dwd - zentrale entwicklung michael rohn, dirk heizenreder ... · 11th workshop on meteorological...

11th Workshop on Meteorological Operational Systems - ECMWF 12-16 November 2007 - M.Rohn, D.Heizenreder 1

Visualisation and Production using NinJo

DWD - Zentrale Entwicklung

Michael Rohn, Dirk Heizenreder

Eleventh Workshop on Meteorological Operational Systems

ECMWF, 12-16 November 2007


� NinJo: Overview

� Visualisation

� Product generation

� Automatisation with NinJoScience

� Status and Outlook


International collaborationof Met Services inCanada, Denmark, Germany, Switzerland

Further joint Developmentand Maintenance

NinJo: Developed by …


Modular structureeg.GRIB data

�GridLayer

Many data typesGRID, BUFR, Radar, Satellite, warnings, cells, points, areas, …

ServerClient ExternalSystems

NinJo has …


Building kit for meteorological applicationsFrom a satellite image viewer to a complex meteorological workstationVisualization and Production tools can be configuredeach application represented by own layer

Flexible programming environmenteach partner constructs his NinJo and adds own data typescompletely Java based

User orienteduser can define his own NinJo desktopeverything is configurable by user action

Generic Meteorological Workstationeasy extension possible (open for new data, new storage systems)

NinJo is …


� NinJo: Overview

� Visualisation





NinJo: Visualisation - ConceptMain Window

Main scenesecondary scenes

up to 3,configurable

Secondary WindowsMeteograms,

CrossSections, based on extendable

diagram framework


Highly configurableNinJo is the GUI for configuration (System-, Site-, User-Level)xml behind the scene

Favourites and Functionality can be configuredMaps/ Geographical data

• Data types• Color Tables• Legend positions

GUIMenuesSelected data typesSelected layers, active layerNumber of (secondary) scenes

Multi-lingual

NinJo: Visualisation - Flexibility


NinJo: Visualisation - Layer specificMainToolbar

LayerSelection Layer

Menue

LayerToolbar

TimeNavigation


Independent visualisationmeteorological datageographical information

Flexible composition of sceneSuperimposed visualisationModes active, invisible, duplicated

Interactive layersspecific applications

duplicated layer

active layer invisible layer

add layer

NinJo: available Layer in 1.2


Geo RasterHeightElevationLanduse

Geo VectorCoastline, boundariesRoads, railway,Rivers, cities, airports

Geo Gridparallels and meridians

Landsat image (50m resolution) + Geovector data

NinJo: Geo Layers


Supports major platformsGOES, GMS, MeteosatNOAA, FENGYUN

ProductsChannel combinationsCompositesSAF products

Basic Image processingcontrast enhancementcolour tables

NinJo: Satellite Layer


observation and forecast dataVisualization of station based data

configurable displaysymbols and plot models

Sorting algorithmsMin/MaxRatingHitlist

Meteogramsat selected locations

NinJo: Surface Layer


Editable plot-model

NinJo: Surface Layer- Configuration

Priorization


Point map

horizontal distribution of selected elements

Sounding diagrams

vertical distribution at specific locations

(next slide)

NinJo: Sounding Layer


NinJo: Sounding Layer - diagramsCloud cover and turbulences

Hodograph


various numerical NWP dataGME, LME, ECMWF, GFS, LFP, HIRLAM...

calculated elementsprecip. sums, differences

display modes:Isolines, Isoareasvalues, symbols, wind barbs

NinJo: Grid Layer


various Radar products

Projectionscartesian, stereographic

NinJo: Radar Layer


Different networks

Visualization of characteristicsCloud to ground, cloud to cloudPolarityno. of strokesamplitude

Color-encodingdepending on age or polarity

NinJo: Lightning Layer


� NinJo: Overview

� Visualisation





Report types: GAFOR, GAMET, SIGMET, TAFAeronautical

warningsaerea forecastspoint forecasts

defined flight paths

NinJo: Aviation Layer


Weather monitoringcontinous alerting to severe weather situations

Monitoring of qualitycomparison of warningsto observed weather

Monitoring ruleshighly configurableall data typesthreshold definitionscombination of criteria

NinJo: AutoMON


Serves to issue warnings: Editing Production Monitoring

Selectionobjects/areas

Editingwarning templateheight selectiontime selection

Productiondissemination to textgeneration/user

Monitoringcomparison to observations assisted by AutoMON

NinJo: EPM Layer


Outlook: Automatic Warning – Project until 2009

WithinAutoWARN Project

Outside ofAutoWARN

Observations

KONRAD / SCIT

Radar

Lightning, BlitzMOS

LMK (+ other models)

WarnMOS

PEPS

MOSMIX

ASGAutoWARN Status

Generator� Automatic

Warning StatusProposal

ASEAutoWARN Status

Editor� Manually

Modified WarningStatus, Export

ExternalProduct

Generation(WV11)

and distribution

AutoMONAutomaticMonitoring�Warning

Events

Further development: based on AutoMON / EPMcontact: [email protected]


