GOESR AWG Calibration Working Group CWG Chair : Changyong Cao

1
GOES-R AWG Calibration Working Group (CWG)
Chair Changyong Cao
Algorithm Working Group Annual Meeting July
20-24, 2009
2
Outline

• Overview
• GOES-R Cal/Val Plan
• Prelaunch calibration
• CDRL review
• Technical interchange meetings
• Calibration algorithms and ground processing
• Calibration case studies
• Reflectance vs. radiance
• NWP/GDAS bias comparison and lessons learned from
  current GOES
• Postlaunch calibration and longterm monitoring
  capability development
• Selection of vicarious calibration sites
• Lunar/star calibration
• Underflight planning
• Space Weather and GLM cal/val support
• Other activities
• Collaboration with other teams/groups
• AWG, L1b/GRB IPT, IMWG, MODIS/MCST
• Collaboration with NIST and MIT/LL

3
Objectives

• Verify and ensure well-calibrated,
  well-navigated GOES-R L1b data for the life time
  of the instruments (ABI, GLM, and SWx)
• Ensure Level 1B data quality. Provide technical
  oversight and IVV for
• Radiometric calibration
• Spectral calibration
• Spatial calibration/navigation check
• Independent verification of L1B data
• Evaluate and mitigate instrument performance
  risks (e.g., possible striping, noise,
  cross-talk, RVS, spectral response uncertainty,
  etc)
• Provide technical support to the GOES-R/PSE
  (Program Systems Engineering)

4
Cal/Val Working Group
CWG Chair Changyong Cao

• NIST
• Eric Shirley
• Allan Smith
• Joe Rice
• Carol Johnson
• Simon Kaplan
• David Allen
• Collaboration with other groups
• NASA
• Jeff Kronenwetter
• Dennis Chester
• Marv Maxwell
• Peter Silverglate
• Chris Rollins
• MIT/LL
• Monica Coakley and co.
• SWPC

• GOES-R PSE
• Management
• Barb Pfar and Bill Mazur
• CWG coordination
• Jared Clark (Laurie Rokke)
• Members funded through PSE
• Fred Wu (ABI)
• Robert Iacovazzi (ABI and GLM)
• Gordana Rancic (Lunar calibration)
• William Rowland (Space Weather)
• Ping Jing (Vicarious calibration sites)
• Jack Xiong (Optics)
• Michael Weinreb
• Tim Schmit

5
GOES-R Cal/Val Plan Volume 1
EXECUTIVE SUMMARY 1. INTRODUCTION 2. INSTRUMENT
PRE-LAUNCH CHARACTERIZATION AND CALIBRATION 3.
GROUND PROCESSING SYSTEM AND OPERATIONAL
CALIBRATION 4. POST-LAUNCH ACTIVATION AND
CHECKOUT 5. ON-ORBIT VERIFICATION (OV) 5.1.
Radiometric performance evaluation 5.2. Spectral
calibration 5.3. Spatial calibration 5.4.
Calibration using celestial targets 5.5. Radiance
inter-comparison of GOES-R and other GEOSS
space-borne sensors 5.6. Vicarious calibration at
reference sites 5.7. Comparison between observed
and CRTM-simulated GOES-R radiances 5.8. GOES-R
OV using airborne radiometer under-flights 6.
LONGTERM INSTRUMENT PERFORMANCE MONITORING 7.
ANOMALY RESOLUTION
GOES-R L1B
To be based-lined next year !
Recent updates included SEISS, EXIS, SUVI, GLM
5
6
Critical Path to GOES-R L1b Data
MRD/PORD
Requires major CWG support
Instrument design/build/test
1
Requires some CWG support
Acceptance storage
2
Cal/Val strategy plan
3
Prelaunch testing, review docs T/V data,
analysis, waivers, spec. compliance
4
Spacecraft integration testing
5
L1b ATBD
6
VV of radiances
IVV
L1b content format
7

NIST traceability/ independent verification
L0, L1b processing system (Algorithm/software/hard
ware) Calibration input parameter dataset (CPIDS)
8
IVV
Navigation check
GOES/POES intercal
Test datasets acceptance test
9

Star/Lunar cal.
OV/PLT
10
IVV
ISO/GEOSS compliant format for interoperability
Science tests
11

• GOES-R L1b Data 24/7

12

7
L1b calibration

• CDRL and calibration equation review (ITAR
  controlled)
• Reflectance/radiance trade study
• Geolocation and solar/view angles
• L1b and GRB IPT participation
• Determining ABI Calibration Data in GRB
• Defining GOES-R Telemetry Needs for CWG

8

• Calibration case studies to reduce risks
• Case study 1 radiance vs. reflectance for ABI
  solar bands

9
Background

• The ABI MRD specifications for level 1b products
  are radiances for all bands,
• The AWG and data users require reflectance (or
  reflectance factor) for the solar bands (1-6),
• Radiance has additional uncertainties due to
  solar irradiance,
• The Calibration Working Group (CWG) was tasked to
  perform a trade study.

