flight MBW: MWI 30-400/37 is the basic model adequate for Geo-

4 Final Report


Oculus, whereas MWI 100-100/100 has a bigger rotor and different motor design.

Prototype MBW

Figure 4.4-1: MBW characteristics and photography of the prototype MBW (courtesy of RCT)

The microvibration levels generated by the prototype MBW wheels have been characterised in 2007 by EADS Astrium GmbH in the frame of the DLR study “High Precision Attitude Control of Earth Observation Satellites”. The results of this study show that the MBW disturbance levels are lower by a factor of 10 to 240 (depending on frequency and wheel rotation rate) than typical BBW levels. However, it shall be noted that the comparison is supposing hard-mounted wheels, whereas a BBW mounted on an elastomeric suspension would be the actual competitor for a high accuracy pointing mission. Moreover, microvibrations is analysed at the source, whereas its impact on LoS at PLM level, largely dependent on structure transmission is the relevant parameter. It is undoubted that the microvibrations will be lower, but the above ratios shall not be taken as granted.

The MBW appear in all cases as a good candidate for the Geo-Oculus mission, due to its inherent low microvibration content. The low maturity level (TRL ~4) and the associated development and technical risks shall also be accounted for. The pre-development needs to be actively pursued to reach TRL 5 at the beginning of phase C/D.

4.4.2.3 Ball Bearing Wheel (BBW) option A second option is to use standard ball bearing wheels mounted on elastomeric mounts developed for LEO observation missions. This option has been analysed in 2007 in the frame of the CNES study “Constraints for High resolution observation on GEO”. A microvibration analysis was performed for three different structural transmissions, without elastomeric suspension, with a 15 Hz suspension, and with a 30 Hz one. Microvibration levels measured on 8 flight models of Pleiades BBW (18 Nms Teldix RSI) are used as input to the structural model, typical of a GEO spacecraft equipped with an optical payload for Earth observation. The results are post-processed so that to show the peak-to-peak variation of the LoS over an integration time of 70 ms. The results of this study are therefore directly relevant for Geo-Oculus.

4 Final Report


Figure 4.4-2: Structure FEM used for the BBW microvibration analysis and computed stability over 70

msec for a typical 2-day wheel rate profile

The following conclusions are drawn from this analysis. First, elastomeric suspension is mandatory since it allows a reduction by a factor 100 of the high-rank harmonics perturbations. Second, the elastomeric suspension frequency shall be set to 15 Hz and the wheels velocity shall be maintained below 45 Hz (2700 rpm) thanks to adequate wheel off-loading process at AOCS level. Then, conservatively considering a linear summation of the harmonics and the worst wheel FM, the worst case performance over a typical wheel velocity profile is 0.24 µrad peak-to-peak over 70 ms. This is above the Geo-Oculus requirement (0.15-0.2 µrad peak-to-peak), but the impact on the achievable resolution would be limited, with a degradation of the nadir ground resolution from 10.5 m to 11.5 m. Of course, the level of this preliminary analysis cannot give commitment on these figures, but the order of magnitude is believed to be valid.

Therefore, the use of BBW for Geo-Oculus shall not been ruled out by microvibration aspects. Furthermore, the technology is fully qualified, flying on previous programmes.

4.4.3 Post-integration

4.4.3.1 Principles Post-integration consists in taking several successive images of the scene with short integration time and to add them together to obtain long-integration images, i.e. with high SNR. Keeping the integration time small is necessary for avoiding pixel saturation and for relaxing LoS stability when drift is the dominant error. Post-integration is of no use for micro-vibrations mitigation and applies only to drift mitigation, by reducing the integration time. One option is to perform this post-integration on ground, but this dramatically increases the downlink data rate since up to several tens of images are required on the most critical channels (e.g. for marine applications with high SNR requirements and low reflectance). On-board post-integration is therefore the baseline for Geo-Oculus, based on the of the experience gained on COMS satellite developed by Astrium for Korea.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20

0.05

0.1

0.15

0.2

0.25

Temps (jour)

Sta

bilit

é su

r 70

ms

(µra

d)

Performance en fonction du temps

FM09FM06Pire cas

4 Final Report


If the LoS motion during the summation of the successive images is large (typically > 1 pixel), the motion must be compensated for by shifting the pixels. The simplest correction is the so-called "nearest pixel motion compensation", where the shift is limited to an integer number of pixels, simply achieved by a shift in memory and an accumulation. The average MTF loss at the Nyquist frequency is 0.64 on the accumulated image, i.e. equivalent to a shift of one pixel over the whole accumulation. To reduce this significant degradation of the MTF, refined offset correction methods based on pixel interpolation are possible, but not retained for Geo-Oculus for their low maturity and the required large on-board computation and storage capabilities.

The first step is however to measure the LoS motion between two integration phases, with an accuracy significantly better than half a pixel (0.2 pixel i.e. ~0.1 µrad).

4.4.3.2 LoS drift measurement The LoS motion information can be extracted from gyroscope measurements, provided they are mounted close to the focal plane. The following figure shows the LoS drift estimation error for two high accuracy gyros (Pleiades Astrix 200 FOG and SIRU HRG): Over the maximum image acquisition time (5 sec), the error is 0.3-0.5 µrad, well above the 0.1 µrad requirement. Gyros are therefore not adequate for LoS drift estimation.

0,0

0,1

0,2

0,3

0,4

0,5

0 1 2 3 4 5Time (secs)

Gyr

o dr

ift (µ

rad)

ASTRIX200SIRU

Figure 4.4-3: LoS drift estimation accuracy using gyroscopes

In the case of GEO-observation with a staring instrument, the motion information can also be extracted from the image itself, which removes the need for additional motion sensor. The principle is to correlate in real-time on board the spacecraft the incoming image with the accumulated image, so as to determine the relative image to be corrected. Either the full image or vignettes of interest are used. In the first case, the processing load is high, but the algorithm is simple (correlation over a small moving window) and repetitive, which is well adapted to FPGA or ASIC implementation. In the latter case, the system shall identify vignettes of interest within the first image using an algorithm detecting areas with contrasted variations. Then the correlation is performed between the selected vignettes extracted from the accumulated & current image.

Such techniques are actively investigated at Astrium, primarily for on-ground processing to improve image quality without relying on pre-defined landmarks. The resulting accuracy of image correlation is in the order of 10% to 20% of a pixel, that is to say below 0.03 to 0.06 µrad, well with in the 0.1 µrad requirement. Image correlation is therefore the selected approach for LoS drift measurement.

4 Final Report


4.5 Satellite

4.5.1 Configuration The basic features of the suggested baseline configuration as derived by the trade-offs documented in [RD 7] are as follows:

• dual wing steerable solar array, • payload on top panel (launch) / Nadir panel (operation), • no yaw flip manoeuvre.

The selected spacecraft configuration based on above inputs has been influenced by the choice of the favourable design of the chemical propulsion system as well as by the heritage available from similar projects. The resulting external configuration of the spacecraft in stowed condition is shown in Figure 4.5-1.

The payload is connected to the platform via 3 bi-pods providing iso-static mounting conditions. The configuration of these bi-pods has still to be iterated and the provision of suitable hard-points in the platform, accordingly.

The solar array wings are stowed on the side panels which correspond to the North and South panel during operational mode.

The PDT antenna system comprises a deployable boom and is also folded to a side panel. The deployable boom is needed in order to provide visibility between antenna and ground station which is challenged by the big payload and, moreover, by the manoeuvres which point the complete spacecraft to the scene of interest.

A similar problem concerns the S-Bd antenna which is roughly Nadir oriented (the complementary Zenith oriented antenna is hidden behind the spacecraft). Since this S-Bd antenna needs a hemispherical field of view, a mounting on the payload close to the entrance of the big baffle has been selected.

This accommodation of platform equipment on the payload does not seem to be necessary for the infrared Earth sensors (IRES). Their field of view requirement is about 20 degree half cone angle and may be provided by putting these sensors on a pedestal.

The situation is even a little more comfortable for the star trackers. They shall be aligned as close as possible with the payload line of sight direction in order to achieve the maximum attitude knowledge accuracy. However, they must also consider a certain Earth exclusion angle which yields an off-Nadir viewing direction. This together with the even narrower field of view (~15 degree half cone angle) allows to accommodate the star trackers directly on the top panel. This position may be revised in a later phase since the neighbouring payload radiator may evolve and the contribution of thermo-elastic deformations to the pointing knowledge budget may be optimised by mounting the star trackers directly on the payload.

4 Final Report


InstrumentX 3500 mmY 2300 mmZ 2300 mm

LAE

S-BdAntenna

S/C BodyX 2450 mmY 2600 mmZ 2200 mm

SpaceEnvironmentSensor

PDTAntenna

S-BdAntenna

StarTracker

IRESSensors

Figure 4.5-1: Stowed Configuration

Figure 4.5-2 depicts the deployed configuration of the suggested baseline concept for Geo-Oculus.

The upper solar array wing is oriented towards the North direction and the lower wing towards the South direction. The wings are steerable around the North – South axis thereby always providing an optimised sun inclination angle.

The PDT antenna is also deployed to achieve a sufficient clearance between the antenna field of view and the payload. A 2-axis antenna pointing mechanism is installed directly under the antenna dish which is foreseen to compensate the manoeuvres for acquiring a new scene.

4 Final Report


StarTracker

IRES Sensors

PDT Antenna

Solar Array

Figure 4.5-2: Deployed Configuration

Finally, in Figure 4.5-3 the side panels of the spacecraft body are folded away such that the internal arrangement of equipment becomes visible.

All equipment is spread over the North panel (on the left) and the South panel (on the right). These are the preferred locations since the conditions for heat rejection are optimum there which is beneficial for the sizing of the thermal control system. Since the available mounting surface of these panels is comfortable also in view of a later optimisation of the spacecraft balancing, no equipment needs to be mounted on the East and West panels. Hence, the East and West panels are designed as light weight closure panels.

As a conclusion it can be stated that for the overall spacecraft configuration no major criticality has been identified. All design concepts applied are based on heritage thus providing a solution with low risk.

4 Final Report


ADE5

SCU

S-Bd Transponder

Coarse IMU

Prop. Tanks

He-Tank(in central tube)

Instrument

IMU Electronics

InstrumentElectronics

BatteryPSR

PDHTElectronics

SADM

SPU

ReactionWheels

PLIU

IMU

StarTracker

SADM

Figure 4.5-3: Internal Configuration

4.5.2 Electrical Architecture The electrical architecture satisfies the need for high reliability and availability. A low risk approach is followed which means that heritage from previous projects is used whenever possible. Geostationary heritage can be derived from the Eurostar platform series and specifically from the COMS satellite which already implements meteorological and communication services on a 3-axis stabilized geostationary satellite, similar to the GEO-Oculus mission.

Minimum complexity is achieved by separation of the electrical architecture into functional modules consisting of elements independent of the mission needs and customized elements which are tailored for mission specifics especially in the payload section. As such, the architecture references already the hardware breakdown as used on Eurostar for the bus elements while still giving flexibility for mission specific adaptations on platform side.

High autonomy and reliability is satisfied by the redundancy concept within the overall electrical architecture and the on-board computer. It provides redundant modules and bus systems thus minimizing the amount of single point failures. Autonomy is also a key requirement on telecommunication satellites and therefore inherently available.

Growth potential is achieved by a scalable electrical architecture. This is given by a scalable solar array in terms of amount of panels, regulator stages, battery size and regulator capability as well as adaptability of the amount of command and data handling interfaces from and to the on-board computer.

For the payload side, a mission specific architecture is necessary due to the data rates of the instrument. Therefore, MIL-1553 bus has been selected for instrument TM/TC between SCU

4 Final Report


(Spacecraft Computer Unit) and instrument and high-speed interface for the instrument data to be transferred to the PDH. Cross-coupling is achieved by redundant Milbuses and cross-coupled interface to the PDH. Since G-Link is now obsolete, newer technology such as Aeroflex UT54 series may be more appropriate.

The proposed electrical architecture is given in the following figure. The payload subsystem contains an Instrument Control Unit (ICU) for self-standing instrument mode control and to facilitate instrument testability.

