1st GPM Workshop on Algorithms & Data Products

1. Global Precipitation Measurement (GPM) Mission An International Partnership & Precipitation Satellite Constellation for Research on Global Water & Energy Cycle Overview & Preliminary Planning • 1st GPM Workshop on Algorithms & Data Products • Eric A. Smith; NASA/Goddard Space Flight Center, Greenbelt, MD 20771 [tel: 301-286-5770; fax: 301-286-1626;easmith@pop900.gsfc.nasa.gov;http://gpmscience.gsfc.nasa.gov] • March 12-13, 2002; Greenbelt, MD

2. Purpose of Workshop I. Review preliminary planning & define continuing agenda for precipitation algorithms & data products development (Erich Stocker's main concern). II. Review & approve conceptual algorithm & data products architecture that enables TSDIS group to make progress now (Ziad Haddad, Chris Kummerow, Bob Meneghini, John Stout, Wai-K. Tao). III. Connect algorithm/data products program to GV program & continuous error characterization, noting separation of science from requirements (Sandra Yuter). IV. Foreword three (3) important recommendations to GPM instrument manager (Mark Flaming) concerning specifications for core satellite radiometer, i.e., GPM Microwave Imager (GMI): (a) exact frequency definitions; (b) contiguity requirements & implications; (c) priorities for high frequency channels -- assuming they are provided.

3. Integrated Precipitation Data Plan (IPDS): Accelerated Plan Primary IPDS Elements • Advanced Data System • Data Archive & Distribution • CalVal Simulator • Science Guidance

4. IPDS Architecture NewDISS Formulation Initiative GHCC SSMIS/AMSR Generic I/F Tools & Data Formats Evolution SAN Interface World Precipitation Data System TRMM Testbed I/F Simulator Testbed I/F Std’s Evolving Advanced GPM Algorithms Experimental Rain Products & Error Characterization Geo IR DAAC Megha Trop Dual Frequency DPR Synthesizer EGPM Validation / Error Estimates Science Guidance CalVal Simulator Field Campaigns Ground Instruments Precipitation Algorithm Release Beta V6g V7g V8g TRMM GPM V0 V1 V2 V3 V4 V6 V7 V8 V9 V5 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

5. Strategy for Populating Constellation Reference NPOESS-1 GPM Core (CMIS) N-GPM b DMSP-F18/20 (N-PMR) (ATMI/DPR) (SSMIS) E-GPM NPOESS-2 (E-PMR) DMSP-F19 TRMM (CMIS) Optimization and Compromise Co-Op Drone Partners Potential New Drones/Partners (SSMIS) DMSP-F16 AQUA FY-3 NPOESS-Lite DMSP-F17 ADEOS-II (CMIS) NPOESS-3 (C-PMR) (CMIS) MEGHA TROPIQUES GCOM-B1 (MADRAS) (AMSR-FO) TBD

6. GPM Operations & Data Flow Core DMSP GCOM TDRSS Continuous Science Data NASA Drone Partner Drone NASDA Operations White Sands IPO Operations Command Telemetry Per Orbit GHCC/MSFC Potential ESA (E-GPM) & ASI (I-GPM) Partnerships Per Orbit Ancillary Data Streams GSFC Global Precipitation Data Center (GPDC) IR Data (GOES) NASA Mission Operations Center AMSU (POES) Coordination GPM Partner Mission Operations/Data Center Near Real-time Rainfall 3-hr Rain Map Climate Data EOSDIS GP Data Broadcast Network Continuous Real Time Data Links Once Per Orbit Data Links Command/Telemetry Links Virtual Ground Station Global Precipitation Network

