Current Status and Lessons from TRMM/GPM XCAL Calibration Activities Wesley Berg Colorado State University GPM: Global Precipitation Measurement TRMM: Tropical Rainfall Measuring Mission PMM: Precipitation Measurement Missions XCAL: PMM intercalibration working group

Global Precipitation Measurement (GPM) Mission GPM Core Satellite Dual-Frequency radar (DPR) Ku-band (13.6 GHz) Ka-band (35.5 GHz) Microwave Imager (GMI) 13 channels 10-183 GHz GPM was designed to provide the next generation of global precipitation products characterized by More accurate instantaneous precipitation estimates, particularly for light rainfall and cold season solid precipitation Unified precipitation retrievals from a constellation of microwave radiometers through the use of intercalibrated brightness temperatures and a common observational hydrometeor database consistent with the combined radar/ radiometer measurements obtained by the GPM Core Observatory. Constellation needed to provide 3-hourly global sampling.

GPM Radiometer Constellation Conical Imagers Cross-Track Sounders

GPM/TRMM Radiometer Constellation Satellite (Sensor) 6-7 GHz 10 GHz 19 GHz 23 GHz 31-37 GHz 85-92 GHz 150-166 GHz 183 GHz GPM GMI   10.65vh 18.7v/h 23.8v 36.64v/h 89.0v/h 166v/h 183.31±3, 7 TRMM TMI 10.65v/h 19.35v/h 21.3v 37.0v/h 85.5v/h GCOM-W1 AMSR-2 6.925v/h 7.3v/h 23.8v/h 36.5v/h 89.0v/h (B) DMSP F16, F17, F18, F19 SSMIS (4) 22.235v 91.655v/h 150h 183.31±1, 3, 6.6 METOP A/B, NOAA 18/19 MHS (4) 89qv 157qv 183.31±1, 3 190.31 Suomi NPP ATMS 23.8qv 31.4qv 88.2 qv 165.5qh 183.31±1.0, 1.8, 3.0, 4.5, 7.0 Megha- Tropiques SAPHIR 183.31±0.2, 1.1, 2.8, 4.2, 6.8, 11 GPM Era (Mar 2014 – Present) GPM Imager Constellation (7) GPM GMI (reference sensor) TRMM TMI GCOM-W1 AMSR2 DMSP F16, F17, F18 and F19 SSMIS GPM Sounders (6) Metop A and B MHS NOAA 18 and 19 MHS Suomi NPP ATMS *Megha-Tropiques SAPHIR TRMM Era (Dec 1997 – Apr 2015) TRMM Imager Constellation (10) TRMM TMI EOS-AQUA AMSR-E GCOM-W1 AMSR2 DMSP F11, F13, F14 and F15 SSM/I DMSP F16, F17 and F18 SSMIS TRMM Sounders (8) NOAA 15, 16 and 17 AMSU-B Metop A, NOAA 18 and 19 MHS Suomi NPP ATMS *Megha-Tropiques SAPHIR *Experimental products currently

XCAL Responsibilities and Goals Primary Tasks The XCAL team is responsible for the Level 1C intercalibrated brightness temperature files used as input to the operational precipitation retrieval algorithm for all of the GPM constellation radiometers (and TRMM constellation radiometers). Specific tasks include: Identify sensor issues affecting the calibration and stability of the Tb for each of the constellation radiometers. This involves Investigating calibration errors across scan and/or along orbit (i.e. time-dependent) Develop and apply corrections for sensor calibration issues. Derive and deliver intercalibration tables to adjust for residual sensor calibration differences in a physically consistent manner. Assess calibration of reference radiometer (GMI) Estimate calibration differences using multiple approaches (e.g. double differences, vicarious, polar matchups) Investigate both cold and warm-scene differences where applicable

XCAL Responsibilities and Goals Additional Tasks Assess uncertainties Investigate errors in RTM and geophysical parameters Results from multiple approaches Evaluate uncertainties in intercalibration techniques Document results (full transparency) Publications Working with other space agencies, related efforts (i.e. GSICS) etc. Work to improve intercalibration techniques Improved retrieval algorithms Identify and use best RTM models Screening (precipitation, RFI, sun-glint, etc.) Develop RTM model corrections?

Satellite/Instrument Characteristics Reference Sensor Attributes GPM GMI: Calibration Reference Sensor GMI Specs 10.65v/h 18.7v/h 23.8v 36.64v/h 89.0v/h 165.5v/h 183+3v 183+7v DT x CT Res in km 32.1x19.4 18.1x10.9 16.0x9.7 15.6x9.4 7.2x4.4 6.3x4.4 5.8x3.8 Beamwidth (deg) 1.72 0.98 0.85 0.81 0.38 0.37 NEDT (K) 0.96 0.84 1.05 0.65 0.57 1.5 Beam Efficiency (%) 91.1 91.2 93.0 97.8 96.8 96.5 95.2 Uncorr Nonlinearity (K) 0.2 0.1 0.5 Band Width (MHz) 100 200 400 1000 6000 4000 3500 4500 Feedhorns 1 Integration Time (ms) 3.6 Satellite/Instrument Characteristics Calibration Summary Data from on-orbit calibration maneuvers were used to check for calibration anomalies and develop corrections for calibration issues (i.e. magnetic anomalies, spillover corrections etc.) Significant changes were made to the spillover corrections based on the on-orbit maneuvers. Given limitations of pre-launch measurements this is a likely cause of significant calibration differences between sensors, particularly at lower frequencies. Calibration corrections are based on data from on-orbit calibration maneuvers and are not dependent on radiative transfer models Independent comparisons with by both Ball/RSS and XCAL indicate that the GMI calibration is very consistent with clear-sky ocean simulated Tb. A conservative estimate for the absolute calibration errors of the GMI window channels are < 1K Comparisons of the GMI 166 and 183 GHz channels with the MHS and SAPHIR cross-track sounders indicate differences of < 0.5K Nominal EIA 52.8/49.2 Orbit Inclination 65.0 deg Local Obs. Time Variable (Precessing) Altitude 407 km Reflector Size 1.22 m Sampling Interval 13.5 km Reference Sensor Attributes Has both imager and moisture sounding channels 65 degree orbit inclination provides regular daily coincident overpasses with other sensors Very well calibrated and stable