Figure 4.5-4 Geo-Oculus Electrical Architecture (Overview)

4.5.3 Power Subsystem The Electrical Power System (EPS) shall serve the satellite in sun and eclipse with the required power as derived from the power budget. The following essential EPS sizing requirements apply:

• Life time: 10 years • Orbit: Geostationary • Maximum eclipse duration: 72min • Unit margin: between 5% to 30% pending maturity • System margin: 10% at EOL plus 10% time margin on energy recharge

The main functions of the electrical power systems are

• generation of electrical power with a photo-voltaic solar array • storage of excess energy during sun phases into the battery • safe distribution of the electrical power to the on-board users • protection of the battery against overcharging and deep discharging • fully autonomous operation.

The EPS consists of solar array, battery, Power Shunt Regulator (PSR), battery charge and discharge regulators as well as bus distribution and protection. Two one axis drive mechanism provide optimum

4 Final Report


sun orientation for each of the two wings of the solar array.

Power Budget Solar array and battery sizing assumptions:

• Solar array temperature: 57°C • Summer solstice EOL • Solar array degradation for required life time • Unit power figures include maturity margins and an overall system margin of 10%.

Power sizing The instrument power value is based on a mean figure without sun avoidance and a dual wing solar array configuration. For the other instrument/satellite alternatives (dual wing solar array with sun avoidance or single wing with/without sun avoidance) the solar array and the battery will be slightly smaller.

This gives the following results:

• Average load power: 1800W • Required solar array area: Minimum 10.1 m2 • Battery size: 135Ah (11 string 3 parallel configuration) • Battery mass: 48kg

Power Profile GEO-Oculus

0

500

1000

1500

2000

2500

396 796 1196min

W

0,0010,0020,0030,0040,0050,0060,0070,0080,0090,00100,00

SoC

Power SA W Power profile Load W Battery SoC

Figure 4.5-5: Power Profile

The solar array area includes 3.5% margin for string failures.

The battery is based on G5 technology. Cell losses are covered by two additional modules (parallel cells). The maximum battery DoD is approx. 62%. The battery sizing is based on nominal in-orbit operations.

4.5.4 Payload Data Handling and Transmission

Data rate assessment For the sizing of the PDHT, an average data rate of 250Mbps coming from the instrument is assumed.

4 Final Report


This is based on a worst case data rate scenario with the following assumptions:

• Post-integration with (simple) image motion compensation is done on-board (necessary for the UV-blue channels of the marine applications with up to 30 images to be summed up).

• Due to the required image summation, 18 bits per pixel are assumed to provide an adequate quantisation. For the sake of a simple and reliable algorithm, the 18 bits per pixel are assumed for all channels even if no image summation is required.

• For the panchromatic channel VNIR7 required for the disaster monitoring mission, it is assumed that 4 images are down-linked which may then be subjected to motion compensation and post-integration with a more accurate and sophisticated algorithm.

• 2 x 2 pixel binning of the UV-blue and red-NIR channels of the marine applications has not been considered for the data rate.

With these partly conservative assumptions, an evolution of the data rate throughout the next study phases may be compensated such that no change of the selected data transmission technology becomes necessary. For details about the choice of the location and technology of post-integration and image summation, see Ref. [RD 8].

The instrument data are routed via cross-coupled high rate serial interface to the PDH. In the PDH the data are buffered and formed to a continuous data stream with formatting (CADU generation), RS encoding and scrambling. The buffer size has to be determined in the coming study phase since it strongly depends on the ratio of average to peak instrument data. The PDH has a fully redundant structure with the input modules, the buffer and the TMFE output modules interfacing with the cold redundant transmission chains. The TMFE outputs provide full cross-coupling to the modulators.

The PDT is based on cold redundant transmit chains with each consisting of modulator and SSPA, followed by a non-redundant chain selection switch and an output filter. In order to reduce power consumption, a high gain satellite transmit antenna of 0.8m has been selected together with a ground station receive antenna of 13m diameter. The satellite transmit antenna has a beamwidth of approx. 3 degree (3dB double sided beamwidth). The peak gain of the antenna is considered in the link budget requiring antenna pointing via a 2 axes pointing mechanism in case of satellite re-orientation.

Payload data handling, modulator, amplifier and antenna are based on existing components or require minor modifications to be suitable for Geo-Oculus. The only exception is the antenna pointing mechanism due to the large amount of operational cycles. It is expected that upgrading of existing designs or delta qualification will be sufficient.

Carrier frequency selection (ITU constraints) The choice of carrier frequency is dependent on a number of technical and regulatory constraints, including the ease of frequency coordination and location of ground station. For the less than 300 MHz required bandwidth proposed, use of X-Band has been chosen as the baseline (maximum bandwidth available in the 8 GHz downlink band is 375 MHz). Use of the currently unused EES spectrum at 26 GHz (Ka Band) could also be feasible, if a ground station is capable of overcoming propagation issues.

Antenna design baseline Due to the high downlink data rate (250 Mbit/s), a High Gain Antenna has been selected as a solution

4 Final Report


for the PDHT system. The need for a body-mounted instrument affects the contact to the ground station. Due to the necessity to transmit while the satellite is repointing, a steering mechanism for the HGA will be required. This will probably involve a boom mounting scheme on the spacecraft. The impact on the spacecraft in terms of pointing disturbance has not yet been assessed. Control of the HGA motion will be within the PLIU.

Other options such as an electrically steerable (phased array) antenna or a Low Gain Antenna (LGA) do not provide sufficient gain for the high data rate and have now been removed.

Modulation and coding scheme As OQPSK has been selected as modulation scheme, standard RS encoding 255/223 is proposed. This results with an instrument output rate of 250 Mbps in an occupied bandwidth of < 300 MHz in X-band. The Reed-Solomon code to be used is in agreement with ECSS standard ECSS-E-50-01A (Telemetry Sync and coding), Chapter 6. A pseudorandomiser could be used to provide sufficient channel symbol transitions and hence improve received symbol lock.

Link Budget Considering a bit error rate of 10-9 and a transmit power of 8W, a satisfactory link margin of 3.9dB is achieved with a ground station of 13m diameter.

Ground station interface Selection of ground station locations would be dependent on the geostationary longitude of the spacecraft, as described in the System requirements document. A location in central Europe such as Italy or Germany would be capable of viewing a spacecraft regardless of whether it is situated at a longitude more than 45 degrees East or West. This would provide some flexibility in choosing a suitable position for the spacecraft. In all cases the spacecraft should be continually more than 10 degrees above the horizon as seen from the ground station in order to provide sufficient link margin.

Additional considerations to the ground station would include the interface to the user segment, including archiving availability, offsite data links, processing centre capability and user access and security. These factors are especially important where mobile terminals could be deployed as part of a network.

The block diagram of the PDT is shown in Figure 4.5-6.

OQPSK-Modulator

Nominal Chain

Data

Clock

Redundant Chain Data

Clock

OQPSK-Modulator

HGASSPAX-Bd

SSPAX-Bd

Figure 4.5-6: PDT Block Diagram

4 Final Report


Reception of the instrument data by mobile terminals is also possible, but affects the necessary on-board transmit power. For example, assuming a 3m diameter mobile reception antenna, the on-board RF power will increase to approx. 50W requiring the use of TWTAs for the amplification instead of SSPAs.

4.5.5 Telemetry and Telecommand General

The S-Band Subsystem provides all classical Tracking, Telemetry and Command (TT&C) services. Ranging and range rate services are implemented according to the ESA ranging standard. Power Flux density requirements are respected during all normal operational phases (except launch).

Transmitter 1(nominal)

Receiver 1(nominal)

Diplexer

Receiver 2(hot redundant)

Transmitter 2(cold

redundant)

NadirAntenna

Transponder 1

Transponder 2

Ranging&

Coherency

Ranging&

Coherency

TM

TM

TC - Data& Clock

TC - Data& Clock

Diplexer

ZenithAntenna

3 dB -Combiner

RHCP

LHCP

Figure 4.5-7: S-Band Subsystem

The S-band communications subsystem consists of two transmitters, two receivers, a 3dB Combiner, two antennas, and RF harnessing. The nominal RF transfer from and to ground will be achieved using a combined receive/transmit quadrifilar helix (QFH) antenna mounted on the nadir side. An identical QFH antenna on the zenith side is used to establish ground contact for off-nominal attitude conditions. While both receivers are running in permanent hot redundancy, one of the two transmitters will be switched on via the Spacecraft Computer Unit (SCU) when required to perform ranging or to downlink the housekeeping telemetry data to the ground station and will, usually, also be switched on prior to

4 Final Report


launch, remaining on until completion of LEOP.

RF signals from the ground stations at Usingen in Germany, or Maspalomas in Gran Canaria, will be received from both receive antennas, superposed in the combiner and routed to both receivers. The first receiver achieving a subcarrier lock will be selected by the SCU telecommand decoder as the ‘active’ receiver.

The nominal transmitter will send the generated RF signal to both nadir and zenith antennas for transmission to ground. In the case of a failure, the nominal transmitter will be deactivated and the redundant transmitter will take over operation.

Telemetry, Tracking and Command

The uplink frequency will be within the 2025 – 2110 MHz band whereas the downlink frequency will be within the 2200 – 2290 MHz band. Coherence will be implemented, applying the turnaround ratio of 221/240, and can be enabled or disabled using appropriate commands.

During TT&C operation, the uplink Telecommand data and the downlink Housekeeping Telemetry data are modulated on to subcarriers with typical bitrates of 2000 bps and 8192 bps, respectively. The subcarriers are phase modulated onto the respective uplink and downlink carriers, together with the ranging signal.

The helix antennas provide hemispherical coverage with a worst case antenna gain of about -3 dBi. The Link Budget has been developed to achieve required minimum link margins, e.g. >3dB nominal TM recovery margin, while fully respecting the maximum Power Flux Density (PFD) requirements.

Link Performances

A sample link budget for S-Band TT&C and Ranging link is shown in Table 4.5-1 below. Basic 223/255 Reed-Solomon encoding has been assumed for the telemetry downlink. The chosen ground station is Maspalomas on Gran Canaria, which has a 15m diameter dish and a reception G/T of 29.2 dB/K.

Table 4.5-1: Assumptions for S-Band TTC Links

Programme: Geo-Oculus Orbit: GEOGround Station: Height: 35800.00 km

S/C Antenna: Elevation: 10 degreesType: Ranging

Amplifier RF Output: 5.0 W Ranging Possible: TRUEDownlink Data Rate: 8,192 s/s Uplink Data Rate: 2,000 s/s

Information Rate: 7,142 bps

Coding Scheme:Modulation Scheme: Subcarrier Type:

AEOLUS LGA (S-Band)

PCM(NRZ-L)/PSK/PM

Maspalomas-1 (S-Band)

Sine Wave223/255 Reed-Solomon Coding

The analysis yields very healthy recovery margins for Telecommand (26.45 dB nom.) and Telemetry (18.65 dB nom.). Power Flux density limits are not violated.

It should be noted that two S-band stations are required for the envisaged spread spectrum ranging method - for an optimal performance, an additional third station should be implemented. Consequently, it makes sense to consider not only Maspalomas but also a second station, such as an S-Band station in Redu, for instance. Table 4.5-2 shows the differences in EIRP and G/T for these two stations. Redu will have a marginally better link than Maspalomas.

4 Final Report


Table 4.5-2: Characteristics of Maspalomas and Redu S-Band stations MAS-1 RED-1 EIRP 72.1dBW 72.5dBW G/T 29.2dB/K 29.6dB/K

4.5.6 Attitude and Orbit Control

4.5.6.1 Introduction The AOCS plays a significant role within the functional process chain to fulfil the demanding pointing requirements of a very high performance mission, both in terms of absolute and relative pointing / pointing knowledge. This overall process chain consists of the payload, the platform (including the AOCS) and the on-ground post-processing (INR).

Based on system pointing requirements and an associated pointing budget for the whole process chain, preliminary requirements for the AOCS have been derived. Out of the various pointing requirements, the following ones are driving the AOCS concept and will be checked in this chapter:

• the absolute pointing error (APE), • the absolute measurement error (AME), • the pointing drift error (PDE) over 100 msec.