10
Scope of the Trade Study

• This study covers four areas
• 1) The impact on level 1b ground data processing
  and data format (collaboration with GS and AWG)
• 2) The impact on MRD (collaboration with MIT/LL)
• 3) Implications on the ITT algorithm delivered in
  CDRL 78 and 80 (collaboration with Flight and
  ITT).
• 4) Solar zenith angle and view angle calculation
  issues (collaboration with AWG and GS).

11
Solar irradiance uncertainties for the GOES-R/ABI
solar bands

• General agreement exists between the two commonly
  used solar irradiance models Thuillier 2002 and
  MODTRAN4/Newkur (the latter is to be used for
  GOES-R/ABI)
• In the spectral regions of interest for the
  GOES-R/ABI solar bands (green curves), the
  differences between the two models are up to 3.
  
• The CWG developed strategies to reduce this
  uncertainty and a briefing has been provided to
  the GOES-R program office.

Top Comparison between the Thuillier (red) and
MODTRAN/Newkur (blue) solar spectral irradiance.
Bottom The ratio between Newkur and Thuillier
irradiance. The spectral response functions of
the GOES-R channels 1-6 are plotted in green in
both figures.
12
Summary of the Radiance/reflectance study

• The Calibration Working Group has completed a
  trade study of the radiance vs. reflectance for
  the solar bands of ABI, with the following
  recommendations
• The L1b postlaunch requirement should be changed
  to reflectance for the solar bands, based on user
  needs and our knowledge of solar spectral
  irradiance,
• The impact of change can be minimized by carrying
  certain parameters in the data processing up to
  level 1b (the GRB scaled integers can be
  converted to either radiance or reflectance)
• The ground system should ensure the necessary
  ephemeris data and the reflectance factor
  coefficients be included in the processing and in
  the level 1b data for users to compute the
  reflectance,
• The CWG should work with both the vendor and the
  ground system to ensure the correct calibration
  algorithm is used in the operations.
• All related documents from this study are
  delivered to GOES-R.

13

• Calibration case studies to reduce risks
• Case study 2 Bias between NWP/GDAS and sounding
  channel observations

14
Calibration study to reduce risks for NWP

• Instrument and calibration changes directly
  impact the data assimilation
• GOES detector temperature setting causes errors
  (contribution to penalty) in NWP
• Methodology being developed to address this issue.

GOES 12 Sounder CH 2
GOES detector temperature winter setting
acceptable level
GOES detector temperature summer setting
Global Data Assimilation System Monitoring
Courtesy of J. Derber, NCEP
Operational calibration directly impacts data
assimilation in NWP
15
Satellite Observation Bias Relative to NWP Model
Predictions
WEATHER
NCEP Global Data Assimilation System (GDAS)
(temperature bias NWP model Satellite)

• Some channels still have large biases
• Study with NIST suggests spectral uncertainties
• As NWP model accuracy improves, it demands more
  accurate satellite data
• Study provide calibration risk reduction for
  GOES-R ABI sounding channels

GOES 11 Sounder CH15
Bias between NWP model and satellite observations
up to 6 degree Kelvin !
Calibration biases need to be resolved for
improved numerical weather prediction
16
GOES-12 Channel 15 Biases (IASI as a Reference)
Ch14 Ch 15
Ch 16
GOES-12 sounder observations
IASI-simulated GOES sounder
The patterns are quite different between GOES
sounder and IASI for Ch 15
17
IASI versus GOES12 Sounder
8.4 cm-1 (34 of bandwidth)
Before shift
After shift
Assuming GOES rad. does not change but is
converted to BT based on new SRF
18

• Vicarious calibration site selection and
  under-flight planning for GOES-R/ABI

19
Vicarious Calibration Sites

• 6 Sites considered
• MODIS clear sky 1km data used to study site
  stability
• Sonoran desert time series
• The only target for GOES-WEST
• Effect of precipitation
• Uyuni salt flats (Bolivia)
• Site comparisons
• White sand overflight

20
Candidate targets for the vicarious calibration
of GOES-R ABI
First five sites reference Teillet et al.
(2007), Prime candidate Earth targets for the
post-launch radiometric calibration of
space-based optical imaging instruments, Proc.
SPIE, Vol. 6677, 66770S (2007)
DOI10.1117/12.733156.
21
Currently focusing on these four sites which
are generally spatially homogeneous
Images from Google Earth
22
MODIS/Aqua clear-sky observations of Sonoran
Desert, Mexico

• The TOA reflectance has both seasonal variations
  and long-term trends due to a number of factors,
  including the bidirectional reflectance
  distribution factor (BRDF), precipitation in the
  rainy seasons (Jan, Feb., and Aug.).

