Ocean EDR Product Calibration and Validation Plan For the VIIRS Sensor for Ocean products

Transcription

1 DRAFT Cal Val Plan VIIRS Workshop Ocean EDR Product Calibration and Validation Plan For the VIIRS Sensor for Ocean products Developed by the Government Ocean Team representing (NOAA, NAVY. NASA, University ) Robert Arnone, NRLSSC April 1, 2009 Corvallis, Oregon 1

2 VIIRS Ocean Cal-Val Overall Approach 1) Develop an integrated cross-agency Cal/Val plan for the ocean EDR products that is based on present infrastructure and architectures used for heritage satellite sensors. (operational at NASA, NOAA, NAVY) 2) Ensure consistency of Ocean EDR products with heritage satellite products (from e.g. SeaWiFS, MODIS, MERIS and AVHRR) through extensive cross-evaluation and inter- comparison. 3) Construct a readiness program beginning with MODIS and MetOp-AVHRR FRAC (1km global) as a pathfinder to ocean calibration and validation for Ocean EDRs for NPP VIIRS and extending to VIIRS on NPOESS C1 and C2. Environmental Data Records (EDRs) Rrs, SST, Chl, IOPs EDRs are calculated from Sensor Data Records (SDRs) which are the calibrated and geolocated at sensor radiances. Navy, NOAA, NASA and NGST will all produce EDRs which need to be compared and evaluated. 2

3 Time schedules in preparation for NPP: I. Prelaunch period emphasis on heritage instruments AVHRR, SeaWiFS and MODIS. review, unify, evolve and document the methods and resources required for operational calibration and validation; set up procedures and protocols to effectively communicate the Cal/Val results within the Team, and between the Team, IPO, NGST and the user Community; set up methodology and tools to Quality Control in-situ data used in validation, satellite-derived products and associated clear-sky radiances over oceans; set up a near-real real time data stream to generate proxy VIIRS SDRs (by reformatting MODIS data into NPOESS HDF5 format and simulating VIIRS performance characteristics); investigate the potential for at least two targeted prototype campaigns to evaluate the credibility of operational NPOESS validation strategies. 3

4 II. Intensive Cal/Val program immediately post launch acquire VIIRS SDRs, generate Navy, NASA and NOAA heritage ocean products and process them through the heritage Calibration and Validation algorithms. conduct targeted validation campaigns and demonstrate methods and use of operational field data sets. conduct a rigorous inter-comparison and cross- evaluation of the SDR-derived government products, and the NGST standard products (SST, Rrs and chl.) Evaluate Government VIIRS ocean products (NASA, NOAA, NAVY) against NGST products. Evaluate Government, University and NGST products against heritage products. Evaluate products against in-situ observations. 4

5 III. Define a long term Cal/Val monitoring system add new data streams (SDRs and government derived products, and associated clear-sky radiances over oceans) to the existing long-term monitoring data bases. increasingly rely on operational in situ monitoring sites such as, buoys for SST comparisons, MOBY or follow-on for ocean color radiances, coastal sites and AERONET Ocean Sites with SeaPRISM and others with reliable validated in-situ ocean data. analyze long-term performance of NPOESS products relative to the heritage products. 5

6 Ocean product Calibration Validation The The Plan and The Execution 6 investment areas 1 Define a VIIRS Proxy Data Stream 2 Define the required in situ data stream for Cal/Val 3 Tuning of algorithms and LUTS (Vicarious calibration and SDR feedback) 4 Ocean Algorithm, evaluation and uncertainty 5 Product validation and product longterm 6 Not in order or priority Satellite intercomparisons, robustness, seasonal and product 6