A feature special to the Geo-Oculus mission is that the whole spacecraft is turned in order to move to the next image which shall be acquired in a step and stare mode. Based on the findings related to the mission scenario trade-offs, a medium agility is requested from the AOCS concept in order to support a reasonable number of mission products within the dedicated revisit cycles. The allocated budget assigned to this medium agility is 70 sec for the total manoeuvre time between 2 images. The agility itself is driven by the AOCS actuator selection and the overall platform design (moments of inertia, flexible modes). Especially the flexible modes come into play when high torque actuators are used. This is due to a high initial deflection and the related long tranquilisation time in order to reach again the required pointing budgets. Sun avoidance manoeuvres in regular intervals also represent an agility aspect but are not driving the design because the slew times can be reasonably long. In order to reduce the negative impact of flexible modes on the manoeuvre time, an active damping strategy for the solar array modes could be assessed. This is currently kept in mind as a back-up solution but, so far, only the performance of actuators without such a damping technique has been evaluated for this study.

For the assessments performed in this chapter it is assumed that the AOCS architecture to support the various operational modes (transfer, acquisition, nominal operation, orbit maintenance, safe mode) can be established on the basis of existing E/O or telecomm platforms (e.g. Eurostar) in order to benefit from long-time heritage, risk mitigation and cost minimisation. The focus for this study is to select a suitable set of sensors and actuators which fit with the dedicated pointing and agility requirements of the Geo-Oculus mission.

4.5.6.2 Pointing budgets for AOCS The requirements in Table 4.5-3 represent requirements assigned for the AOCS performance (thermo-elastic distortions are covered by a separate budget):

4 Final Report


Table 4.5-3: AOCS 100% pointing requirements Pointing Index

Preliminary Values (100%) Remark

APE ±100 μrad Derived to minimise overlap of neighbouring images PDE 0.5 μrad over 100 ms

1.0 μrad over 100 ms For VNIR7 channel (panchro) For other channels

AME ±100 μrad Currently the same value assumed as for APE

4.5.6.3 AOCS sensors and actuators Given the high performance requirements for Geo-Oculus, only high performance sensors are considered . The main sensor will be a star tracker (STR) which will be operated together with an inertial measurement unit (IMU) in a gyro-stellar estimator set-up (GSE). In the GSE the STR data is combined with the IMU data to benefit from the advantages of both sensors. The STR provides noisy but stable attitude information and the IMU provides low-noise data which drifts over time. The IMU cancels the noise from the STR and the STR cancels the drift in the IMU to a large extent. Several options for STR are available on the European market:

• Sodern Hydra • Jena Optronik Astro APS • Galileo AA

Several options also exist for the IMU selection: • EADS Astrium Astrix 120 HR • EADS Astrium Astrix 200 GEO • Northrop Grumman Scalable SIRU

The baseline STR is the Astro APS and the baseline IMU is the Astrix 200 GEO, as these are the baseline sensors for a reference mission which is similar in many respects. The performance of the three listed STR are similar and the baseline can easily be changed if necessary. For the IMUs there is a clear performance difference between the Astrix 120 on one hand and Astrix 200 and SIRU on the other hand. The performance level of the Astrix 200 and SIRU is necessary to meet the relative pointing requirements. The SIRU is produced in the US and is subject to ITAR restrictions. It can therefore not be selected as baseline.

The attitude control and manoeuvre actuator selection has gone through several iterations, and the current choice stands between using Magnetic Bearing Wheels (MBW) and an Electric Propulsion System (EPS).

• Rockwell Collins MBW • EPS system

The Rockwell Collins (RCD) MBW is the only MBW option available in the European market. It is currently not flight proven but RCD indicates that they will have flight proven models available in 2013. Thales in Ulm, Germany, is currently developing a HEMPT based EPS system called HEMPT 3050. This system is much to powerful for fine attitude control and a theoretical, scaled down microHEMPT thruster has been developed for comparison.

4 Final Report


Table 4.5-4: Comparison of HEMPT 3050 and microHEMPT HEMPT 3050 microHEMPT Force 30-50 mN 100-500 µN Mass flow 1.2 mg/s 12 µg/s

For the final suggestion of preferred actuators for the Geo-Oculus mission, the properties of above options in terms of pointing performance, agility and fuel consumption is checked for attitude control tasks and manoeuvres in the following.

4.5.6.4 Attitude control performance The attitude control performance of Geo-Oculus has been derived from simulations using a high performance attitude control simulator as already applied in a similar project. The baseline sensor suite has been used for simulations for both the MBW and EPS option. Below follows a summary of the attitude performance figures for steady state attitude control and manoeuvres.

MBW system and performance The configuration of a MBW based actuator system is equal to that of a ball bearing reaction wheel system. Four or five MBW can be selected, depending on the impact zero-crossings has on the pointing accuracy. Five wheels are needed if zero-crossings are judged to be of importance, as a five wheel system will not experience zero-crossings even if one of the wheels should fail. A four wheel configuration is the minimum for redundancy, but will have wheel zero-crossings if one wheel should fail.

The worst values of the analysed performances are

• APE 14.1 μrad, AME 11.3 μrad, PDE 0.32 μrad/100ms (all values for 100% probability).

All performance values are better than the requirements of Table 4.5-3.

EPS system and performance: The EPS analysis is based on the recently finished HOPAS-3 study, investigating the use of EPS as the sole actuator on spacecraft in GEO. The study has been done for DLR by EADS Astrium GmbH. The baseline EPS configuration is derived from this study.

The EPS based attitude control system will have a 12 thruster configuration based on the microHEMPT. A HEMPT 3050 based system has been analysed and rejected based on its high fuel consumption (~100 kg over 10 years) and problems with meeting the given minimum lever arm and torque level requirements. Also, the high power consumption of the HEMPT 3050 (1.2 kW per active thruster) does not allow more than 4 thrusters for attitude control to be active at the same time. This impacts the attitude control performance, especially after the completion of manoeuvres, when only attitude control thrusters are used for settling. In comparison, the microHEMPT can operate six thrusters at the same time, with a total power consumption of 0.2 kW.

The attitude control thrusters are configured with four thrusters around each axis, as can be seen in Figure 4.5-8. The special configuration has been developed especially to meet the lever arm and thruster plume direction requirements, and at the same time provide decent torque levels and fuel consumption.

4 Final Report


Figure 4.5-8: EPS thruster configuration

The analysed performance values are


The APE and AME steady-state performances are well below the requirements with the APE being slightly worse than for the MBW solution. The EPS attitude performance is largely a function of the thruster controller dead zone. The dead zone determines the level the torque command must reach before the thruster will start firing. It is there to avoid excessive thruster firing due to noise and must be selected large enough to prevent continuous firing and counter firings. A small dead zone gives higher pointing accuracy whereas a large dead zone reduces the fuel consumption. The PDE steady-state performance for a EPS system is better than that for the MBW system, due to the lower torque exercised on the system from the EPS thrusters.

The power and fuel consumption is acceptable but it may be further decreased by a larger dead zone. This is possible since there is still a good margin to the absolute attitude requirements. Doubling the dead zone from 1/6 of the available torque to 1/3 reduces the fuel consumption by more than 80%. The attitude performance is reduced but still within the requirements.

4 Final Report


Table 4.5-5: EPS fuel and power consumption and system mass for attitude control over 10 years Deadzone size Fuel consumption [kg] Power consumption [kW] System mass [kg]

DZ 1/3 0.7 0.2 61 microHEMPT DZ 1/6 3.9 0.2 64

HEMPT 3050 120 2.3 307

1000 1500 2000 2500 3000 3500 4000 4500 50000

0.5

1

1.5

2

Thruster activation timeline (DZ: 1/6)

Thru

ster

s ac

tive

[-]

Time [s]

1000 1500 2000 2500 3000 3500 4000 4500 50000

0.5

1

1.5

2

Thruster activation timeline (DZ: 1/3)

Thru

ster

s ac

tive

[-]

Time [s]

Figure 4.5-9: Thruster activation timelines for dead zone sizes of 1/6 (top) and 1/3 (bottom) of maximum available torque

The analysed performance values are


A major issue with the EPS based attitude control system is that it is based upon currently non-existent technology which is believed to be available on the European market within the next five to seven years. No major technological showstoppers are identified for the development of a microHEMPT system but should such a system prove itself to be infeasible for use on Geo-Oculus, other technologies such as microHET and FEEPT thrusters can be considered.

4.5.6.5 Manoeuvre performance One of the key issues for Geo-Oculus is the ability to image multiple locations throughout Europe several times per day. The limiting factor for manoeuvrability is the available torque to perform the manoeuvre in the shortest time possible, and the settling time needed after the manoeuvre to reach the required attitude performance again. Manoeuvrability is only required around the x- and y-axis, to

4 Final Report


scan North/South and East/West, respectively. A summary of the manoeuvres required over 24 h is presented in Table 4.5-6.

Table 4.5-6: Manoeuvre summary Manoeuvre [deg] Daytime manoeuvres

over 9 h [-] Nighttime manoeuvres over 15 h [-]

Total manoeuvres over 24 h [-]

Manoeuvre time allocation [s]

0.25 27 0 27 70 0.40 27 0 27 70 2.00 324 720 1044 70

All manoeuvre performance data presented in this section are based on one-axis manoeuvres.

There are several options for manoeuvre actuators. The MBW option is clear, but also two EPS options, using either the HEMPT 3050 or microHEMPT, have been considered. The MBW option uses the same wheels for attitude control and manoeuvres, as the torque output from one wheel is scaleable from 0 to 400 mNm. The wheel drive electronic quantisation is 55 µNm. The EPS can not throttle its output to the same degree and an additional EPS system is needed for manoeuvres. The first option is to use the HEMPT 3050 thrusters for manoeuvres which can produce a torque of ±85 mNm around each axis.

In the EPS analysis another option has been introduced as well. MicroHEMPT for attitude control has a few advantages such as lower fuel and power consumption and system mass, and one obvious drawback: the long resulting manoeuvre time. An overview of the theoretical, time optimal manoeuvre time for the various options are listed in Table 4.5-7. Note that these numbers do not allocate time for a settling period after the completion of the manoeuvre.

Table 4.5-7: Theoretical, time optimal manoeuvre times for HEMPT 3050 and microHEMPT based configurations, and duty cycle over 24 h

Manoeuvre [deg] HEMPT 3050 (30 mN) MicroHEMPT (2 mN) MicroHEMPT (3 mN) MBW (400 mN) 0.25 24.2 s 132.6 s 108.3 s 9.4 s 0.40 30.6 s 167.8 s 137.0 s 11.9 s 2.00 68.5 s 375.2 s 306.3 s 26.5 s Duty cycle over 24 h 84% 463% 378% 33% Est. fuel consumption over 10 years

635 kg 18 kg 27 kg 0 kg

In the following, the complete manoeuvres are assessed which includes the time where the actuators are operated (corresponds to the manoeuvre times of above table) plus the settling time (mainly driven by the solar array) needed to reach the pointing requirements. Only the APE and PDE are evaluated according to the requirements of Table 4.5-3. The absolute measurement error AME shows the same performance before and after a manoeuvre such that this parameter does not need to be checked. From above figures, the EPS options could already be eliminated since the HEMPT solution needs too much fuel (635 kg of noble gas) and the micro HEMPT system is not suitable because of a duty cycle significantly higher than 100% which means that all the required manoeuvres can not be performed in the required time frame. Nevertheless, all of above options are evaluated to give an impression what is feasible with each of the options.

MBW performance The MBW option has the most available torque for both the manoeuvres and the stabilizing after the

4 Final Report


manoeuvre. It also has a close to linear torque response to the commanded torque within its torque limits, something that favours the MBW option over the EPS solution.

In the following table the total times needed for the selected 3 typical manoeuvres are listed. The criterion for the end of the manoeuvre is the fulfilment of the pointing requirements according to Table 4.5-3 where for the PDE the less stringent value has been considered (1μrad / 0.1sec).

Table 4.5-8: Total MBW manoeuvre time, including settling time for APE and PDE over 0.1 sec Manoeuvre [deg] APE [s] PDE 0.1 s [s] 0.25 8 58 0.40 14 129 2.00 63 126

It can be seen that the APE settling time for all manoeuvres is within the allocated manoeuvre time of 70 seconds. However, the PDE settling is far longer due to vibrations of the solar array induced by the manoeuvre. The length of the PDE settling can be possibly reduced by increasing the stiffness of the solar arrays and by lowering the applied torque when performing a manoeuvre, thus increasing the manoeuvre duration. Another option is to increase the simulated damping factor of the solar array, thus reducing the PDE settling time. A conservative value of 0.3% is used as default, but increasing the damping factor to 0.5% reduces the PDE settling from 126 to 74 seconds for a 2 deg manoeuvre. It is assumed that if the manoeuvre is optimized further, it will be possible to reduce all settling times to below 70 seconds.