23
Multi-year average reflectance value with one
sigma uncertainty (percent)
The Uyuni Salt Flats site is the brightest in the
visible bands.
24
Overall Site Assessment in Bands 0.64 um and 0.86
um
25
Upcoming AVIRIS Underflight over White Sands

• White Sands BRDF Measurements
• Airborne Visible/Infrared Imaging Spectrometer
  (AVIRIS) is to be flown on-board the NASA ER-2
  over WSMR in four flight lines that intersect
  at (32 deg 55' 8" N, 106 deg 21' 04"W)
• Each line to be flown consecutively in both
  directions, to capture the BRDF over as many
  relative zenith/azimuth angles as possible
• Coordinate (if possible) with overflights of
  MODIS and/or AVHRR

• Simulating the Solar Noon Zenith Angle Seasonal
  CycleAcquisition Time Frame Early-August 09
  (Late-May 2010 if Summer monsoon cloudiness
  obscures site)Acquisition Duration Four hours
  during largest diurnal change of solar zenith
  angle

• Supporting MeasurementsAtmospheric Extinction
  NASAs AErosol RObotic NETwork (AERONET)
  sunphotometer data from White Sands-HELSTF
  Radiosonde El Paso, Texas NWS Line Office

White Sands Wx Forecast Support - El Paso, Texas
NWS Line Office

• AVIRIS CalibrationLaboratory calibration with
  respect to NIST reflectance standard before and
  after flightOverpass of Dryden airplane hanger
  roof immediately after departureOn-board lamp
  calibration

25
26

• Lunar calibration studies and recommendations for
  GOES-R/ABI

27
Lunar calibration

• ABI will scan moon 5 times faster -gt more moon
  observations per month
• Stable reflectance
• Moon provides independent monitoring for the
  solar diffuser
• GOES-R relies on scheduled moon observations
• Progress
• Precision is improved from 3.5 to 1.9
• Degradation rate is obtained with 97 confidence

28
Method 2 Masked Round Moon Integration with line
by line space count estimate
28
29
Lunar Calibration Issues

• Issue I
• Lunar Irradiance MODEL is necessary for Relative
  Response trending analysis
• ROLO/USGS is a sophisticated Lunar Irradiance
  model with uncertain funding
• Possible solutions
• Find an alternative for ROLO/USGS Lunar model
• Develop a Lunar Irradiance model
• Radiance Calibration find a stable site on the
  Moon
• Observe Moon at similar phase and scan angles

29
30
Initial recommendations for ABI

• Schedule regular moon observations (8-16) every
  month (the more the better)
• Observe the moon at similar phase angles whenever
  possible
• Keep scan mirror angles as similar as possible

31
Space Weather Instrument Calibration Support

• SEISS
• Identified discrepancies between levels in the
  treatment of accuracy
• Investigated gaps between requirements and
  planned calibration (what can or cannot be done
  prelaunch)
• Uncertainty analysis
• EHIS dead time correction
• EXIS
• Identified issues in CDRL 80
• Level 1b metadata (working with SWPC)
• SWX Metadata Splinter session on 7/23/09
• Supported peer review meetings

32
GLM cal/val support
GLM PDR Dry Run Reviewed Lockheed-Martin
cal/val plans in accordance to the GLM PORD. A
splinter cal/val meeting motivated two CWG
actions in regards to determining ABI and GLM
post-launch degradation in the solar reflective
channels 1) Explore use of deep convective cloud
method to test degradation of background images
of the TRMM Lightning Imaging Sensor (LIS) (a
proxy for GLM). 2) Create a vicarious calibration
target characterization plan.
33
GOES-12 Cooler Radiator Temperature
Instrument performance monitoring capability
development
2-day diurnal
10-day diurnal
Past 365 days (1 year)
Past 365 days (1 year) _at_satellite midnight
Since operational _at_satellite midnight (1700Z)
G-12 decontamination 02/07, 07/07, 01/09
G-12 decontamination 02/07, 07/07, 01/09
34

• Collaboration with NIST
• Developed the MOU for FY2009
• Filter SRF studies
• TXR measurements
• Use of the Synchrotron facility
• Planning

35
Spectral Response Functions

• Independent measurement of the GOES-R spectral
  response functions
• GOES-13 SRF piggy back
• Resolving on-orbit spectral issues by measuring
  the witness samples
• The use of travelling SIRCUS

36
GSICS Correction of GOES Calibration Biases

• Bias found between GOES-12 Imager Ch6 and AIRS
  (gray dots) in the GEO/LEO intercomparison
• GSICS correction (red dots) based on spectral
  shift
• GOES-12 decontamination causes large jumps in
  biases
• Serves as calibration risk reduction for
  GOES-R/ABI

37
Collaborations with other groups

• MIT/LL
• AOL TAP
• NASA
• CDRL reviews
• TIM

38
Other CWG activities

• Tour of the SOCC GOES facility (June 17, 2009)
• Meeting with the MODIS Characterization support
  team (March 31, 2009)

39
Summary

• CWG has made significant progress in most areas
• CDRL reviews and interaction with ITT
• Cal/val plan
• Calibration algorithms
• GRB L1b content in collaboration with IPTs
• Calibration risk reduction studies
• Vicarious calibration sites and underflight
  planning
• Lunar calibration
• Collaboration with NIST and MIT/LL
• Leverage GSICS developments
• Supporting ABI, SWX, and GLM
• Annual review for PSE and Calcon session in
  August