7. Tracibility to ESE's Strategic Plan IV. What are consequences of change in Earth system for civilization? (Consequences) 1. How are variations in local weather, precipitation & water resources related to global climate variation? 2. What are consequences of land cover & use change for sustainability of ecosystems & economic productivity? 3. What are consequences of climate & sea level changes & increased human activities on coastal regions? V. How well can we predict future changes in the Earth system? (Prediction) 1. How can weather forecast duration & reliability be improved by new space obs, data assim, & modeling? 2. How well can transient climate variations be understood & predicted? 3. How well can long-term climatic trends be assessed & predicted? 4. How well can future atmospheric chemical impacts on ozone and climate be predicted? 5. How well can cycling of carbon through Earth system be modeled, & how reliable are predictions of future atmospheric concentrations of carbon dioxide & methane by these models? How is Earth changing and what are consequences for life on Earth? I. How is global Earth system changing? (Variability) 1. How are global precip, evap, & water cycling changing? 2. How is global ocean circulation varying on interannual, decadal, & longer time scales? 3. How are global ecosystems changing? 4. How is stratospheric ozone changing, as abundance of ozone-destroying chemicals decreases & new substitutes increases? 5. What changes are occurring in mass of Earth’s ice cover? 6. What are motions of Earth & its interior, & what information can be inferred about its internal processes? II. What are primary forcings of Earth system? (Forcing) 1. What trends in atmospheric constituents & solar radiation are driving global climate? 2. What changes are occurring in global land cover & land use, & what are their causes? 3. How is Earth’s surface being transformed & how can such information be used to predict future changes? III. How does Earth system respond to natural & human-induced changes? (Response) 1. What are effects of clouds & surf hydrology on climate? 2. How do ecosystems respond to & affect global environmental change & carbon cycle? 3. How can climate variations induce changes in global ocean circulation? 4. How do stratospheric trace constituents respond to change in climate & atmospheric composition? 5. How is global sea level affected by climate change? 6. What are effects of regional pollution on global atmosphere, & effects of global chemical & climate changes on regional air quality? Asrar, G., J.A. Kaye, & P. Morel, 2001: NASA Research Strategy for Earth System Science: Climate Component. Bull. Amer. Meteorol. Soc., 82, 1309-1329.

8. GPM Mission is Being Formulated within Context of Global Water & Energy Cycle with Foremost Science Goals Focusing On • Improved Climate Predictions:through progress in quantifying trends & space-time variations of rainfall & associated error bars in conjunction with improvements in achieving water budget closure from low to high latitudes -- plus focused GCM research on advanced understanding of relationship between rain microphysics/latent heating/DSD properties & climate variations as mediated by accompanying accelerations of both atmospheric & surface branches of global water cycle. • Improved Weather Predictions:through accurate, precise, frequent & globally distributed measurements of instantaneous rainrate & latent heat release -- plus focused research on more advanced NWP techniques in satellite precipitation assimilation & error characterization of precipitation retrievals. • Improved Hydrological Predictions:through frequent sampling & complete continental coverage of high resolution precipitation measurements including snowfall -- plus focused research on more innovative designs in hydrometeorological modeling emphasizing hazardous flood forecasting, seasonal draught-flood outlooks, & fresh water resources prediction.

9. Is Water Cycle Accelerating?

10. TRMM-retrieved Radar & Radiometer Rainfall Anomaly Time Series with MSU-retrieved ice index (DCI) (anomalies normalized by annual mean) (F.R. Robertson; GHCC)

11. (1) Global Water & Energy Cycle Processes & Modeling: Role of Precipitation (Prof. Eric Wood, Princeton) (2) Climate System Variability & Climate Diagnostics: Role of Precipitation (Dr. Franklin Robertson, NASA/MSFC @ GHCC) (3)Climate Model Simulations & Reanalysis, NWP Techniques, & Data Assimilation: Role of Precipitation (Dr. Arthur Hou, NASA/GSFC) (4) Land Surface Hydrology & Hydrometeorological Modeling: Role of Precipitation (Dr. Harry Cooper, Fla. State Univ.) (5) Ocean Surface & Marine Boundary Layer Processes: Role of Precipitation (Dr. Vikram Mehta, NASA/GSFC) (6) Coupled Cloud-Radiation Modeling: Physical Interpretation of Precipitation Processes (Prof. Gregory Tripoli, Univ. Wis. & Dr. Wei-K. Tao, NASA/GSFC) (7a) Precipitation Retrieval: Reference Radar-Radiometer Core Algorithm & Radar Simulator (Dr. Ziad Haddad, NASA/JPL) (7b) Precipitation Retrieval: Parametric Radiometer Constellation Algorithm & Radiometer Simulator (Prof. Christian Kummerow, Colo. State Univ.) (7c) Precipitation Retrieval: Cal Transfer, Bias Removal, & Merged Products (Prof. Eric Smith, NASA/GSFC) (8) Calibration & Validation of Satellite Precipitation Measurements (Dr. Sandra Yuter & Prof. Robert Houze, Univ. Washington) (9) Forecast Apps, Public Service, TV, & Educational Outreach (Dr. Marshall Shepherd, NASA/GSFC) GPM Mission's Nine (9) Science Themes USGCRP WCSG, 2001: A plan for a new science initiative on the global water cycle. UCAR Report., sponsored by USDA, DOE, NASA, NOAA, NSF, USGS, ACE, BREC, & EPA, 118 pp. Atmosphere Weather, Climate, & Hydrology Water Land