EPS performance: An EPS based manoeuvre system requires higher torques than the EPS based attitude control thrusters can produce in nominal operations. Therefore an additional set of manoeuvre thrusters are needed.

A manoeuvre system based on the HEPMT 3050 thrusters can produce a torque of ±85 mNm around the x- and y-axis, giving the theoretical time optimal manoeuvre times and estimated fuel consumption listed in Table 4.5-7.

It is also possible to use an additional set of microHEMPT thrusters, in combination with the attitude control thrusters, for manoeuvres. This will cause the manoeuvre times to increase dramatically, as the available thrust force only will be in the range of a few mN. This requires the microHEMPT thruster to be able to operate at both 100 µN and 500 µN. The 100 µN operational mode is used for attitude control and 500 µN for manoeuvres. If the two thruster pairs are operated at maximum force simultaneously, a total of 2 mN will be available. To reduce total manoeuvre time additional sets of microHEMPT thrusters can be added.

When using the HEMPT 3050 configuration the following performance can be achieved.

Table 4.5-9: Total HEMPT 3050 manoeuvre times, incl. settling time for APE and PDE over 0.1 sec Manoeuvre [deg] APE [s] PDE 0.1 s [s] 0.25 18 20 0.40 153 57 2.00 300 160

It can be seen from Table 4.5-9 that the APE settling is the largest problem. Even though the time optimal manoeuvre time for the HEMPT 3050 configuration is below the 70 seconds allocated to

4 Final Report


manoeuvres, the APE settling time is so long that the duty cycle becomes as high as 368%. This only allows Geo-Oculus to perform 27.2% of the required manoeuvres over 24 hours, which is around the same performance as a configuration using microHEMPT can achieve. The 2 mN option can perform 21.7% of the required manoeuvres and the 3 mN option can perform 26.5% of the required manoeuvres, assuming that the low force manoeuvres are so slow that no settling time is needed after the completion of the manoeuvre. In this context it can be seen that some fuel mass can be saved using microHEMPT for manoeuvres.

Table 4.5-10: Total manoeuvre times for HEMPT 3050, microHEMPT and MBW based configurations, duty cycle over 24 h

Manoeuvre [deg] HEMPT 3050 (30 mN) MicroHEMPT (2 mN)** MicroHEMPT (3 mN)** MBW (400 mN) 0.25 20 s 132.6 s 108.3 s 58 s 0.40 153 s 167.8 s 137.0 s 70 s * 2.00 300 s 375.2 s 306.3 s 70 s * Duty cycle over 24 h 368% 463% 378% 89%*

*Assumption, performance currently not achieved in simulations

** Theoretical performance, no settling time included

Overall performance A MBW configuration provides very good attitude control performance and is assumed to be able to meet the manoeuvre requirements with some additional tuning of certain parameters. The EPS attitude control performance is also very good, but the manoeuvre performance is far worse than what is required. By using large EPS thrusters, the actual manoeuvre time is within the allocated time, but the settling time after a manoeuvre is much to long for the 0.4° and 2° manoeuvres due to the low available torque from the attitude control thrusters. This leads to the result that not all manoeuvres can be performed which reduces the mission value. Also, using the HEMPT 3050 causes a high fuel and power consumption. The microHEMPT option for manoeuvres has not been investigated in detail, but it is clear that the fuel consumption will be lower, and that such a configuration will be able to perform as many manoeuvres as a HEMPT 3050 system if the assumption that no settling time is needed after the manoeuvre completion holds.

The MBW option is the clear favourite of the two, and is selected as a baseline nominal mode actuator. The only drawback is that the MBW development is in an early phase, and might not be available as expected in 2013. If indications of significant delays in the MBW development, or shortcomings in the performance surfaces, the EPS option can be considered again. Figure 4.5-10 shows the baseline AOCS configuration and Table 4.5-11 summarizes in which operational modes the various AOCS equipment is used.

The hybrid option discussed in [RD 7] has not been considered further, as it leads to unnecessary high costs and system complexity. The CPS system can perform wheel offloading and East/West station keeping every three weeks, with a total outage of less than 10 min each time. The North/South station keeping is performed twice every year. If the total mass of the spacecraft should exceed the capabilities of the desired launcher, the EPS system can again be considered as it has the potential to lower overall system mass.

4 Final Report


Figure 4.5-10: Baseline AOCS configuration

Table 4.5-11: AOCS equipment and modes Sensors Actuators

Mission phase Coarse GYP Earth IRES Star STR Sun BASS Fine LiASS IMU MBW CPS Transfer & Acquisition On station: Normal mode On station: Station Keeping Safe mode Transfer Safe mode On-Station ( )

4.5.7 Propulsion System The Propulsion System of the mission is composed by:

• Chemical Propulsion System (CPS): it is intended for GTO-GEO Transfer and, in the option without EPS, also for Station Keeping, wheel off-loading and de-orbiting.

• Electric Propulsion System (EPS): it is optional and, if present, it is intended for reaction wheel off-loading or for active pointing (for an AOCS without reaction wheels).

The following table summarises the possible options and the tasks allocated to CPS and EPS.

Table 4.5-12: Propulsion System Options for Geo-Oculus Option 1 2 3 No EPS EPS + MBWs EPS onlyGTO-GEO CPS CPS CPS NSSK + EWSK CPS EPS EPS Deorbiting CPS CPS EPS Wheel off-loading CPS EPS / Pointing Manoeuvres / / EPS

The main propulsion requirements, as deriving from AOCS analysis, are the following:

• Transfer ΔV: 1500 m/s + 50 m/s margin,

4 Final Report


• De-orbit ΔV: 10 m/s • NSSK ΔV 400 m/s, • EWSK ΔV 10 m/s • Wheel off-loading + Rate Damping + Safe Mode = 15 kg (applicable to CPS only)

The main trade-off regarding the propulsion system is about the use of EPS:

• no EPS onboard (propulsion tasks entirely performed by CPS, all attitude control tasks performed by reaction wheels)

• EPS only for reaction wheels downloading (GTO-GEO transfer and de-orbiting performed by CPS, fine pointing manoeuvres for image acquisition performed by reaction wheels)

• EPS for attitude control (GTO-GEO transfer and de-orbiting performed by CPS, no reaction wheels)

In this scenario, the results of the various propulsion system options are then to be analysed in a trade-off analysis at system level, i.e. involving also AOCS and main satellite level design choices.

Therefore, the purpose of this section is just to prepare the input for such trade-off analysis.

In following sections, the mass budgets (main trade-off criteria) for the options in Table 4.5-12 are derived from requirements and briefly analysed.

4.5.7.1 CPS Geo-Oculus is a geostationary mission. Astrium has a long heritage of supplying geostationary spacecraft, dating back to the 1970s. In addition to the telecom fleet there is a successful fleet of scientific and earth observation missions including Mars Express which has achieved two years in Mars Orbit, Venus Express currently in Venus orbit, and Rosetta and Cluster missions which are now flying with bipropellant NTO / MMH propulsion systems. The combined experience of this wealth of heritage shall enable Astrium to complete the study and return conclusions for the optimal CPS to meet the Geo-Oculus mission requirements.

Astrium currently has 3 generic platforms for geostationary missions which can be considered for Geo-Oculus. These are:

Eurostar 2000+

MON-3/MMH bipropellant propulsion system, an evolution of the Eurostar 2000 platform, featuring 4 propellant tanks, on a central cylinder supported structure

Eurostar 3000

MON-3/MMH bipropellant propulsion system, a larger version of the E2000+ platform. The design has been expanded to include larger tanks to increase the mission capabilities of the design, featuring 4 propellant tanks on a central cylinder supported structure, in 4 sizes

Eurostar 3000C

The E3000C is an evolution of the E3000 design, based upon the successful Mars Express and Venus Express spacecraft. It remains a MON-3/MMH bipropellant propulsion system, with heritage from E3000 and Mars/Venus Express, but the platform is smaller than both E3000 and E2000+ to suit a smaller payload requirement. It features 2 propellant tanks are (supported by a “single H” type structure, demonstrated by the Mars/Venus Express spacecraft). As with Eurostar 3000, the tank size is interchangeable.

4 Final Report


The CPS considered for Geo-Oculus is the same of E3000 platform. The variable is given by the propellant tanks capacity.

The CPS main features are:

• He pressurized bi-propellant system (MMH + MON-3) • Four cylindrical tanks / one central pressurant tank • Common propellant storage and feed system • One 450 N LAE • Seven pairs of 10N RCT’s

Figure 4.5-11: Geo-Oculus CPS Schematic

The basic CPS sizing option is performed for option 1 of Table 4.5-12 which is the solution without EPS. The result is presented in the following table.

4 Final Report


Table 4.5-13: CPS Mass Budgets for considered options Budget Element Option 1 Total CPS Dry Mass 162.9 Total residual 26.82 Total CPS EOL Mass 189.7 Useful Propellant load 1793.9 Total CPS BOL Mass 1983.6

4.5.7.2 EPS The Electric Propulsion System (EPS) architecture is deriving from:

• Option 2: existing and available commercial solutions • Option 3: iterations with AOCS definition and is the one used for most simulations of section

4.5.6 (Attitude and Orbit Control)

The necessary iterations with AOCS definitions are needed to optimise on board resources request (power, mass, volume) while fulfilling mission requirements.

The architecture is summarised in the following table:

Table 4.5-14: EPS Architecture summary for the two options, Option 2 and Option 3 Option 2: EPS + MBW Option 3: EPS only Main Manoeuvre Thruster (MMT) 2 main + 2 redundant

+ 2 thruster pointing mechanisms 8 main + 8 redundant

MMT Thrust 80 mN 30 mN MMT Duty Cycle 100% during NSSK

manoeuvres (twice a day) 85% during

re-pointing manoeuvres MMT PCU 1 main+1 redundant, each driving 2

thrusters 2 PCU, each driving 4 main + 4

redundant MMT Fine Pointing Thruster (FPT) / 12 main + 12 redundant FPT Thrust / FPT Duty Cycle / 1% during fine pointing FPT PCU / 2 PCU, each driving 4 main + 4

redundant MMT

EPS for Option 2

The EPS for Option 2 can be based on existing EPS for Eurostar 3000, with an architecture as per Option 2 of Table 4.5-14, using SPT-100 as thrusters (HET type, 80 mN of thrust, 1510 s of Isp).

4 Final Report


FUFUFU

PPU A*PPU B*

XST

FDV

XFC XFC XFC XFC

FDV

Orientation mechanism (PY) Orientation mechanism (MY)

HET2PYThruster

HET1PYThruster

HET2MYThruster

HET1MYThruster

FU

LPT (4)

HPT (2)

Isolation and regulator solenoid valves

Plenum

XRFS

FDV

PV

XEF

VBA

PY TMA MY TMA

TSU TSU

FUFUFU

PPU A*PPU B*

XST

FDV

XFC XFC XFC XFC

FDV

Orientation mechanism (PY) Orientation mechanism (MY)

HET2PYThruster

HET1PYThruster

HET2MYThruster

HET1MYThruster

FU

LPT (4)

HPT (2)

Isolation and regulator solenoid valves

Plenum

XRFS

FDV

PV

XEF

VBA

PY TMA MY TMA

TSU TSU

Figure 4.5-12: GeoOculus EPS Schematic, Option 2, Eurostar 3000 type

4 Final Report


As secondary trade-off, the same architecture can be analysed by using HEMPT thrusters with 80 mN nominal thrust and an Isp of 3000 s.