Point Forecast Editing

Selectionarea / pointforecast timeheight level

Editingabsolute / relativeinterpolation

Consistencycompares elementsoptional corrections

NinJo: MMO Layer


Interactive graphical editor

Editing Graphical ObjectsFronts, Jet, Clouds, SigWx elementsWeather areas,…Texts and icons

MetObjectsencodingdistribution for other users

NinJo: IGE


Field editingphysically based modificationsresult storage for further processing

Edit Modespoint, single and multiplestructuredrag points on isolines

black listfilter observations

Prototypicaladapt balanced fieldse.g. wind - mass

NinJo: On Screen Analysis


� NinJo: Overview

� Visualisation





SatWetter

BlitzMOS

RadVorOP

SYNOP

METAR

SPECI

GMEObservation

ECMWF

COSMO-EU

COSMO-DE

GME-DMO

GME-MOS

ECMWF-MOS

MIX-MOS

COSMO-EU-DMO

COSMO-EU-KALMAN

COSMO-DE

Motivation: NinJo-Science

NWP Postprocessing

large variety of data:only partial used in forecastclose to impossible to manually assess

Automatisationsupport forecastermore time for extreme weather forecasting/warning

Local forecast(site specific

guidance)Nowcasting


10 11 12 13 14 15 48 120 168

Observation

Nowcasting

Model

(post processing)

Adaptation/Objective Optimisation

Example: Update at 11:00 UTC

[h]

COSMO-DE

EPS-Mean

<planned>

Data merging: Model postprocessingNowcastingobservation

Individual weightstime dependantelement specific


Adaptation: Interpretation of nowcasting products

SatWetter RadVor_OpBlitzMOS

Heterogeneous dataformats, resolutioninformation content, probabilistic

Interpretation required

SatWetter: cluster analysis imagery + synopBlitzMOS: probabilities for lightning, MOS basedRadVor-OP: precipitation from Radar + translation


Adaptation: Combination of products

ww p

3 0.2570 0.1071 0.0985 0.1087 0.1089 0.1076 0.0477 0.0478 0.0479 0.0416 0.10

ww p(norm)17 0.04580 0.01781 0.01782 0.01783 0.01784 0.01785 0.01786 0.01787 0.01788 0.01789 0.01790 0.01795 0.04596 0.04597 0.03698 0.04599 0.036

Sum 0.435wwi(wk4) (1-0.435)/50

„Linear Opinion Pool“:Aggregation of probability distribution(eg. „Significant weather“)

individual weightstime dependantelement specific


NinJo ServerTier

External SystemsTier

Adaption & Guidance Service

Access Layer

Science F

ramew

orkAdaption

Guidance

NinJo ClientTier

GRIB DataServer

GRIB DataImport

Image DataServer

Image DataImport

SecurityService

Point DataServer

Point DataImport

Alerting DataServer

Alerting DataImport

Monitoring Service

Watchdog Service

Monitoring

GRIB DataMonitor

Point DataMonitor

Alerting DataMonitor

External Distribution(AFD)

Decoding(GLOBUS)

LDAPDirectory

NinJo Client

AlertingLayer

Point DataLayer MMOMonitoring

Layer

Access LayerWatchdogClient

central production integrated in NinJoPointDataServer to hold result data

VisualisationSurfaceLayer (operational forecast)OOG-Layer („Expert view“)

NinJo-Science: Technical View


Integration in meteorological workstation NinJo


Objective Optimisation / Adaptation

Site specific forecast data merging

currently being evaluated at DWD

AutoSWIS

site specific road weather forecast + energy balance model

… more to come

Cell detection from Radar data: DWD, current project until 2009

Streamlines, Cross sections,Trajectories: MeteoSwiss

NinJo-Science modules


� NinJo: Overview

� Visualisation





Visualisationcomplete package of layers supporting the entire forecasting process at DWD (determinstic, warning, aviation, etc.)

ProductionInteractive layers to enable manual quality control, editing, and productgeneration

NinJo Scienceto support automatisation at DWD

NinJo StatusDWD: replaced legacy system MAP in October 2007BGIS (military): replaces legacy system in December 2007MSC: operational introduction early 2008

Summary


NinJo Viewcurrent Version 1.2DWD: since 2006 operationally used parallel with legacy system

October 2007: legacy system has been switched offDMI and MeteoSwiss: in 2007 operationally usage parallel with legacy systemMSC: operational introduction in 2008

NinJo Product (1.3 Q3 2008)New: Interactive graphic editing / Production work bench

NinJo Enterprise (1.4 Q3 2009)Enabling third parties to develop their own layers and functionalities

Current licence holdersINM – SpainSAWS - South Africa

NinJo: Outlook


www.ninjo-workstation.com

[email protected]

More

… information… examples

… news

NinJo: Licensing

dwd - zentrale entwicklung michael rohn, dirk heizenreder ... · 11th workshop on meteorological...

Documents