s2. euv irradiance & calibration. february 13-17, 2006 aia/hmi science team meeting eparvier - 2...
Post on 21-Dec-2015
220 Views
Preview:
TRANSCRIPT
S2. EUV Irradiance & Calibration
AIA/HMI Science Team Meeting Eparvier - 2February 13-17, 2006
EUV Observations
Most of the new missions that make the next 5 years of solar observations look so exciting carry EUV/SXR instruments Solar-B EIS, XRT STEREO SECCHI EUVI GOES SXI, XRS
Two of the three SDO instruments are strongly focused on the EUV Calibration of these EUV instruments is essential for a number of
reasons: EVE calibration is important for understanding the effects of
irradiance variability on the atmosphere AIA calibration is important for understanding the thermal structure of
the corona Even scientific investigations that don’t explicitly rely on calibrated
EUV observations will benefit from cross-calibration of EUV instruments
AIA/HMI Science Team Meeting Eparvier - 3February 13-17, 2006
Agenda
(slightly modified since the announcement was posted)
1. Overview of EVE calibration
2. Overview of AIA-EVE cross-calibration
3. Discussion• cross-calibration with other instruments• problems• priorities• procedures
Wanted: practical ideas and questions, not necessarily solutions (yet…)
EVE and those other instruments on SDO
Frank Eparvier
LASP / University of Colorado
eparvier@lasp.colorado.edu
AIA/HMI Science Team Meeting Eparvier - 5February 13-17, 2006
Reminder: EVE Instrument Overview
ESPMEGS B/P
MEGS A
SAM
EOP
EEB
EVE Resources
Power (orbit average) 43.9 Watts
Mass 54.2 kg
Data Rate 2 kbps (engineering)7 Mbps (science)
Dimensions (L x W x H) 99 cm x 61 cm x 36 cm
Key Components EVE Optical Package (EOP)
MEGS MEGS A + SAM MEGS B + P
ESP EVE Electrical Box (EEB)
Processor & Memory Interfaces (1553 & HSB) Power / Heaters / Control CCD power converters ESP power converters
Spacecraft Coordinates
+X
+Y+Z
SDO Spacecraft
AIAHMI
EVE
AIA/HMI Science Team Meeting Eparvier - 6February 13-17, 2006
How does EVE measure the EUV? Multiple EUV Grating
Spectrograph (MEGS) At 0.1 nm resolution
MEGS-A: 5-37 nm MEGS-B: 35-105 nm
At 1 nm resolution MEGS-SAM: 0-7 nm
At 10 nm resolution MEGS-Photometers: @ 122
nm Ly- Proxy for other H I
emissions at 80-102 nm and He I emissions at 45-58 nm
EUV Spectrophotometer (ESP) At 4 nm resolution
17.5, 25.6, 30.4, 36 nm At 7 nm resolution
0-7 nm (zeroth order) In-flight calibrations from ESP and
MEGS-P on daily basis and also annual calibration rocket flights
0.1
1
4
7
10
nm
AIA/HMI Science Team Meeting Eparvier - 7February 13-17, 2006
EVE Science Requirements
Parameter Minimum Success Requirements
Comprehensive Success Criteria
Design Goals
Range 6 or more emissions to specify the chromosphere, TR, and corona, plus the He I I 30.4 nm emission
0.1-105 nm 0.1-105 nm
Resolution 0.2 nm for these lines 0.1 nm for 18 or more emissions to specify the chromosphere, TR, and corona, and 5 nm or better elsewhere
0.1 nm
Time Cadence 60 sec < 20 sec 10 sec Accuracy 35% for 5 nm
intervals and daily average
25% for 5 nm intervals and daily average
20% for brighter emissions
Mission Life 3 years 5 years 5 years, long enough to sample low and high solar activity
AIA/HMI Science Team Meeting Eparvier - 8February 13-17, 2006
EVE Data Products
Level Algorithm purpose
Scientifically Useful
Description File Duration
Daily Volume
(MB)
0A Fast validity check
No TLM consistency/quality checking ~1 minute 76000
0B Assemble images, detailed data verification
No Data checks for CRC and pixel parity, parse data packets, merge image data, separate by science channel and filter wheel position
~1 minute 76000
0C Space Weather Yes Quick-look indices, MEGS-A, MEGS-B, SAM, MEGS-P & ESP