Following table summarises the results for mass and power

Table 4.5-15: EPS Option 2 Mass Budgets for considered thruster options (figures in kg) SPT-100 80 mN HEMPT TOTAL EPS DRY MASS 92.8 103.2 Total Propellant Load 83.0 41.1 TOTAL EPS BOL MASS 175.8 144.3

Table 4.5-16: EPS Option 2 Power Budgets for considered thruster options (figures in W) SPT-100 80 mN HEMPT Main Thruster(s) Assembly 2404 5553 PCU & Ancillary 227 464 Total 2632 6017 Total (Including Margins) 2895 6619

Although the system based on HEMPT technology is about 20% lighter, it needs twice the power to be run. Decreasing power to match the system based on SPT-100 technology could be achieved by:

• either having a 30 mN HEMPT, but thrust times would increase around the nodes, decreasing the efficiency of firing and thus increasing the quantity of propellant to be used

• or setting the HEMPT at a lower Isp, but being a more massive thruster of SPT for the same combination of Thrust and Isp, this would not be an option to be considered

The final trade-off solution will depend on spacecraft level trade-off analysis.

EPS for Option 3

The EPS for Option 3 is based on the general architecture envisaged for carrying on AOCS simulations:

• 8 Main (Attitude) Manoeuvre Thrusters (MMTs) of 30 mN each, Isp of 3000 s; • 12 Fine Pointing Thruster (FPT) providing down to 0.1 mN of thrust each

The rest of the system has been completed as per Option 3 of Table 4.5-14.

Two MMT options have been considered, namely HEMPT and GIT, while smaller GIT (Isp of 3000 s) and FEEP (Isp of 6000 s) have been considered as FPT, making four possible combinations for Option 3. There is little difference in propellant mass (between 574 and 578 kg); the overall EP mass and power budget are shown in Table 4.5-17and Table 4.5-18, respectively..

4 Final Report


Table 4.5-17. EP mass budget for Option 3 [kg] Main Manoeuvre Thruster HEMPT GIT

GIT 860 902 Fine Pointing Thruster

FEEP 834 876

Table 4.5-18. EP power budget for Option 3 [W] Main Manoeuvre Thruster HEMPT GIT

GIT 2571 2138 Fine Pointing Thruster

FEEP 2600 2167

4.5.7.3 Propulsion System Summary The mass of the propulsion system for the three options is summarised in Table 4.5-19.

Option 3 is not practical since too massive, complex and expensive. Option 1 is the traditional configuration and is feasible. Option 2 allows to save between 250 and 340 kg (depending on the selected technology) on Option 1 by using EP for station-keeping. However, this mass reduction should be somewhat reduced as it does not take into consideration the additional mass due to an increase in solar array as well as batteries, and potentially PCDU too. An in-depth analysis would be required at a later stage to determine the better of Options 1 and 2.

Table 4.5-19. Geo-Oculus propulsion options summary

213821672571260066192895/Power requirements [W]

3014.92964.92935.42889.31642.51737.21983.6TOTAL PS MASS AT LAUNCH

2466.22445.124342412.21405.51487.21793.9TOTAL PROP. LOAD

1888.01869.51857.71838.31364.41404.21793.9CPS Total Prop. load

578.2575.6576.3573.941.183/EPS Total Prop. load

548.7519.8501.4477.1237.0250.0189.7TOTAL PS DRY MASS

194.3189.7189.7189.7133.8157.2189.7CPS dry mass

354.4330.1311.7287.4103.292.8/EPS dry mass

1668.31668.31668.31668.31668.31668.31668.3S/C dry mass (no PS)

GIT+GITGIT+FEEPHEMPT+GITHEMPT+FEEPHEMPTHET

Option 3Option 3 Option 3 Option 3Option 2 Option 2 Option 1

213821672571260066192895/Power requirements [W]

3014.92964.92935.42889.31642.51737.21983.6TOTAL PS MASS AT LAUNCH

2466.22445.124342412.21405.51487.21793.9TOTAL PROP. LOAD

1888.01869.51857.71838.31364.41404.21793.9CPS Total Prop. load

578.2575.6576.3573.941.183/EPS Total Prop. load

548.7519.8501.4477.1237.0250.0189.7TOTAL PS DRY MASS

194.3189.7189.7189.7133.8157.2189.7CPS dry mass

354.4330.1311.7287.4103.292.8/EPS dry mass

1668.31668.31668.31668.31668.31668.31668.3S/C dry mass (no PS)

GIT+GITGIT+FEEPHEMPT+GITHEMPT+FEEPHEMPTHET

Option 3Option 3 Option 3 Option 3Option 2 Option 2 Option 1

4.5.8 Structure and Thermal Concept

4.5.8.1 Structure

Structure design Figure 4.5-13 depicts the structure of the satellite with 4 propellant tanks, based on Astrium’s Eurostar

4 Final Report


3000 platform.

The overall S/C structure has a classical "box" shape with a central cylinder (800 mm) as the main structural load path to the launcher. The design will fulfil the satellite strength and stiffness requirements. The instrument is mounted at the top of the platform.

Currently, the instrument is connected to the platform by means of three isostatic mounts. It is recommended for future work to revisit the instrument mechanical configuration so that it is supported by four sets of isostatic mounts, turned upside down. This would allow the “head” of an isostatic mount to be fixed to the top of the shear walls, and thus provide a better load path than what is currently depicted. Clearly, this topic has to be iterated with the mechanical design of the payload considering the mechanical load path and thermo-elastic distortions.

It should also be noted that due to the width of the instrument being much larger than the central cylinder, it is not possible to fix the isostatic mounts in the current configuration directly to the central cylinder.

Figure 4.5-13. Satellite Configuration showing the main structure

Primary Structure The satellite primary structure consists of,

• A launcher interface ring • A central cone/cylinder structure • 4 shear walls • ±X Upper and lower floors • Upper and lower tank floors • Tank support struts

ZS/C

XS/C

YS/C

4 Final Report


The primary structure design is founded on the E3000 heritage design.

The central structure consists of a lower conical section and two upper cylinder segments. With the exception of the Aluminium Alloy upper ring segment all other parts are made from filament wound CFRP sandwich panel section. Bonded rings are located at each end of the CFRP cone cylinder and at the intersection between the cone and cylinder.

Considering each of these rings the lower ring provides the launch vehicle adapter (LVA) interface, through this ring interfacing with an opposite ring on the launch vehicle, combined with the use of a clamp band, the launch vehicle interface is made. The at the cone cylinder junction supports the tank floors, to this an important part of the central structure, the tank support struts, linking this floor at the tank interfaces to the LVA ring, these provide axial support to the tank. The upper tank floor, which provides lateral support only allowing tank expansion, mounts to the ring at the top of the cylinder.

Focusing on E3000 heritage all shear wall and tank floors are made from Aluminium alloy sandwich panels. If future, more detailed, distortion analyses would show the need for a CFRP panels, the material of the shear walls could be switched from aluminium skin to CFRP skin but retaining the Aluminium alloy honeycomb core.

Secondary Structure The satellite secondary structure consists of,

• ±Z equipment panels and ±Y closure panels • Local support brackets/panels as e.g.

− Connector brackets − Thruster supports − EMC covers − Liquid apogee engine and pressurant tank supports − etc.

Focusing on E3000 heritage the equipment panels are made from Aluminium alloy sandwich panels. Currently proposed is the use of CFRP skinned panels, however if future, more detailed, distortion analyses would show that Aluminium Alloy panels could be accommodated, the material of the equipment panel walls could be switched from CFRP to aluminium skin but retaining the Aluminium alloy honeycomb core.

The panel thickness will be typically 35-40mm. The design and the materials of local support structures will be defined in a later project phase.

Solar Array The solar array substrate will be a lightweight CFRP sandwich panel with typically 20 mm thickness.

Structure Load Paths The circular central structure will collect the individual loads over its height and will ensure a homogeneous load distribution over the launcher interface circumference. Hence the launcher I/F overflux requirement will be fulfilled.

The axial (in-plane) equipment panel loads will be transferred to the central cylinder via the shear webs.

4 Final Report


The lateral in-plane equipment panel loads will be transferred to the central cylinder via the floors in shear. The lateral out-of-plane equipment panel loads will be transferred to the central cylinder via the shear webs and floors in tension/compression.

The tanks are grouped closely around the central structure. Thus the tank load path is very short and mass-effective, the tanks will laterally be supported by the tank floors and axially by struts from the lower tank boss to the LVA ring side.

CFRP Outgassing Water and carbohydrate evaporation in the vicinity of the instruments needs to be limited to avoid e.g. ice on cold optical surfaces. Any critical CFRP surface shall be sealed by an aluminium barrier foil of typically 20 microns. In the past, this was done successfully e.g. on all inner surfaces of the 6.8 m long XMM telescope tube structure.

Distortion Assessment Pointing stability depends on thermal and moisture release distortions and therefore on configuration, material selection, method of construction, changes in average temperature and temperature gradients.

A similar stringent requirement also exists for MTG. It is therefore assumed that, through the high degree of similarity between the two satellites, that the Geo-Oculus environment would yield distortions of similar magnitude. Therefore, it is confidently believed that distortions should not be an issue for Geo-Oculus.

Launch Vehicle Vibrations The Frequency requirements for the S/C hard mounted at the I/F to the launch adapter are taken from Soyuz since it is the worse case compared to Ariane 5:

• 15 Hz in lateral + 15% margin • 35 Hz in longitudinal + 15% margin

Considering stiffness the main driver for the fist axial frequency is tank mass and the stiffness of the underlying support. To optimise stiffness performance the heavy and light tanks are diagonally opposite as to maintain a central centre of gravity position, also the supporting struts are categorised into heavy or light struts each providing the best support stiffness for the respective tank mass.

The first lateral mode is largely driven by the X axis centre of gravity position, this is heavily influenced by the mass of any +X top floor mounted equipments and instruments.

To estimate the first lateral frequencies of the Geo-Oculus spacecraft, the performance of the reported performance of MTG spacecraft has been examined and scaled as appropriate. Scaling has taken account of the comparative propellant mass and instrument mass associated with the Geo-Oculus spacecraft.

The MTG spacecraft is viewed as a suitable foundation given the spacecraft architecture is similar to Geo-Oculus and likewise is largely based on E3000 heritage, the most significant differences is an extra 150kg approx located on the spacecraft +X floor and an extra propellant mass of 178Kg.

The first lateral and longitudinal frequencies are as follows,

4 Final Report


Direction MTG Scaled MTG Requirement Margin

Lateral 19.32Hz 18.86Hz 17.25Hz 1.61Hz

Longitudinal 40.65Hz 40.37Hz 40.25Hz 0.12Hz

4.5.8.2 Thermal Control

Thermal Environment The Geo-Oculus spacecraft will circle the Earth in a geostationary orbit. It is positioned directly over the equator and follows it’s path in the equatorial plane at a speed matching the Earth’s rotation. Thus, the spacecraft completes one rotation around its North / South axis per day.

The sun will traverse through an angle of ±23.5° perpendicular to the orbit plane during the year. The extremes will occur at the Winter Solstice and the Summer Solstice. Eclipses will occur during the Equinox seasons, with maximum eclipse duration of 72 minutes. The North face of the spacecraft will receive direct solar illumination for 6 months centred on the Summer Solstice, while the South face will receive direct solar illumination for 6 months centred on the Winter Solstice. The other faces of the spacecraft will receive varying solar illumination during each day.

The Earth varies it’s distance from the Sun over a period of 1 year. This means that the solar constant at Earth’s location changes over the year from 1420 W/m2 at Winter Solstice to 1327 W/m2 at Summer Solstice.

Thermal Control Concept The Geo-Oculus spacecraft body thermal control will rely primarily on passive means supported by electrical heaters. The North and South faces of the spacecraft are used as the main heat rejection paths. Externally they will be covered by Optical Solar Reflectors (OSR). The exact area of OSRs exposed to space will be regulated by the use of Multi-Layer Insulation (MLI).

The inside of the panels will have a black finish. Aluminium doublers and heat pipes, as appropriate, will be used to spread the heat within the panel. All electronic units are mounted inside the spacecraft primary structure. Heat transfer from the dissipating units to the radiators relies mainly on conduction..

Thermal Performance The total dissipation of the equipments on the spacecraft is 1402 watts. 423 watts is the dissipation of the externally mounted units, leaving 979 watts dissipated within the spacecraft body. The payload electronics, 300 watts, is mounted on the North radiator. In addition, some of the bus equipment will also be mounted on the North panel such that the total amount of heat rejection capability adds up to 479 watts. The rest of the spacecraft bus electronics, 500 watts, is mounted on the South radiator. This allows the calculation of the radiator sizes and the required heater power. The analysis results are shown in Table 4.5-20 and Table 4.5-21.