7 Table 1: Tasking linked to the Major Ocean Cal Val Components 1 Proxy VIIRS Data Stream 2 "Insitu data collection 3 Vicarious Calibration and LUT tuning 4 Algorithm Performance 5 Product Validation 6 Inter-satellite calibration 1. Software versions - MODIS to VIIRS Proxy 6. Moby or AHAB calibration insitu data 12. Evaluation of Vicarious cal using MOBY, AERONET, Open Ocean Gyre water 17. Evaluation of ocean color Algorithms with VIIRS Proxy Data 21. Automated match up of ocean color imagery products with insitu data 25. Establish protocols for ocean color products continuity (MODIS/MERIS) )Seasonal/Temporal evaluation 2. Integrate Sensor characteristics into Software versions (Crosstalk, Polarization, etc) 7. Aeronet SeaPrism data stream 13. Automated tuning of the SST algorithms 18. Evaluation of SST Algorithms with VIIRS Proxy data 22. Automated (Bulk/Skin) with matchup 26. Establish protocols for SST product continuity- MODIS/AVHRR/ METOP Seasonal/Temporal evaluation 3. Integrate Proxy software \into NRTPE 8. Ship Cruise Data field data stream, bio-optical and MERI- buoy measurements 14. Coupling skin and bulk measurements MERI and Buoys comparison and analyses for SST 19. Evaluation of the Atmospheric Correction methods 23. Ship cruise data (SST and Color) match up with satellite imagery 27. Demonstrate real time continuity of Ocean Color products 4. Demonstrate VIIRS data stream to users. 9. Insitu data Management tools 15. Develop methods For automated Vicarious Calibration 20. Comparison of radiative transfer SST models 24. Determine if the available insitu data can validate Products space/temp 28. Demonstrate real time Continuity of SST products 5. Model VIIRIS proxy data using MODIS Lw to TOA 10. Buoy match ups Of SST 16. Generation of Rayleigh Look up tables 11. Instrument upgrades and NIST traceability 7

8 Ondrusek 15, DiGiacomo 43 Siegel 44 Letelier 8

9 Current Efforts Organ. Sponsor Contact Objective Cross thread the agency contributions with the SST investments Vicarious Proxy VIIRS Data "Insitu" data calibration and Stream collection LUT tuning Demonstrate VIIRS data stream to users. Model VIIRIS proxy data using MODIS Lw to TOA Insitu data management tools Buoy match ups of SST Instrument upgrades and NIST traceability Develop methods for automated Vicarious Calibration Generation of Rayleigh Look up tables 4 Algorithm Performance Comparison of radiative transfer SST models 5 Product Validation Determine if the available insitu data can validate products space/tempora l 6 Inter-satellite calibration Establish protocols Evaluation of Automated for ocean color Vicarious cal Evaluation of match up of Software Identify versions - Moby or AHAB the Gaps and products responsibilities. continuity using MOBY, ocean color ocean color MODIS to VIIRS calibration (MODIS/MERIS/ AERONET, Open Algorithms with imagery Proxy insitu data SeaWIFS) Ocean Gyre VIIRS Proxy Data products with Seasonal/Temporal water insitu data evaluation Coordinate across disciplines Establish protocols Integrate Sensor Automated for SST product characteristics into Aeronet - Automated Evaluation of SST matchup of continuity- Ocean Software versions SeaPrism data tuning of the algorithms with SST MODIS/AVHRR/ Color (Crosstalk, stream SST algorithms VIIRS Proxy data (Bulk/Skin) METOP, Polarization, etc) with matchup Seasonal/Temporal evaluation Identify international efforts Ship Cruise Coupling skin Ship cruise Data - field and bulk Evaluation of the data (SST and Demonstrate real Integrate Proxy data stream, measurements Atmospheric Color) match time continuity of software into NRTPE bio-optical and MERI and Buoys Correction up with Ocean Color MERI- buoy comparison and methods satellite products measurements analyses for SST imagery Demonstrate real time Continuity of SST products NRL SPAWARS Rowley Data assimilation / diurnal variability NAVY IPO May SST algorithms evaluation NOAA IPO Ignatov SST algorithms evaluation U. Miami IPO Minett, Evans SST algorithms evaluation NOAA NOAA STAR Research activities NASA NASA Feldman, Patt NASA PEATE / prepare for the NPP U/ Wisconsin IPO?Gunther? upgrade the MERI instrument NAVY NAVO May, Coulter prepare for NPOESS, IDPS etc NOAA NOAA - NDE Staphaluss Data formats for VIIRS NRL SPAWAR Lyon Coulter, May, NAVO- NAVO Patterson Gap fillers, Automated validation sites, APS upgrade identify data stream, IDPS setup at NAVO Atmospheric correction - Rayleigh tables NOAA IPO Wang NOAA NOAA Ondrusek Moby buoy, AHAB Oregon State IPO Davis cal val report NASA NASA Feldman. Patt NASA PEATE / prepare for the NPP NASA NASA - NPP Turpie Evaluation of VIIRS sensor and product NOAA NESDIS DiGiacomo Research for VIIRS into Coast Watch NOAA NOAA Operations Staphaluss Preparing VIIRS for NOAA Operations NOAA NOAA - NDE Staphaluss Data formats for VIIRS NIST IPO Johnson Sensor characterization NRL IPO Lutz, Snyder Sensor characterization and impact on atmospheric correction NASA IPO Esaias Ocean color NRL IPO Arnone Coordination team 9