1 minute 36
1 Apply calibration
Yes (SAM, ESP, MEGS-P)
Use measurement equations to produce irradiance units
1 hour 1095
2 Re-grid, extract lines
Yes Bin data to fixed wavelength scale, integrate over emission features with background removal
1 hour 1160
3 Daily average Yes Merge all component data into daily averages, bin to 0.1 and 1 nm
24 hours 0.026
AIA/HMI Science Team Meeting Eparvier - 9February 13-17, 2006
Calibration is a Lifetime Commitment
The Calibration Essentials: Understand the Measurement Equation:
Know all the parameters that go into the measurement to irradiance conversion and assess how to best quantify each
Do a thorough error analysis and uncertainty budget Calibrate pre-flight:
Use a standard radiometric EUV source Primary standards, such as NIST SURF-III source, are preferred (note: SURF
beam flux known to <1% for EUV ranges) Track in-flight:
Any instrument changes that will affect results E.g. detector flat fields, gain changes, temperature effects, background signals, …
Re-Calibrate in-flight: As close after launch as possible (changes since pre-flight calib.) On a regular basis thereafter in order to track absolute changes E.g. redundant channels, on-board sources, rocket underflights, proxy models
Validate: With measurements made with other instrumentation Comparisons with models
AIA/HMI Science Team Meeting Eparvier - 10February 13-17, 2006
MEGS A & B Measurement Equations
Where:
E Solar spectral irradiance(x,y) Detector pixel location
S Raw signal from detectort Integration timeG Detector gain
fFFFlatfield correction
fLinLinearity correction
CBkgBackground signal
CSLScattered light signal
fImagePixel contribution weighting to slit image
Good(x,y) “Good” pixels in slit image
ASlitSlit area
Dispersion (bandpass of single detector element)
RcResponsivity at center of FOV
fFOVPointing within FOV correction
fDegradDegradation correction
f1AUNormalization to 1-AU
Wavelength
EOSHigher order correction
C1 x,y St GfFF fLin
C2 x,y C1 CBkg CSL
E1 C2
ASlit RC fFOV fDegrad f1AU
hc
EMeasured E1 EOS
C3 f Image x,y C2 x,y
Good(x,y )
f Image x,y
Good(x,y )
AIA/HMI Science Team Meeting Eparvier - 11February 13-17, 2006
MEGS-A & B Error Analysis
The uncertainties of the various correction factors must be propagated through to determine the accuracy of the measured irradiance (note: denotes uncertainty in the units of the variable):
For bright solar emission features the primary contributors to accuracy are the uncertainties in RC (the responsivity of the instrument) and the fDegrad (degradation correction)
For dim solar emissions, other uncertainties dominate, such as the precision of the measurement and the various corrections to the signal
EMeasured E1
2 EOS
2 1
2
C1C1
S2
S 2 G
2
G 2 fFF
2
fFF2 fLin
2
fLin2
12
C2 C1
2 CBkg
2 CSL
2 1
2
E1E1
C32
C32 ASlit
2
ASlit2
2
2 RC
2
RC2 fFOV
2
fFOV2
fDegrad2
fDegrad2 f1AU
2
f1AU2
2
2
12
C3
2 fImage x,y C3
fImage x,y Good(x, y )
2
fImage x, y 2
fImage x,y 2 1fImage x,y
fImage x,y Good(x, y )
2
C2 x, y 2
C2 x,y 2
Good(x, y )
AIA/HMI Science Team Meeting Eparvier - 12February 13-17, 2006
EVE Uncertainty Budget and Verification Matrix
SymbolParameter
DescriptionError
BudgetComponent
Level
Instrument
Level
Spacecraft
Level
On-Orbit
Level
S Signal 34% X X X X
t Integration Time 0.02% X
G Gain 1% X X
fFF Flatfield 2% X X X X
fLin Detector Linearity 0.2% X X
CBkg Background 20% X X X X
CSL Scattered Light 20% X X X
fImage Slit Image Weight 2% X X
ASlit Slit Area 8% X
Dispersion 6% X X
RC Responsivity at Center
12% X X
fFOV FOV Correction 10% X X
fDegrad Degradation Correction
18% X
f1AU 1-AU Correction 0.02% X
Wavelength 0.2% X X