4 Final Report


Table 4.5-20: Thermal Results for North Radiator North Radiator Dissipation 479 watts Margin 153 watts Total dissipated 632 watts Required radiator area 3.24 m2 Heater power – Equinox sunlight 50 watts Heater power – Equinox eclipse 105 watts

Table 4.5-21: Thermal Results for South Radiator South Radiator Dissipation 500 watts Margin 160 watts Total dissipated 660 watts Required radiator area 3.54 m2 Heater power – Equinox sunlight 88 watts Heater power – Equinox eclipse 139 watts

The North and South panel provide up to about 5 m2 of radiator surface each which leaves sufficient margin for further evolution.

4.5.9 Satellite Budgets

Geo-Oculus Budgets

S/C Mass Dry Mass 1858 kg Launch Mass 3652 kg Power Power Demand 1800 W S/A Size (installed) 11 m2 Battery 135 Ah Propulsion Tanks 4x406 ltr Communication PDT 250 Mbit/sec

4.6 Ground Segment

4.6.1 Ground Segment Architecture The architecture of the Ground Segment for the Geo-Oculus system takes into consideration the heritage of the Agency in operating EO satellites and within this context the consistency of the functionalities and of the implementation solutions with other EO systems operated by the Agency. In particular, the interoperability of Geo-Oculus with other EO systems serving the GMES needs is of paramount importance.

4 Final Report


4.6.1.1 High Level Functional Architecture At the first level of breakdown, the functional architecture identifies two main components:

• The Flight Operations Segment (FOS), • The Payload Data Ground Segment (PDGS).

This breakdown is illustrated and further refined in Figure 4.6-1.

Facilities Ground Segment

• Acquisition• Ingestion• Processing / reprocessing• Archiving and Inventory• Production Requests Handling• Dissemination• Circulation• Monitoring and Control

User Services & Mission Planning

Sensor Performance, Products & Algorithms

• Routine Quality Control• Product Quality Control• Product Calibration• Product Validation• Instrument Calibration• Processing and Instrument Data

Files Generation• Instrument Performance Monitoring• Algorithms and Instrument

Processing Facility Development• Instrument Processing Facility

Maintenance & Evolution• User Support• Precise Orbit Determination

Mission Control System

Flight Dynamics

Spacecraft Simulator

• Platform Model• Payload Model• Ground Segment Model

• Spacecraft Operational Database

• Housekeeping Telemetry Processing

• Time Management• Telecommand• Mission Planning• On-Board Software

Maintenance• Ground Station Network

Interface• Off-line analysis• Authentication and

Encryption

Ground Stations and Networks

• Orbit Determination, Prediction and Control

• AOCS Monitoring• AOCS Command

Generation• Test and Validation

• TMTC Ground Stations• NDIU• PSS• Ground Communication

Network

• General Web • Catalogue • On-line Ordering and Order Handling• User Management • Mission Planning • Help and Documentation Desk • Statistics & Reporting

FOS

PDGS

Figure 4.6-1: Geo-Oculus -Ground Segment breakdown into Domains and Functions

4.6.1.2 Proposed Architecture In the following, the proposed overall architecture of the Geo-Oculus Ground Segment is presented. The single elements of the architecture will be considered in the subsequent subsections.

The Geo-Oculus Ground Segment Architecture features:

For Mission Monitoring and Control:

• A Flight Operations Control Centre, • a nominal TM/TC station, • a back-up TM/TC station and • a network of additional TM/TC stations used only for LEOP.

For Payload Data Reception, Exploitation and Processing:

• A “core” Payload Data Ground Segment and

4 Final Report


• one or several deployable Payload Data Reception and Processing facilities.

The Figure below depicts the Ground Segment Architecture. It also identifies the main internal and external interfaces. The external interfaces comprise the interfaces to the users and additional data sources, which provide auxiliary data for processing and meteorological forecast data supporting the scheduling of observations.

FOS PDGSReceivingStation(s)

TMTCstation

LEOPNetwork X-BandS-Band

S-Band

Backup

EGSE

Service SegmentService

Segment

External Data

Sources

External Data

Sources

External Data

Sources

External Data

Sources

TMTC

Instrument Raw Data

Flight Operations Control Centre

Decryption, Processing, Archiving &

Dissemination

Users

Customised Services

Payload Data Reception


Users ServicesCoordination & Control

UserRequests

Sensor Performance, Products and

Algorithms

Sensor Performance, Products and

Algorithms

Basic Products

External Auxiliary Data

Payload Mission Planning

Meteo Forecast Data + MTG real time data




Dissemination


Dissemination

User Reception and Processing

Terminal

Basic Products

Figure 4.6-2: Preliminary Architecture of the Geo-Oculus Ground Segment

The concept for the GEO-Oculus Ground Segment takes into account the particular principles of operations and technical constraints resulting from a spacecraft on a geostationary orbit. Moreover, the ground segment architecture is adapted to the needs of its customers for the different targeted applications, in terms of revisit time, flexibility in satellite observations programming and latency from observation to end of product delivery. In addition, the Ground Segment concept for the Sentinel missions, which will be operated by ESA in the GMES era, has been considered as a "loose" design guideline.

The mission for an optical high spatial resolution satellite operating from a GEO orbit must be regarded as the conjunction of routine monitoring missions (sometimes termed “background” mission) and of one or several emergency monitoring missions, which are by nature less schedulable than the routine ones. An example of routine monitoring mission is the monitoring of coastal areas for which the revisit times and the response times are comparatively long. Emergency monitoring missions are e.g. fire or disaster monitoring. In this case, the revisit time as well as the response time need to be much

4 Final Report


shorter. As a consequence, the Ground Segment baseline architecture takes into account both types of missions in order to be equally suited for routine type applications and emergency type applications.

In the frame of a study for CNES, it has been envisaged that routine type applications could be served via a “Central” PDGS, due to the fact that (1) they are less time critical than emergency applications and (2) they require some value added processing (via service segment entities) in order to actually satisfy the needs of end users who are not experts in the manipulation and understanding of satellite images. This is compliant with the generic PDGS architecture model with, however, the PDGS providing products up to Level 1b only and further processing up to the delivery of a value added service being performed by specialised value added providers within the service segment.

For emergency applications such as disaster monitoring (fires, floods), the study for CNES privileged an architecture model with specific decentralised user’s facilities for data reception, processing and dissemination. The crisis headquarters would be equipped with a small reception terminal and with the means to at least perform basic processing and dissemination towards the crisis actors on the operations theatre. The rationale for such a decentralised model is that it will provide quicker response times by delivering the data directly on the operations field.

The decentralised model supposes that (1) the crisis headquarters or even the mobile command posts are equipped with all means to process data and (2) the end users can cope with a relatively basic level of processing.

The projects currently on-going in the field of natural risks management do not foresee a direct delivery of basic products (e.g. Level 1b, which could be provided by processing on-board the S/C) to the end users but rather still foresee the involvement of specialised service providers in the value adding chain. To be easily understandable by non experts, the basic products must be geo-referenced and combined with geographical or socio-economic information, and finally be integrated into a Geographic Information System (GIS).

All of these steps might be performed in the spirit of a Service Oriented Architecture (SOA) of the overall GEO Oculus PDGS even in the field. However, they are basically relying on the know-how of the service providers for executing the value adding such as Web Map Services (WMS), Web Feature Services (WCS) or Web Processing Service (WPS). Note that these services are Web-based by principle, such that a high-speed WAN connection is mandatory for this kind of applications.

In addition, the processing cannot be executed without input of auxiliary data such as calibration data, orbit data and attitude data, which in turn requires specific interfaces between the “Core” Payload Data Ground Segment and the decentralised part of the Payload Data Ground Segment.

So, the main advantage of the decentralised architecture model, which is to optimise the response time is counterbalanced by the need for a high-speed WAN connection even in the field (which might as well provide the products generated by the "Central" PDGS) or even the shortcomings of providing only products of low value, possibly usable only by experts.

For more details, see [RD 10].

4.6.2 Geo-Oculus dedicated Ground Segment issues This section addresses specific issues linked to the characteristics of the Geo-Oculus mission, i.e. the fact that (1) the satellite orbit allows permanent contact with the ground both for TM/TC operations and

4 Final Report


for the reception of payload telemetry and (2) the mission is focused on frequent revisit for systematic observations and real time instantaneous access emergency management.

4.6.2.1 Command and Control Geo-Oculus will allow permanent contact with the Mission Control System. One main advantage is that commanding the satellite is possible at any time, without having to wait for a ground station contact. On the other hand, the Mission Control System will receive a permanent flow of Housekeeping telemetry, during day and night. Hence, a reasonable schedule for operating the spacecraft should be established in order to minimise the operations cost.

4.6.2.2 Flight Dynamics As a baseline for the Geo-Oculus orbit determination the spread spectrum ranging method using a S-Band repeaters and S-Band Ground Stations has been selected. The stations will have to include the necessary ranging equipment, and the Flight Dynamics System will have to process the ranging information from the stations in order to perform high accurate orbit restitution.

4.6.2.3 Mission Planning Efficiency of mission planning operations is essential to gain the full benefit of the satellite agility. As far as the routine (or background) mission is concerned, the baseline observation schedule can be established well in advance and loaded on-board the satellite at given times. However, actual meteorological conditions must also be taken into account in order to minimise the likelihood of cloudy scenes and thus useless observations.

For the emergency monitoring mission(s), reactivity is at stake. This means that, as soon as a cata-strophic event requiring fast scheduling of an observation occurs, it shall be possible to superimpose an emergency observation to the nominal plan. As for the routine missions, the emergency obser-vations shall also take into account the meteorological conditions during the mission planning stage.

From a mission planning point of view, simple conflicts management rules can be implemented, which allow for giving priority to emergency observations over routine ones in an automatic manner. Additional conflict management strategies are needed to dissolve possible resource conflicts between different emergency monitoring missions, which may coexist within the same period of time. For handling these situations it will be useful to dynamically assign priority levels to the different emergency events.

Strategy baselines on the implementation of emergency observation requests into the schedule have already been given in [RD 3], including scenarios for combining routine and emergency observations. The straight forward approach to avoid congestion is to allocate only a given percentage of observation capabilities to routine observations, so that in total, enough resources will be available to include both routine and emergency observations.

As can be seen from the above considerations, one major issue in the context of the on-demand scheduling of emergency observations is staffing, unless the whole process could be automated. If not the case, personnel will need to be available both on PDGS and FOS side to take into account and schedule unforeseen requests. For solving this issue, an intermediate approach like for Sentinel-3 Fire Monitoring mission can be adopted, i.e. working during normal working hours only (8/24 5/7) outside periods of natural disasters, and working round the clock during periods when such events are the most likely to occur (e.g. April to October for the fire season in Southern Europe).

4 Final Report


4.6.2.4 Cloud Cover Nowcasting and Very Short Range Forecasting Currently, mission planning for optical satellites uses cloud forecast issued by e.g. Meteo France at 6 hour intervals. Such cloud forecast belongs to the category of “Nowcasting" (NWC) and “Very Short Range Forecasting" (VSRF), which is defined in a very broad sense as “user-driven services using appropriate meteorological and related science to provide information on expected conditions up to 12 hours ahead”, covering inter alia air pollution, ocean, and hydrology at these timescales.

The Satellite Application Facility (SAF) for Nowcasting provides operational services to ensure the optimum use of meteorological satellite data in Nowcasting and Very Short Range Forecasting. Currently the SAF NWC generates so called Cloud Mask Products at three hour intervals.

In addition to EUMETSAT's SAF NWC, there exist numerous public web sites which provide prediction maps of weather, temperatures, and cloud cover at various resolutions and various time intervals. A cloud forecast map is published each hour from current time until 5 days later. However, the reliability of the information is difficult to be verified.

Regarding the perspectives for the timeframe from 2015 onwards, one can obviously consider the new generation of meteorological satellites, i.e. Meteosat Third Generation (MTG). The MTG missions capitalise on the continuation and enhancement of the MSG capabilities.