10 1. VIIRS Proxy Data Stream. 1. Establish a proxy data stream using MODIS (RTPE) data. Address the sensor issues and data format issues to prepare the govn t t and community for VIIRS data. 2. Include realistic estimates of errors and sensor uncertainties based on laboratory calibrations data to determine SDR uncertainty prior to launch. 3. Establish spiral development of software updates to the Proxy data stream 1. Coordinated with the IPOs GRAVITE for 5 software distribution 4 Product validation and Ocean product longterm 4. Provide proxy software and data stream Algorithm, 3 evaluation and in real time for operational and uncertainty research evaluation. 1 Define a VIIRS Proxy Data Stream Tuning of algorithms and LUTS 2 (Vicarious calibration Define the and SDR feedback) required in situ data stream for Cal/Val 6 Satellite intercomparisons, robustness, seasonal and product 10 Not in order or priority

11 What being executed.. 1) (cont.) Proxy Data Stream FY ) V 1.0 In place -- Role of GRAVITE data Format NRPE NRTPE (NASA) (Global) Data stream -- Joe Zajic, 2) VIIRS Ocean color - HDF5 format into Navy APS Lyon, Martinolich - Hyperspectral Convolution with VIIRS / MODIS - P. Lee, P. Lyon, M. Wang, C. Davis 3) VIIRS SST - D. May, Cayolga,, B. Evan, P. Minnett, - Ignatov, B. Emery 4) Integration with Updates from VIIRS sensor specs - Cross Talk, Spectral, Polarization etc. - Continued development of Proxy data stream as we gain more information about the specific characteristics of VIIRS data. A Real-time Data Stream being developed --- To Prepare for VIIRS Data Stream Assess the Algorithms 11

12 Ocean Color Proxy Data Stream Level 1- MODIS Top of Atm Radiance MODIS - channels Top of Atm Radiance VIIRS (Lt (Lt (v) ) MODIS water Remote Sensing Reflectance Derive Water Properties (IOP( IOP) At 6 λ s Optimization Algorithm Atm. Correction Rayleigh Aerosols ρw sw Level 2- MODIS Absorption Scattering Expand IOP at MODIS λ ρ w To Continuous Spectrum Note: ρ = reflectance Expand MODIS to Hyperspectal Aerosol Spectrum Expand MODIS to Hyperspectal Rayleigh Spectrum MODIS hyperspectral Continuous Water Spectrum Expand Continuous Spectrum To Top of Atmosphere τ a2 = λ 1 τ a1 λ 2 λ -4 [ ] η ρ w Lt (v) = Lw (v) + La (v) +Lr (v) La(λ) Lr(λ) Lw(λ) Convolve the VIIRS spectral response with 3 HSI components. White Paper in draft Looking for Contributions 12 Team forming

13 2. In-situ data stream for cal val Addresses 22 types of insitu data - one for vicarious calibration (e.g. MOBY) and a second for product validation. 1. Define where and what in-situ data is required and how it is integrated into the end-to-end product cycle. 2. Define the critical validation site locations and ocean extremes required for validation. 3. Define the spatial and temporal uncertainty of the observations and the measurement protocols for both calibration and validation. 4. Determine protocols and the NIST traceability which are required for each insitu data set. 5. Identify the data base manager for these insitu data. Insitu Cruise measurements SST Buoy and Skin Ocean Color Radiance, IOP, Chl. 1 Define a VIIRS Proxy Data Stream 3 Tuning of algorithms and LUTS 2 (Vicarious calibration Define the and SDR feedback) required in situ data stream for Cal/Val 4 Ocean Algorithm, evaluation and uncertainty 5 Product validation and product longterm 6 Satellite intercomparisons, robustness, seasonal and product 13 Not in order or priority