EOS Order Sorting 2% X X
EMeasured Irradiance Product
25%
AIA/HMI Science Team Meeting Eparvier - 13February 13-17, 2006
EVE In-Flight Calibration Activities
Continuous Internal Cross-Calibrations: Overlapping Channels within EVE
Daily: Filter wheel movements (dark, alternate filters) Flat field lamps for MEGS CCDs (LEDs)
Quarterly Maneuvers: Cruciform Scans: ±150 arcmin in 3 arcmin steps
Gives gross FOV changes and locates edges of FOV for relative boresight calibrations to SAM and AIA guide telescope
FOV Maps: ±10 arcmin in 5 arcmin steps (5x5 map) Gives finer FOV changes over nominal FOV pointing area (with margin)
Also get bonus mapping when AIA and HMI require maneuvers (though their mappings are different and not optimized for EVE needs).
Annual Rocket Underflights: Fly prototype instruments on sounding rocket periodically. Calibrate rocket instruments at NIST before and after flight to transfer best
calibration to EVE.
Cross-calibration:AIA-EVE,
SDO-everybody else
AIA/HMI Science Team Meeting Eparvier - 15February 13-17, 2006
Overview: AIA/EVE cross-calibration
Spectral response η(λ) (effective area) of AIA channels determined by component-level calibration measurements Mirrors (primary determination of bandpass) Filters CCDs System-level effects
Estimated BOL relative calibration accuracy for AIA is 15% Absolute calibration is more difficult Calibration will change due to contamination, degradation, etc. Therefore, cross-calibration with EVE is highly desireable
First-order cross-calibration procedure: Use EVE MEGS-A measurements of full-disk solar spectral irradiance to
predict a full-disk count rate in each AIA channel Compare EVE-predicted count rate with AIA’s measured full-disk count rate,
and produce a scaling factor for each channel
AIA/HMI Science Team Meeting Eparvier - 16February 13-17, 2006
Refining the Cross-calibration
First-order calibration should be easy to implement, but a few questions remain: What cadence? (yearly? monthly? daily? 10 seconds?) How do we interpret the resulting scale factors?
Contamination? Something else? or is it just an empirical correction, and we don’t worry about it?
There are some potential pitfalls to the first-order AIA-EVE cross-calibration: Field of view Spectral resolution Bandpass uncertainty
AIA/HMI Science Team Meeting Eparvier - 17February 13-17, 2006
Field of View
AIA field of view is 41 arc-minutes (to edge of CCD) / 46 arc-minutes (vignetting circle) 1.3-2.0 pressure scale heights
(at T = 3.0 MK)
Based on Yohkoh observations, we estimate that AIA will observe ~ 96 % of the total coronal radiance Higher fraction for lower-
temperature lines Depends on size and location
of particular structures
Estimated X-ray radiance at 3 MK as observed by Yohkoh/SXT as function of limb height.
Yohkoh/SXT 8 May 1992
AIA/HMI Science Team Meeting Eparvier - 18February 13-17, 2006
Spectral Resolution
Spectral resolution of ~ 1 Å results in calibration errors Less than 1% for longer-wavelength (broad) channels Up to 25% for 171 and 94 Å
Can be corrected by modeling higher-resolution spectrum
Simulated full-disk spectrum (10% AR, 90% QS) shown in blue. Blurred with 1 Å FWHM gaussian and binned at 6 pixels/Å in black. Response of AIA 194 channel shown in red. Folding the black spectrum through the red instrument response results in errors of 1-25% compared to using the blue spectrum.
AIA/HMI Science Team Meeting Eparvier - 19February 13-17, 2006
Bandpass Uncertainty
First-order cross-calibration only allows us to correct the overall scale of the AIA response functions
Uncertainties in the bandpass shape are more important; can we use EVE to correct those?