In addition to the afore-mentioned remote sensing missions, there are also scientific research works that have demonstrated the ability of so called “advanced advection methods” to provide robust short-term top forecasts of cloud motion. The advection technique is based on a cross-correlation algorithm that computes local motion vectors by tracking identifiable cloud features across pairs of time-sequential satellite images. Satellite data are first processed by cloud detection and cloud property retrieval algorithms to identify, classify, and stratify cloudy features by altitude. Cloud information is remapped to a standard map projection and the correlation algorithm applied. If available, NWP winds are used to reduce processing time and to eliminate obviously incorrect motion vectors.

All these elements concur to the conclusion that significant progress for nowcasting of cloud cover is being made, which, together with the advent of a next generation of LEO and GEO meteorological satellites in the time frame 2015 – 2025 should contribute to improve very significantly the accuracy, timeliness and update frequency of cloud cover forecast products used to optimise the scheduling of operations for space based optical remote sensing.

For the optimisation of the instrument's schedule, the NWC and VSRF information are needed to be provided to the payload mission planning. The interface between the mission planning and the meteo services should be realised such that the cloud cover information can be directly extracted. Based on this information the cloud cover ratio of the scenes to be acquired shall be computed automatically within the PDGS. After this the schedule of the instrument will be elaborated accordingly to be then provided to the FOS, which is in charge of incorporating it into the spacecraft's overall mission plan.

4.6.2.5 User Access The realisation of the user interfaces for the access to the system shall be realised via a standardised central user portal. Via this user portal it shall be possible to place general user requests. These general user requests can be related to catalogue inquiries, ordering of available products from the archive up to the placement of new acquisition request for the Spacecraft.

The user portal should be based on state-of-the-art Web technology. When realising the user portal, it

4 Final Report


is mandatory to take into account not only existing but also evolving standards, which are commonly agreed not only in the remote sensing domain but also within a broad community of Geo Information Service providers and users. Utilising standards guarantees interoperability between different systems and a unified access to these from a user point of view.

At present, the Open Geospatial Consortium (OGC) is the most relevant entity which defines and drives the relevant open standards. Within the OGC, the Sensor Web Enablement (SWE) focuses on sensors and sensor networks. The definition of the SWE standards aims to access and, where applicable, to control all types of sensors, instruments and imaging devices via the Web.

For this purpose, the SWE comprises seven major elements, i.e.:

(1) Observations & Measurements Schema (O&M) – Standard models and XML Schema for encoding observations and measurements from a sensor, both archived and real-time. Current standard: OGC 07-022r1 Observation and Measurements – Part 1 – Observation Schema

(2) Sensor Model Language (SensorML) – Standard models and XML Schema for describing sensors systems and processes; provides information needed for discovery of sensors, location of sensor observations, processing of low-level sensor observations, and listing of taskable properties. Current standard: OGC 07-000 Sensor Model Language

(3) Transducer Markup Language (TransducerML or TML) – The conceptual model and XML Schema for describing transducers and supporting real-time streaming of data to and from sensor systems. Current standard: OGC 06-010r6 Transducer Markup Language

(4) Sensor Observations Service (SOS) - Standard web service interface for requesting, filtering, and retrieving observations and sensor system information. This is the intermediary between a client and an observation repository or near real-time sensor channel. Current standard: OGC 06-009r6 Sensor Observation Service

(5) Sensor Planning Service (SPS) – Standard web service interface for requesting user-driven acquisitions and observations. This is the intermediary between a client and a sensor collection management environment. Current standard: OGC 07-018 Sensor Planning Service Application Profile for EO Sensors

(6) Sensor Alert Service (SAS) – Standard web service interface for publishing and subscribing to alerts from sensors. Draft standard (not yet released): OGC 06-028r5 Sensor Alert Service

(7) Web Notification Services (WNS) – Standard web service interface for asynchronous delivery of messages or alerts from SAS and SPS web services and other elements of service workflows.

With relation to the "Inspire" directive of the EC, which aims at establishing a Geo-data infrastructure for Europe, the OGC specifications are the key drivers for defining the standardised exchange of Geo-information within the European Community.

As can be seen from the above considerations, the realisation of the user portal following the OGC standards allows to define generic user requests via the Web, where user requests can be among others:

• Catalogue browsing (primarily the request/mission catalogue) • Image requests from the archive and re-processing requests of archived data • Performing new acquisitions (routine and emergency) • Monitoring of events and submittal of alerts and notifications

Apart from the above-mentioned user services, the user portal is also in charge of handling all related

4 Final Report


management functions, e.g. provisions of access rights, granting of user privileges, etc. By defining the access rights, a user may have the privilege only to browse the request/mission catalogue, to subscribe to request to currently scheduled acquisitions products, up to requesting new acquisitions. In this context, the User Portal is also responsible of providing prioritized access for users requesting new acquisitions in the frame of emergency applications. This is in particular necessary to guarantee the high reactivity of the overall system for emergency situations.

The UACC (User Access, Coordination and Control) domain is currently responsible of the interface with the end-user and for all the services pertaining to the user interfaces including the ordering function, the mission planning, the master catalogue and the help desk (with the exception of the dissemination). The UACC domain is composed of a number of heterogeneous elements, some of them being natively multi-mission (e.g.: MUIS, MMMC, EOLI), some other elements, which were not-natively multi-mission, have been adapted. The UACC domain is located at ESRIN.

For Geo-Oculus, the evolvement of the current UACC towards multi-mission planning can be considered as a preliminary baseline for defining the realisation of user accesses and all related functionalities.

4.6.2.6 Payload Data Reception, Processing and Dissemination As already stated earlier, the reception of Payload data can make use of:

• A nominal Payload Receiving Station collocated with the main Processing and Archiving Facility

• A set of smaller user dedicated Payload Receiving Stations, with smaller size dishes.

At time of ingestion, the Payload data will be decrypted first. After that, the Payload data have to be screened and the metadata attached to this Payload data are extracted in order to feed the catalogue. A Moving Window Display function may be included to provide the capability to display the raw data before it is processed.

Within the Payload Data Ground Segment, the processing should proceed up to Level 1B or even further, depending on the availability of auxiliary and ancillary data required for processing. These include auxiliary data resulting from calibration (generated inside the PDGS), predicted or restituted orbit data (generated by the FOS) and other auxiliary / ancillary data for instance for more precise geo-localisation or ortho-rectification.

Dissemination should occur mainly towards the service segment, the latter being in charge of further processing and delivery of value added services in compliance with the GMES Service provision model.

The above-mentioned processing chain for the image product has to be fully automatic in order to minimise the processing delays in the context of emergency observation missions. The assignment of different priority levels for each self-contained image product may be suitable to additionally accelerate the processing and transmission of urgent data.

Regarding the required processing performance, it is suitable to analyse a worst-case scenario. For oil-slick detection, an image size of approx. 6E+09 Bits can be assumed. A product consists of 15 image parts, giving in total 9E10 Bits to be processed. As a reference for a first approximation we can assume that the required floating point operations (FLOPS) per bit are comparable to the processing of a Spot 5 image. For a Spot 5 image, 2000 FLOPS per bit are required, which results in a total of

4 Final Report


1.8E14 FLOPS per product.

The currently available hardware capabilities can be estimated using the example of a SUN SPARC Enterprise M 9000 Server, which provides a computing power of 1.032 TFLOPS. The M 9000 is one of the most powerful servers available today. As a result, the M 9000 would be able to compute the above-mentioned product within 180 seconds (1.8E14 TFLOPS / 1E12 TFLOPS).

The automatic dissemination of the products to the dedicated Service Segment for the generation of higher level products / value adding can be facilitated by assigning Request-IDs, by which each user request, the subsequent tasking within the Spacecraft's schedule and the finally delivered product can be identified and automatically routed to the end-user.

4.6.2.7 Encryption Concept From a generic point of view, the encryption of the payload data shall ensure the confidentiality of its content. The reason why this data shall be kept confidential is basically motivated by the following two headlines:

• Public safety • Commercial aspects

The consideration of public safety aims at safeguarding the EO data from misuse by illegal groupings such as criminal associations or even terrorists. This aspect has already been considered by different legalisation authorities worldwide.

The commercial aspects cover all those issues which are related with the property rights of the image data. In this respect, the satellite data should be safeguarded from eavesdropping or theft in order to be able to retail the image products to commercial users.

As a result of the previous considerations, the encryption concept should be simple. A commercial level encryption concept should be adequate.

One of the main issues will be to ensure that, in case of deployment of users’ terminals for emergency applications, only the user terminal which submitted the observation request should be able to receive the image(s) acquired (in addition to the core Payload Data Ground Segment, which would receive all images in parallel for the sake of their long term archiving).

Among the possible concepts for encryption, the following alternatives can be considered:

• One key per image: this is the most secure but also the most complex alternative • One key per time slot: in this concept, the keys would be changed at regular time intervals

e.g. each day, each week, each month • One key per receiving station: each “user” receiving station would then be able to decrypt

only the images which have been encrypted with this station’s key. The core PDGS would receive all images and then would also need to receive all keys.

One specific issue with respect to users’ receiving terminals will be how to distribute the keys, which may be an issue in case of deployment of the stations where no permanent network is available. Note that this issue extends to the distribution to the users’ stations of any other kind of auxiliary data needed for data reception (pointing data if the satellite is not on a geostationary orbit, time slot for data reception) and for processing (auxiliary data, orbit data).

4 Final Report


4.6.2.8 Ground Stations for Geo-Oculus For the preliminary architecture of the Geo-Oculus ground segment it is foreseen to establish two TT&C stations to provide redundancy for the TT&C functionality.

TT&C Ground Stations

The TT&C Stations are responsible for exchanging telecommands and telemetry with the satellite and to provide ranging functionality. For the envisaged orbit determination based on the spread spectrum ranging method, at least two S-Band stations are required; to achieve optimum performance three S-Band Ground stations should be foreseen. It is recommended to use dedicated S-Band GEO ground stations for the TT&C functionality providing permanent contact with the satellite.

Using the S-Band for TM/TC and ranging makes the system compatible to the LEOP G/S network. During LEOP, commissioning and verification phase the LEOP G/S network can provide backup and failsafe capabilities for the operational system within the initial verification phase. Additionally using a lower frequency band like the S-Band in combination with a big G/S antenna size increases also the ranging accuracy of the station.

The location of the primary S-Band TT&C ground station can principally be selected at free choice. The only constraint is that the S-Band G/S needs a direct communication link to the operations facilities (being ESOC in Darmstadt as a baseline), which allows a seamless exchange of the TM/TC data between ESOC and the G/S. At present stage, the Agency's ESTRACK facilities located at Maspalomas (Spain) is considered as primary ground station. The secondary ground station is assumed to be located in Redu, Belgium. The monitoring and control of the S-Band ground stations can be achieved remotely from ESOC by the staff already in place.

TT&C Standards and Interfaces

For compatibility reason to the ESOC/ GSOC G/S network, it is also recommended to follow the CCSDS standard for the TM/TC data packets and to provide SLE (Space Link Extension) interfaces.

PDT Ground Stations

As a baseline for the Payload Data Ground Segment the main data reception facility shall be equipped with a dedicated ground station to provide permanent contact with the satellite. As has already been mentioned before, it is recommended as far as feasible that the receiving station is collocated with the Payload Data Ground Segment in order to reduce the latency between data reception and processing.

What concerns the usage of a dedicated frequency band for payload data transmission, the ITU allows for the data reception of GEO earth observation satellites to use the X-, DBS- or Ka-Band. At present stage, the X-Band has been selected as baseline for the Geo Oculus PDT. The X-band is in the earth observation domain the most commonly used frequency range for TM data transfer.

However, a GEO based S/C the utilisation of the X-band for payload data transmission may produce interferences with the LEO systems. A LEO Ground Station located within this foot print can probably cross this permanent GEO TM link through the tracking process of its LEO spacecraft. As a result of the interference, it may be disturbed in its link and might loose its track. A small foot print of the GEO S/C can reduce this risk, on the other hand, this increases the size of the TM transmit antenna on board the S/C.

5 Final Report


5 Recommendations on further Analysis

5.1 System Analysis

Detailed analysis of the manoeuvre times A driving aspect of the mission performance is the manoeuvre time of the system to point from on observation pattern to the next. Depending on the configuration of the AOCS, the limiting factor is the settling time to achieve the desired level of stability after the active part of the manoeuvre. It has been identified that especially the characteristics of the solar array are of relevance for the settling time.