14 2. (cont.) Insitu Data Collection 1) SST match up using data from Buoys 10000/ 2 weeks ( May ) 2) Moby Ondrusek et al. (Real time Web access) 3) A-MERI (surface skin temperature) Cruises Minnett/ Evans 1) Must correct for Skin vs. Bulk relationships 2) International UK, Web sites ( Minnett, May, Robinson) 4) AUV- SST Emery 5) Aeronet/ SeaPRISM Goal Real-time Web Availability) 1) CUNY New York (Ahmed) 2) Gulf of Mexico (Lyon, Vaughn, Scarborough, Ladner, Martinolich) 3) West Coast (Davis, Jones) 4) East Coast MVCO, Ch. Bay (Hou, Hooker) 5) International (Zibordi( et al., ) ( Paper on the Software and protocols) 6) Australia (Decker) 6) Ship Cruises 1) Insitu match up data sets 1) West Coast Time Series Davis, Letelier 2) Navy / NURC Cruises Lyon, Gould, Trees. 3) NOAA Cruises (Chesapeake Bay ) Ondrusek, Stumpf 6) Data management tools (Fargron( / NASA -- SeaBASS) 7) International IOCCG (Davis, Lee DiGiacomo, Fargron,, Trees) 8) Sampling Strategy / Uncertainty (Trees/ Alverez) 14

15 3) Tuning of algorithms and LUTS (Vicarious Calibration and SDR feedback) 1. Integrate the tuning of the Look up Tables that are introduced at the SDR level with ocean observations for both ocean color and SST 2. Define the timeliness and update cycle by defining the temporal uncertainty and of the sensor and the algorithms 3. Develop capability to perform rapid matchups of satellite and ocean o parameters to optimize the ocean products. 4. Automate the vicarious calibration procedures and integrate within in the SDR- EDR levels. 5. Define vicarious calibration strategy for initialization (address s immediate post-launch) and for long term. 6. Address multiple approaches for initial vicarious calibration. Address the needs for short-time time scales. 6 5 (Climatology, satellite matchup and insitu.) 1 Define a VIIRS Proxy Data Stream 3 Tuning of algorithms and LUTS 2 (Vicarious calibration Define the and SDR feedback) required in situ data stream for Cal/Val 4 Ocean Algorithm, evaluation and uncertainty Product validation and product longterm Satellite intercomparisons, robustness, seasonal and product 15 Not in order or priority

16 3. (cont.) Vicarious Calibration / Tuning LUT FY09 FY 10 1) Automate Ocean Color Vicarious methods on Different data sets (Aeronet, Ship, MOBY) 1) Lyon, Vaughn, Navy NAVO (transition) Inversion of L2Gen 2) Wang / Ondrusek - NOAA - New Methods 3) NASA (Werdell, Bailey, Franz) 2) Rayleigh Look up Tables (Wang) 3) SST --- Skin/bulk relationships -- Tuning Minnett/ Evans 4) Radiative transfer simulations - Ignatov Web sites NOAA 5) GRIST Minnett., May International - Web sites ) 6) Integration of Vicarious Calibration within GRAVITE and IDPS Zijac, Guenther 16

17 4) Ocean Algorithm, evaluation and uncertainty 1. Evaluate aspects of algorithm performance with VIIR proxy data and follow-on sensor data. (e.g. the atmospheric correction in ocean color products) 2. Examine algorithms over the full range of expected conditions and define the limitations and uncertainties associated with seasonal variations, latitudinal variations, coastal vs. open ocean, cloud / land boundary affects, varying atmospheric types (aerosols and Total Perceptible Water). 3. Examine algorithms uncertainty from variety of default inputs to the algorithms - (Determine the ensemble spread or the algorithm s uncertainty maps.) 4. Determine the impact of sensor on the algorithm and product uncertainty. 1 Define a VIIRS Proxy Data Stream 2 Define the required in situ data stream for Cal/Val 3 Tuning of algorithms and LUTS (Vicarious calibration and SDR feedback) 4 Ocean Algorithm, evaluation and uncertainty 5 Product validation and product longterm 6 Not in order or priority Satellite intercomparisons, robustness, seasonal and product 17

18 4) Algorithm Evaluation - FY ) VIIRS proxy data / MODIS / MERIS data - Proxy Data Stream 1) NRL Stennis (Lyon, Vaughn, Arnone, Gould, Ladner) 2) West Coast ( Davis, Letelier) 3) Coastal Areas Stumpf 2) NIR/ VIIRS Atmospheric Correction 1) Wang NIR/ SWIR VIIRS / MODIS algorithms 2) Gao - MODIS/ VIIRS 3) SST 1) SST regional studies VIIRS / MODIS/ AVHRR 1) May, NAVY 2) Radiative Transfer Simulations / Studies Ignotov Web sites 3) Regional studies MODIS / to AVHRR to VIIRS (Evans, MInnett) 4) Others