TRMM TMI V8 Calibration Updates Spacecraft attitude (updated: PR roll estimates, sun-sensor data, gyro data) View-angle offsets (updated: roll, pitch, ½ cone, start angle, timing) Updates to TMI along-scan correction (updated cold/warm scene) RFI in cold reflector correction (new) TMI antenna-pattern correction (updated) TMI reflector emisivities (updated) TMI hot-load correction (new) TMI reflector temperature (updated)

TMI Along-Scan Correction Cold and Warm Scenes *from Rachael Kroodsma, NASA GSFC Precipitation Processing System/University of Maryland

TMI Along-Scan Correction Cold and Warm Scenes (pitch/roll offset applied) *from Rachael Kroodsma, NASA GSFC Precipitation Processing System/University of Maryland

TRMM TMI V8 Geolocation/Pointing Pitch/roll offset: -0.08/-0.08 Swath Cone Angle Start Angle First pixel time S1 (10v) 49.40 63.91 -0.0966 S1 (10h) 49.50 S2 (19, 21, 37) 49.28 64.36 0.1272 S3 (85) Wentz 2015 numbers Channel Cone Angle Start Angle 10 49.43 63.70 19-85 49.30 64.10 TMI forward – aft look Tb differences before (top) and after (bottom) adjustments for the 10 GHz v-pol and h-pol channels. *from Rachael Kroodsma, NASA GSFC Precipitation Processing System/University of Maryland

TRMM TMI Emissivity and Spillover Values (Derived from Deep Space Calibration Maneuvers) Channel Pre-launch spillover RSS-Wentz spillover UCF-CFRSL spillover RSS-Wentz emissivity UCF-CFRSL emissivity 10.65 V 0.01600 0.01960 0.03218 0.03163 10.65 H 0.01930 0.02495 0.02654 19.35 V 0.02180 0.02545 0.03601 0.03586 19.35 H 0.02250 0.02466 0.03682 0.03390 21.30 V 0.02430 0.02213 0.03688 0.03832 37.00 V 0.01250 0.01839 0.03793 0.03940 37.00 H 0.01230 0.01731 0.03818 0.03680 85.50 V 0.01210 0.01773 0.04858 0.04253 85.50 H 0.01080 0.02106 0.04852 0.04045 *from Faisal Alquaied and Linwood Jones, University of Central Florida

TMI Hot Load Correction: Post-boost (2003-2013) Sun beta angle K Phase from orbit midnight (deg) *from Faisal Alquaied and Linwood Jones, University of Central Florida

Main Reflector Physical Temperature (2003-2013, yaw 0) Before Th correction After Th correction Sun beta angle Time Since Eclipse (min) *from Faisal Alquaied and Linwood Jones, University of Central Florida

1D Physical Reflector Temperature Table (Beta Angle = 43) Earth shadow (Eclipse) Post-boost array shadow (post-boost) Post-boost (2003-2013) array shadow (pre-boost) Pre-boost (2000-Jul 2001) Tphy, K Time since eclipse (min) *from Faisal Alquaied and Linwood Jones, University of Central Florida

TMI 21.3 GHz V-Pol vs. GMI 23.8 GHz H-Pol XCAL TMI/GMI Calibration Differences Radiative Transfer Model Dependency TMI 21.3 GHz V-Pol vs. GMI 23.8 GHz H-Pol MonoRTM Absorption Rosenkranz Absorption Raises the issue of consistency. If XCAL uses MonoRTM for calibration, but retrieval algorithm uses other model, inconsistency will lead to retrieval errors.

XCAL TMI/GMI Calibration Differences Radiative Transfer Model Dependency

Summary GMI is extremely well calibrated and stable, providing an ideal calibration reference for the constellation radiometers. Potentially an excellent absolute calibration reference! The current operational Level 1C Tb datasets for the radiometer constellation have been intercalibrated to GMI V4 and is publicly available from NASA PPS. (Berg et al. 2016: Intercalibration of the GPM Microwave Radiometer Constellation, J. Atmos. Oceanic Technol.) GPM V5 reprocessing (April 2017) Changes to the GMI calibration, primarily to low frequency channels (~1K max from V4) Associated updates to the GPM constellation radiometers TRMM V8 reprocessing (Late 2017) TMI calibration updates Intercalibrated to GMI V5 Goal is high-quality TMI/GMI reference dataset from December 1997 to present Constellation sensors back to December 1997 (SSM/I, AMSR-E, AMSU-B, etc.) Lessons Importance of on-orbit calibration maneuvers Value of multiple approaches for calibration analysis Importance of transparency Value of working with instrument developers to identify instrument issues Important to identifying/correct errors based on root cause