12. 1 • GPM Disciplinary Research Themes, Science • Requirement Groups, & Main Provisional Science Objectives 1 6 7a 2 3 7b 7c 4 8 9 5

13. Proposed GPM Precipitation Products I. Level 2 Orbit Swath Products (similar to TRMM) II. Level 3 Grid Products (modification to TRMM) ∑ arithmetically consistent succession: (1) 3-hourly @ 0.1 x 0.1 deg; (2) daily @ .25 x .25 deg; (3) pentad @ 1.0 x 1.0 deg; (4) monthly @ 5.0 x 5.0 deg ∑ simplified & integrated parameter set: (1) surface rainrate/rainfall & grid variance; (2) convective-stratiform separation; (3) bulk DSD parameters; (4) latent heating profile; (5) confidence index ∑ four (4) types of retrieval results ordered by retrieval quality: (1) core sat result (combined DPR - GMI); (2) constellation sat result (rain radiometers) (3) non-rain radiometer result (e.g., AMSU); (4) geo-infrared result (GOES, GMS, METEOSAT, MSG) III. Full Resolution Pixel Tables (extension to TRMM) ∑ tabulates all instantaneous rain pixels from all satellite sources within individual 3-hourly 0.1 x 0.1 deg grid elements, identified in space & time IV. Level 4 Scale Invariant Blended Products (new) ∑ properly transformed to constant space-time scales Goal is relatively seamless transition of precipitation products from TRMM to GPM eras, thus minimizing mission change impacts on science community.

14. How is Seamlessness Achieved? I. Level 1 Swath Products: Mostly same procedures & definitions, but with greater emphasis on calibration quality assurance -- this is business of both project & science team with project providing continuous quality checking, TSDIS conducting zero-biasing of constellation members relative to core, & science team providing oversight. II. Level 2 Swath Products: Mostly same procedures & definitions, but with greater emphasis on accuracy of instantaneous retrievals & validation of "unfolded" physical algorithms -- this is strictly business of science team coordinated with TSDIS. III. Level 3 Grid Products & Full Resolution Pixel Tables:Modified procedures & definitions -- this is strictly business of TSDIS because these products are directly derivable from level 2 products, should be user-friendly, & should be arithmetically consistent. IV. Level 4 Scale Invariant Blended/Gridded Products: New procedures & definitions -- this is a research problem to be addressed by science team and to be implemented on TSDIS when properly mature.

15. Accelerated Algorithm Development 1. Design & test GPM-era "precipitation product" generation system, as continuation of level-1 & level-2 "algorithm version" system adopted for TRMM, to enable seamless product transition from TRMM-era to GPM-era [GPM-era products will eventually be brought on-line for routine processing based on approval of"Integrated Precipitation Science Team"]. 2. Develop, code, & test three (3) main new GPM-era precipitation retrieval algorithms: (i) level-2 reference DPR-GMI combined, (ii) level-2 parametric PMW radiometer, & (iii) level-3 merged-gridded products at time-space resolutions of: (a) 3-hour / 0.1'deg (b) 1-day / 0.25'deg (c) 5-day / 0.5'deg (d) 1-month / 5.0'deg to determine computational performance in TSDIS / GPDC environment -- to be ready for implementation in late TRMM era [advanced GPM algorithms will not be vulnerable to drop-outs of any particular satellite, instrument, or channel -- including core satellite instruments]. 3. Develop level-3 merged-gridded products at hierarchy of above space-time scales, all arithmetically consistent, containing 5 types of results: (a) DPR-GMI combined result [#allpix, aveRRallrain , sigRRallrain , #convpix, aveRRconv , sigRRconv , #stratpix, aveRRstrat , sigRRstrat , #unknpix, aveRRunkn , sigRRunkn , confindexREF, LHREF(4), DSD(3)] (b) constellation rain radiometer result [#allpix, aveRRall , sigRRall , #convpix, aveRRconv , sigRRconv , #stratpix, aveRRstrat , sigRRstrat , #unknpix, aveRRunkn , sigRRunkn , confindexCON, LHCON(4)] (c) non-rain radiometer (e.g., AMSU) result [#allpix, #rainpix, aveRRrain, sigRRrain , confindexNRR] (d) GEO IR result [#allpix, #rainpix, aveRRrain , sigRRrain , confindexIR] (e) merged-weighted combination of above 4 results [mergeRR , confindexMERGE] 4. Develop pixel-level precipitation table product (sattype, time, lat, lon, RR, raintype, confindex) using all existing satellite rain sources, with pointers to "gridded" 3-hourly level-3 product, to enable detailed/high resolution space-time analysis [must not break TRMM paradigm that precipitation products are sized for "reasonable" communications & computer analysis efficiency, i.e., "no Huge & Unwieldy files"].