19 5) Product validation and product long-term (initially based on VIIRS Proxy data, MODIS, METOP etc) 1. Comparisons of ocean waters observations (SST and Rrs - Color) with derived satellite products including their uncertainty. 1. Develop procedures and protocols on match up methods for insitu and satellite products. 2. Define the spatial and temporal uncertainty of the match-up of insitu data to the satellite retrieved geophysical parameter. 2. Determine how to evaluate ocean products in a range of environmental ntal conditions which are defined through required data sets. 1. Ocean product evaluation should include the atmospheric uncertainty nty (aerosols and TPW) and the ocean product. 2. Define the error estimates of the products in different conditions, ns, to include (sensor configurations, solar cycles, coastal vs. open ocean, o etc.) 5. Determine how the ocean product validation is influenced by the sensor? 6 5 Satellite intercomparisons, 6. Time series Buoy SST monitoring and SeaPRISM RRS tracking 4 Product robustness, validation and seasonal and 1. Define Protocols for tracking validation procedures. product longterm product 3 2. Leverage off on MODIS/ SeaWIFS, AVHRR/ METOP 1 Define a VIIRS Proxy Data Stream Tuning of algorithms and LUTS 2 (Vicarious calibration Define the and SDR feedback) required in situ data stream for Cal/Val Ocean Algorithm, evaluation and uncertainty 19 Not in order or priority

20 5) (cont.) Product Validation for FY ) Ocean Color Validation Match- up and uncertainty 1) Automated match up Ship (Trees, Davis, Lyon, Stumpf, Ondrusek) 2) Automated matchup Aeronet / SeaPRISM 1) (Lyon, Ahmed, Davis, Jones, Zibordi 2) Web site - for near real-time Spatial / Temporal / Measurements and uncertainty analysis 3) Coordinate with IOCCG International efforts (DiGiacomo, Davis, Lee) 2) SST match up 1) Buoy Bulk temperature Match up May, Minnett 2) Spatial uncertainty - AUV Emery 3) Ship transects skin/bulk uncertainty Minnet/ Evans 4) International GHRIST (May, Minnett, Evans, Robinson) 3) Uncertainty of insitu data (SST / Color) Almeriz/ / Trees 20

21 6) Satellite Continuity Inter- comparisons, robustness, seasonal and product Baseline approach (Define the methods which VIIRS will be evaluated based on existing ing ocean products.) 1. Define the protocols to compare ocean satellite products from different satellites 1. Use as a pathfinder MODIS, SeaWIFS, MERIS, METOP,. 2. Determine the spatial and temporal uncertainty of match-up differences between satellite products. 1. Spatial variability of sensors regional differences 2. Examine the seasonal differences 3. Demonstrate the ability for satellite ocean product continuity in near real time. 1. Time series analyses of regional products (e.g. SST variability) 2. Determine how to define methods to track product 4. Demonstrate the methods which NPP will evaluated against heritage ocean products 1. Spatial and temporal statistical analyses (binning, histograms etc.) 1 Define a VIIRS Proxy Data Stream 3 Tuning of algorithms and LUTS 2 (Vicarious calibration Define the and SDR feedback) required in situ data stream for Cal/Val 4 Ocean Algorithm, evaluation and uncertainty 5 Product validation and product longterm 6 Satellite intercomparisons, robustness, seasonal and product 21 Not in order or priority

22 6. (cont.) Satellite Data Continuity 1) Ocean Color Continuity 1) Protocols for SeaWiFS / MODIS / MERIS / OCM / Sentinel 2 / HICO 1) Coastal and Regional Studies (Stumpf) 2) VIIRS proxy / MODIS/ MERIS color products 1) West Coast - (Davis, Letelier, Jones) 2) East Coast - (Wang, Ondrusek, Ahmed) 3) Gulf of Mexico & International Areas (Arnone, Lyon, Gould) 3) Continue using MOBY for vicarious calibration (Ondrusek) 2) MODIS/ MERIS - Global Studies - Bailey, NASA 3) IOCCG GlobColour International Coordination (Fargion( Fargion) 2) SST AVHRR / METOP / MODIS / 1) Integrate MODIS cal val procedures into Operations May / Minnett 2) MODIS / AVHRR Protocols May / Minnett / Evans 3) VIIRS/ MODIS comparison / regional studies, May 4) NOAA Operation SST Ignatov. 22

23 Summary We have begun to identify roles and responsibilities Beginning execution Interagency and Integrated Program Office (IPO)/NGST efforts are being coordinated. Preparing in situ capability. Preparing way Forward through heritage Sensors. Looking forward to soonest possible launch of NPP. 23