Measurements of the MSSTA multilayers. This is not data from an AIA telescope, but the illustration of bandpass variations over the mirror surface is relevant.
See the poster by R. Soufli et al.
AIA/HMI Science Team Meeting Eparvier - 20February 13-17, 2006
Questions (1 of 3)
For AIA-EVE cross-calibration: How often should we perform "first-order" calibration? What data products are necessary for this cross-calibration? What sort of operational coordination is necessary? Coordination with rocket
underflights? How do we interpret the resulting scaling factors?
contamination? something else? not at all?
How do we deal with the field-of-view discrepancy? How do we use EVE to correct the bandpass shape of the AIA?
To what extent will cross-calibrations rely on spectral modeling? What improvements in spectral modeling can be made to enhance
calibration accuracy? Can AIA-EVE cross-calibration be used to constrain Fe abundance? What role can DEM extraction from AIA play in cross-calibration, and
extending the spectral range of EVE?
AIA/HMI Science Team Meeting Eparvier - 21February 13-17, 2006
Questions (2 of 3)
For EIS-AIA-EVE inter-calibration: How does EIS-AIA cross-calibration feed back into AIA-EVE cross-
calibration? Is it possible to get full-disk spectra with EIS? If not, how do we cross-calibrate with EVE? If so, how can we coordinate this cross-calibration? Will it be possible to cross-calirbate EIS with the LASP rocket this
year?
For XRT-EVE cross-calibration: Can the EVE SAM and ESP be used to cross-calibrate with XRT? Would this be useful? What sort of coordination is necessary? How often should this be
done? etc.
Can XRT and AIA be cross-calibrated? How? (Using DEM extraction?)
AIA/HMI Science Team Meeting Eparvier - 22February 13-17, 2006
Questions (3 of 3)
Are TRACE and EIT going to be observing during SDO? If so, how do we cross-calibrate with AIA? If not, how do we establish continuity between the AIA dataset and
the EIT/TRACE datasets? How important is this cross-calibration?
For AIA, how important is it to have accurate: Absolute calibration? Relative calibration (channel-to-channel)? Bandpass shape calibration?
AIA/HMI Science Team Meeting Eparvier - 23February 13-17, 2006
Backup Slides
AIA/HMI Science Team Meeting Eparvier - 24February 13-17, 2006
EVE and AIA Inter-Calibrations
EVE spectra can be convolved with AIA bandpasses and compared with integrated images to transfer an absolute irradiance calibration from EVE to AIA. What’s needed for this transfer?
AIA bandpasses (!) AIA image conversion to irradiance EVE irradiances
Can EVE be used to track changing AIA bandpasses? Probably, but how the bleep do we do that? Logistical Questions:
Do AIA and EVE integrations need to be coincident? Do special data products need to be made for inter-calibrations? How frequently should comparisons be done? Are there special calibration activities on-orbit that should be planned?
In conjunction with rocket underflights?
AIA/HMI Science Team Meeting Eparvier - 25February 13-17, 2006
Action Items from EVE Science Workshop (Nov, 2005)Action Item
NumberResponsible
PersonAction Item Description Status Due Date
1Woodraska,
EparvierBetter define the MEGS Quick-Look data for SURF and I&T (e.g. consider display procedures for MEGS Level 0B).
open Dec-05
2 TateCan EVE observations be timed to be consistent with TAI time on intervals of 0.0 (like HMI and AIA)?
open Dec-05
3Eparvier, Kohnert
Add spacecraft roll calibration to EVE calibration plan (CMAD). Should use HMI/AIA roll calibration plan that has discrete roll angles (+/- 26°).
open Jan-06
4 Eparvier, JonesInclude wavelength shift effects more completely in irradiance algorithms and error analysis
open Apr-06
5Pesnell and SDO
data teamsSDO project needs to define standard (common) meta-data specifications open Apr-06
6Woodraska,
Woods
Define the SAM higher level data products (beyond the photon events defined for SAM Level 1A). For example, it might include SAM Level 2 that contains solar irradiance in constant energy bins. Also consider SAM images as virtual product that user specifi
open Apr-06
7Woodraska,
WarrenDefine the critical emissions to extract for MEGS Level 2 data products (and also used in EVE Level 3). At least include the 7 EUV emissions measured by AIA.