16. Scientific Challenges for GPM Mission 1. Shift intellectual inquiry paradigm from "curiosity driven" to "quintessential problem driven" -- through GPM science team working group coordination. 2. Shift research paradigm from "measuring takes precedent" to "prediction takes precedent" -- through mandate for GPM science team. 3. Shift derived products paradigm(e.g., latent heating, DSDs, macro/microphysical cloud properties, error characterization, solid precipitation, vertical rain mass flux) from "cautious release" to "aggressive release" -- through modeler involvement in product assessment. 4. Shift fast delivery data paradigm from "only operational users need them" to "research users need them too" -- through transfer of specialized data products from GPM-WPDC to research partners conducting prediction experiments. 5. Shift validation paradigm from "comparison scatter diagrams" to "physical error modeling" involving inverting flow of data from & to validation center and deploying new ground instruments at various validation supersites. 6. Shift cloud/precipitation paradigm from "these are separate & distinct problems" to "this is microphysical continuum" leading to integrated cloud-precipitation missions, research programs, textbooks, and teaching. Coordinate GPM Science Implementation Plan with GEWEX

17. Backup Slides

18. Error Characterization 1 Bias & Bias Uncertainty based on: ∑ physical error model ( passive-active RTE model ) ∑ matched satellite radiometer/radar instrument on ground with continuous calibration (eyeball) ∑independent measurements of observed inputs needed for error model all retrievals from constellation radiometers & other satellite instruments are bias adjusted according to bias estimate for reference algorithm from core satellite

19. Error Characterization 2 Precision & Space-Time Error Covariance based on: ∑ ground Radars ( polarization diversity enables cross-checking ) ∑ high-quality, uniformly distributed, dense, & hi-frequency sampled raingage networks

20. Principles of Physical Error Modeling ∑ difference between retrieved & RTE modeled hydrometeor profiles yields retrieval "bias" ∑ mismatch in 2-ended RTE model solution based on absorption-scattering properties assigned to characterization volume yields retrieval "bias uncertainty" Note: ground Radars & regional raingage networks in conjunction with coincident satellite retrievals can generate "space-time correlation structure functions" & "space-time error covariance matrices". Radiance Tube TOA TOA meteorology & microphysics characterization volume observed by independent instruments ∑ Doppler Profilers (2) for precip/cloud hydrometeor profiles ∑ Radiosonde for T(z)/q(z) profiles ∑ Downward-pointing Radiometer for surface emissivity ∑ Raingage & Disdrometer Network for cross-checking matched core satellite radiometer/Radar ground instrument “Eyeball” [with active target calibration] Surface Surface

21. Mid-Lat Continental Tropical Continental Tropical Oceanic Extratropical Baroclinic GPM Validation Strategy I. Basic Rainfall Validation ∑ Raingauges/Radars new/existing gauge networks new/existing radar networks Research Quality Data Confidence sanity checks II. GPM Supersites  Basic Rainfall Validation hi-lo res gauge/disdrometer networks polarametric Radar system  Accurate Physical Validation scientists & technicians staff data acquisition & computer facility meteorological sensor system upfacing multifreq radiometer system Do/DSD variability/vertical structure convective/stratiform partitioning GPM Satellite Data Streams Continuous Synthesis ∑ error variances ∑ precip trends Calibration Algorithm Improvements Supersite Products III. GPM Field Campaigns  GPM Supersites cloud/ precip/radiation/dynamics processes  GPM Alg Problem/Bias Regions targeted to specific problems FC Data Research ∑ cloud macrophysics ∑ cloud microphysics ∑ cloud-radiation modeling High Latitude Snow

22. Potential GPM Validation Sites Canada England Germany NASA Land Spain Italy South Korea ARM/UOK Japan NASA KSC Taiwan France (Niger & Benin) India NASA Ocean Brazil Australia Regional Raingage Site Supersite Supersite & Regional Raingage Site