It is highly recommended to analyse the manoeuvre times in more detail, especially concerning the characteristics of large structures as the solar array, in order to consolidate this important aspect of mission performance at a high level of confidence.

Detailed investigation of microvibration aspects At the required level of attitude stability, the microvibrations from momentum wheels, solar array drive and, if applicable, of an antenna pointing mechanism have to be minimized and/or compensated. A detailed analysis of the microvibrations and the means of reductions is highly recommended for the further study phases.

5.2 Mission Objectives and Data Processing At least two major issues remain at this stage of the GEO study. First the need to strongly consolidate the user’s requirement, second the temporal coverage specificity of the GEO (compute the optimal revisit frequency to get one clear image per days, based on the Eumetsat cloud products archive and taking into account ocean colour geometrical limitations).

Additional proposed tasks: • Justification of the GEO concept for OC. A less demanding requirement on the temporal

coverage is to be able to detect the daily oceanic structures (like Chlorophyll gradient). Contrary to purely numerical techniques of “optimal interpolation” trying to fill the gap of LEO, the GEO concept could directly supply the physical data in a progressive way among the day. It is proposed to analyse the progressive detection of Chlorophyll structure (i.e. progressive improvement of the gradient computation with increasing clear zones along the day), as a function of the number of acquisitions, and to derive the minimal revisit requirement which suits current operational services in structure detection.

• Air mass issue and atmospheric correction. There is a big need to have a radiative transfer modeling tool in spherical coordinates, in order to access the realistic air mass requirements. To our knowledge such a code is not available to the Ocean Colour community in Europe and could be developed.

• Coverage analysis The user requirement on temporal coverage refers to two main aspects: − need to have several images per day in order to built one daily cloud-free synthesis (e.g.

phytoplancton map) − need to have several clear images per day in order to follow rapid events (e.g. tides, NRT

water quality monitoring). The analysis conducted so far on "availability coverage" used a very high revisit time (15 min) and is thus not exactly scaled to the requirements and potential of Geo-Oculus (agility for pointing on cloud-free region). An acquisition scenario that would optimise the cloud-free region, taking into account the realistic duration of acquisition, pointing and stabilization

5 Final Report


based on EUMETSAT images could be elaborated. The analysis should also include other constraints: minimization of air mass fraction and glint avoidance.

• Data processing. Ocean colour measurements from GeoOculus are characterised on one side by high frequent, spatial high resolution, which are superior to current LEO orbit data, on the other side by radiometrically less favourable conditions, including probably lower SNR, possibly spectral and spatial misregistration between bands and a very large viewing angle at higher latitudes. For detailed assessment and identification of required developments of data processing methods, the relevant processing chain should be analysed for sensibility to the characteristic of geostationary observation. Beside the traditional approach it should be studied to develop products which require simplified processing. The processing would either use directly TOA radiance without explicit atmospheric correction, or perform a simplified AC.

6 Final Report


6 Conclusion

As a conclusion the principle feasibility of the proposed Geo-Oculus mission is confirmed. The results show that the challenges in terms of instrument design and LoS performance can be met.

It is however clear that a lot of assumptions concerning the selected applications, the derived product requirements and the mission scenarios have to be re-iterated before the coming study phases. As a consequence the modified requirements will then lead to a re-iteration of the system design and performance.

Nevertheless, the results of the study are very promising to allow a clear recommendation for a continuation of the activities also in view of the potential of Geo-Oculus to become an operational mission e.g. as a part of the GMES program.

A Final Report

Doc. No: GOC-ASG-RP-002 Page A-1Issue: 2 Date: 13.05.2009 Astrium GmbH

Annex A Abbreviations

AD Applicable Document

ADCS Attitude Determination and Control System

AEF Apogee Engine Firing

AGA 8PSK Eight Phase Shift Keying

ATSR Along Track Scanning Radiometer

AATSR Advanced ATSR

AIDCO EuropeAid Co-operation Office, EU

AIT Assembly Integration & Test

AIV Assembly Integration & Verification

ALADIN Atmospheric Laser Doppler Instrument

AMAP Arctic Monitoring and Assessment Programme

AME Absolute Measurement Error

AOCS Attitude and Orbit Control System

APE Absolute Pointing Error

APS Antenna Pointing System

APSK Advanced Phase Shift Keying

ASAR Advanced Synthetic Aperture Radar

ASD Astrium D (Deutschland)

ASF Astrium F (France)

ASIC Application Specific Integrated Circuit

ASU Astrium UK (United Kingdom)

AVHRR Advanced Very High Resolution Radiometer

BAe British Aerospace

BCR Battery Charge Regulator

BDR Battery Discharge Regulator

BOL Begin of Life

BRDF Bi-directional Reflectance

Distribution Function

BTDF Bi-directional Transmission Distribution Function

CCD Charged Coupled Device

CDH Command and Data Handling

CFRP Carbon Fibre Reinforced Plastic

CMG Control Momentum Giros

CMOS Complementary Metal Oxide Semiconductor

CNES Centre Nationale d´Etude Spatiale

COMS Communications Operational Meteorological Satellite

CORINE Coordinated Information on the European Environment

COTS Commercial of the shelf

CPS Chemical Propuslions System

CSA Canadian Space Agency

CSS Coarse Sun Sensor

DC Direct Current

DET Direct Energy Transfer

DG European Union Directorate General

DGA La délégation générale pur l’armement

DLR Deutsches Zentrum für Luft und Raumfahrt

DMC Disaster Monitoring Constellation

DSNU Dark Signal Non-Uniformity

DUE Data User Element

DUP Data User Project

EADS European Aeronautic Defence and Space Company

EC European Community

ECSS European Cooperation for Space Standardization

EEA European Environmental Agency

EIONET European Environment

A Final Report

A-2 Page Doc. No: GOC-ASG-RP-002 Issue: 2 Astrium GmbH Date: 13.05.2009

Information and Observation Network, EEA

ENS Earth Observation, Navigation and Science

ENTR Enterprise

ENV Environment

EO Earth Observation

EOGeo Earth Observation from GEO

EOL End of Life

EOPP Earth Observation Preparatory Programme

EOS Earth Observation System

EPS Electrical Propulsion System or Electrical Power System

ERS Earth Remote Sensing Satellite

ESA European Space Agency

ESOC European Space Operation Centre

ESRIN European Space Research Institute

EU European Union

EUROSTAT Statistical Office of the European Union

EWSK East West Station Keeping

fAPAR Fraction of Absorbed Photosynthetically Active Radiation

fCover Fraction of Land Cover type

FDIR Failure Detection Identification and Recovery

FDS Flight Dynamics System/Software

FM Flight Model

FOG Fibre Optic Gyro

FOS Flight Operations Segment

FOV Filed of View

FPGA Free Programmable Gate Array

FTS Fourier Transform Spectrometer

GAC GMES Advisory Counsel

GCP Ground Control Points

GEMS Global Earth System Modelling Using Space and in-situ data (FP6 IP)

GEO Geostationary Earth Orbit

GFRP Glass Fiber Reinforced Polymer

GIFTS Geostationary Infrared Fourier Transform Spectrometer

GIS Geo Information System

GMES Global Monitoring Environment and Security

GMFS Global Food Security Service

GNC Guidance Navigation and Control

GNSS Global Navigation Satellite System

GOFC-GOLD Global Observation for Forest and Land Cover Dynamics

GOME Global Ozone Monitoring Experiment

GOSIS GMES Organisation and Systems Integration Scenarios

GPS Global Positioning System

GS Ground Sampling

GSD Ground Sampling Distance

GSE Ground Support Equipment

GSE GMES Service Element (ESA Projekts)

GSO Geosynchronous Orbit

GTO GEO Tranfer Orbit

HEMP High Efficient Electromagnetic Plasma

HEMPT High Efficient Electromagnetic Plasma Thruster

HRS High Resolution Instrument

HW Hard Ware

ICEMON Sea ice monitoring in the polar regions (GSE project)

A Final Report


ICU Instrument Control Unit

IEEE Institute of Electrical and Electronics Engineers

IG Implementation Group

IGSO Inclined Geosynchronous Orbit

IMU Inertial Measurement Unit

INR Image Navigation and Registration

IP Integrated Project (European Commission)

IR Infrared

ITAR International Traffic in Arms Regulations

ITD Infoterra Deutschland

ITT Invitation to Tender

ITU International Telecommunications Union

LAI Leaf Area Index

LC Land Cover

LEO Low Earth Orbit

LEOP Lauch and Early Operations

LHCP Left Handed Circulary Polarized

LMCS Land Monitoring Core Service

LNA Low Noise Amplifier

LOS Line of Sight

LR Low resolution

LST Local Satellite Time

LUSI Land Use and Spatial Information (European Topic Centre)

LWIR Long Wave Infrared

MARS Monitoring of Agriculture by Remote Sensing

MCT Mercury Cadmium Telluride

MIR Medium Infrared

MLI Multi Layer Insulation

MMU Minimum mapping unit

MOU Memorandum of

Understanding

MPPT Maximum Power Point Tracker

MR Medium resolution

MS Member State of the European Union

MSG Meteosat Second Generation

MTF Transfer Function

MTG Meteosat Third Generation

MTR Mid Term Review

MWIR Medium Wave Infrared

NDVI Normalised Differential Vegetation Index

NEDL Noise Equivalent Delta

NIR Near Infrared

NRC National Reference Centre

NRT Near Real Time

NSSK North South Station Keeping

OBC Onboard Computer

OC Ocean Colour

OCM Orbit Control Manoeuvre

OD Orbit Determination

OQPSK Offset Quaternary Phase Shift Keying

OSPAR The 1992 OSPAR Convention; the current instrument guiding international cooperation on the protection of the marine environment of the North-East Atlantic

PB Programme Board

PCM Pulse Code Mode

PDGS Payload Data Ground Segment

PDH Payload Data Handling

PDHT Payload Data Handling and Transmission

PDT Payload Data Transmission

PF Platform

PFD Power Flux Density

A Final Report

A-4 Page Doc. No: GOC-ASG-RP-002 Issue: 2 Astrium GmbH Date: 13.05.2009

PM Progress Meeting

PPS Pulse per Second

PRM Product Requirements Meeting

PRNU Pixel Response Non Uniformity

PSK Phase Shift Keying

QE Quantum Efficiency

RAMSAR Convention on Wetlands, signed in Ramsar, Iran, in 1971

RAAN Right Ascension of Ascending Node

RD Reference Document

REGIO Directorate General Regional Development (Europ. Com.)

RF Radio Frequency

RFI Request for Information

RFP Request for Proposal

RHCP Right Handed Circulary Polarized

RME Relative Measurement Error

ROM Rough Order of Magnitude

RPE Relative Pointing Error

RS Reed Solomon

RW Reaction Wheel

SA Solar Array

SADM Solar Array Drive Mechanism

SAGE Service for the Provision of Advanced Geo-Information on Environmental Pressure and State (ESA GSE project)

SAR Synthetic Aperture Radar

SCU Spacecraft Computer Unit

SK Station Keeping

SNR Signal to Noise Ration

SOLAS Surface Ocean Lower Atmosphere Study

SOW Statement of Work

SSPA Solid State Power Amplifier

SSH Sea Surface Height

SSP Sub Satellite Point

SST Sea Surface Temperature

STR Star Tracker

SWIR Short Wave Infrared

TBC to be Confirmed

TBD to be determined

TC Telecomand

TESI TerraSAR Exploitation and Service Infrastructure

TIR Thermal Infrared

TM Telemetry

TMA Three Mirror Anastigmat

TMTC Telemetry and Telecomand

TN Technical Note

TOA Top of Atmosphere

TOC Table of Contents

TRL Technology Readiness Level

TV Thermal Vacuum

TWTAs Travelling Wave Tube Amplifiers

UK United Kingdom

UN United Nations

UNCLOS United Nations Convention on the Law of the Sea

UNFCCC United Nations Framework Convention on Climate Change

UV Ultra Violet

VGT Vegetation

VHR Very high resolution

VIS Visible

VLWIR Very long wavelength Infrared

VNIR Visible Near Infrared

WBS Work Breakdown Structure

WFD Water Framework Directive

WFE Wave Front Error

WP Working Package

A Final Report