open Apr-06
8Woodraska,
Viereck, Tobiska
Develop concept for irradiance server (common interface / platform / format). For example, resolve FTP process (push vs pull) for distributing EVE space weather data products.
open Apr-06
9Pesnell, Eparvier, Bush, Schrijver,
Viereck
Define solar event log for SDO mission (e.g., flare intensity and peak time, flare location on solar disk). Consider creating log as near real-time file using SDO's space weather data products.
open Apr-06
10 Eparvier
Plan to have Solar EUV Irradiance Validation Workshop for spring 2007 to discuss EUV measurements from rocket, TIMED-SEE, SOHO-SEM, GOES-EUVS, STEREO. Also include solar EUV irradiance models in this comparison / validation effort (either part of same wo
open Nov-06
11Jones, McMullin,
WoodraskaDevelop more completely the ESP algorithms that depend on solar EUV irradiance spectra (from MEGS)
open Nov-06
12Woodraska,
CrotserDevelop algorithms to compress MEGS images into spectra. Will need example (predicted) images at different FOV angles.
open Nov-06
13 Bryant, EparvierOptimize filter in-flight operations to reduce data gaps and using only single channel in calibration mode at a time
open Nov-06
14McCaffrey, Eparvier
Select some focussed topics in space weather research for EPO activities open Oct-07
15 WarrenDevelop algorithm to determine differential emission measure (DEM) from EVE spectra. Consider routine production of the disk-averaged DEM and for comparison to AIA image DEM results.
open Oct-07
AIA/HMI Science Team Meeting Eparvier - 26February 13-17, 2006
Comments from EVE Science Workshop (Nov, 2005)Comment Number
Responsible Person
Comment Description
1Schrijver,
Warren, PesnellIdentify the SolarSoft procedures for removing particle "spikes" in CCD images that might be applicable for EVE data processing
2 KohnertConsider taking prototype (rocket) MEGS to SURF first for calibrations to identify any possible instrument or calibration issues
3 SchrijverConsider full-disk intensity data with corrections for flat-field and despiked as part of the science-quality data products for AIA
4 WarrenSpectral modeling of the EVE data at 0.1 nm resolution to obtain higher spectral resolution is important for AIA-EVE calibration / validation for some of the AIA channels
5 WoodraskaResolve if it is important to include ensemble uncertainties for the ESP Level 1 data that includes an average of 4 measurements taken at 0.25 sec cadence
6 WoodraskaEVE data products do not have to have constant array size. That is, gaps in the measurement sequence are acceptable, and flags for gaps are not expected.
7 WoodraksaSAM Level 1A data product should include the photon events in energy units and location in units of CCD pixels.
8 WoodraskaFor EVE Level 3 data, the daily averaged irradiance will be the mean over the day (instead of median).
9Fuller-Rowell,
Smithtro, Sojka, Solomon
Define optimal wavelength bins for near realtime atmospheric modeling. Also include the FUV range.
10 WoodraskaUse pseudo-calibration correction factors for generating the EVE space weather data products and to subtract off dark measurements with constant scalar (unique scalar for each diode). For example, E = (S - K*D) / R
11 Woodraska Have MEGS data in 1-nm intervals for space weather data products.
12 WoodraskaSpace Weather data products to have 1-min cadence (average) for ESP / MEGS-P data and for MEGS A & B data.
13Eparvier, Schrijver
Define plots / movies for EVE space weather data products and overlap with AIA
14 Woodraska Time stamps in Space Weather data product includes date too.15 Woodraska Do not include SAM data in Space Weather data product.16 Tobiska EVE's photoionization rate results are part of SOLAR2000 model already17 Bailey EVE's photoelectron flux results are still needed for LWS I-T studies / mission.
18Schrijver, Warren
Develop algorithms to link magnetic field and active region evolution to irradiance variations
top related