Status report from Atmosphere Product Teams Teruyuki Nakajima EORC/JAXA 2nd Japan-Australia GEO-LEO Applications Workshop Tokyo, Japan Sep. 1st-2nd, 2016

JAXA Earth Observation Satellite missions Targets (JFY) 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Disasters & Resources Climate Change & Water Cycle Water Cycle Climate change Greenhouse gases [Land and disaster monitoring] ALOS-2 / PALSAR-2 ALOS-Next Optical TRMM / PR 2013~ [Precipitation 3D structure] with NASA Feasibility study GPM / DPR with NASA [Wind, SST , water vapor, precipitation] GCOM-W / AMSR2 [Vegetation, aerosol, cloud, SST, ocean color] GCOM-C / SGLI [Cloud and aerosol 3D structure] EarthCARE / CPR with ESA [CO2, Methane] [CO2, Methane, CO] GOSAT / FTS, CAI 2009~ GOSAT-2 with MOE JMA meteorological satellites MTSAT-1R (Himawari-6) MTSAT-2 (Himawari-7) [Cloud, aerosol, SST] [Cloud, SST] Himawari-8/AHI Himawari-9 (standby) Mission status On orbit Development Study Pre-phase-A

GCOM-C science targets: Radiative forcing Today’s the most significant factor: atmospheric CO2 EarthCARE/CPR 3D structure of cloud and aerosol Aerosol radiative forcing change Cloud GCOM-C Global/horizontal distribution of cloud and aerosol Monitoring and process investigation about cloud and aerosol by GCOM-C & EarthCARE Today’s the most significant uncertainty of radiative forcing is direct/indirect role of cloud-aerosol system Climate model prediction present and future cloud and aerosol roles in the global warming scenarios Improvement Evaluation of model outputs and process parameterization

Visible and Near-infrared Radiometer SGLI instrument Polarization (along-track slant) telescopes (P1-2, three pol. angles) SGLI channels CH   Lstd Lmax SNR at Lstd IFOV VN, P, SW: nm T: m VN, P: W/m2/sr/m T: Kelvin VN, P, SW: - T: NET m VN1 380 10 60 210 250 VN2 412 75 400 VN3 443 64 300 VN4 490 53 120 VN5 530 20 41 350 VN6 565 33 90 VN7 673.5 23 62 VN8 25 VN9 763 12 40 1000 VN10 868.5 8 30 VN11 200 P1 P2 SW1 1050 57 248 500 SW2 1380 103 150 SW3 1630 3 50 SW4 2210 1.9 211 T1 10.8 0.74 340 0.2 250/500 T2 12.0 Visible and Near-infrared Radiometer (VNR) Non-polarization telescopes (VN1-11) InfraRed Scanner (IRS) (SW1-T2) Multi-angle obs. for λ= 670nm and 865nm Here is the SGLI spec. SGLI consist of two main components. One is VNR and second is IRS. VNR is further devided into NP and P. This table shows channel spec. Main resolution is 250m which is finer than that of GLI. Backward (-45deg.) Forward (+45 deg.) Nadir Along track Multi-channel Obs. @Nadir (17 chs) Swath: 1150km（VN, P） 　　　 1400km（SW, T） (250m resol.: 11chs, 500m resol.: 2 chs, 1km resol.: 4chs） SGLI can also cover the globe every 2days

GCOM-C Science targets Integrated use of multiple sensors and parameters Data merger through sensor cross-calibration and product cross-validation GCOS requirements GCOM-C observation Algorithm feedback SGLI VNR one-day coverage Surface albedo and land cover Surface solar irradiance and Photosynthetically available radiation Aerosol amount and properties NPP/VIIRS Sentinel-3/ OLCI synthetic use of the global data Geophysical connection Land vegetation, ocean chlorophyll-a, and primary production Himawari-8.9 /AHI Sea and Land surface temperature Cloud fraction and properties 5

GCOM-C aerosol estimation: SGLI NUV (380nm) band Hokkaido Alaska Lake Baikal Arctic Sea Ice North Pole GLI：19th May 2003 R: 0.678μm G: 0.545μm B: 0.380μm Greenland Scandinavian Peninsula (T.Y.Nakajima, Tokai Univ.)

GCOM-C aerosol estimation Use of SGLI polarization observation POLDER experience Experience of satellite POL data analysis POLDER BPDF data base (function of land cover classification and vegetation index) has been provided by Dr. Bréon under JAXA/SGLI and CNES/POLDER/3MI collaboration Global aerosol optical thickness in June 2003 using POLDER-2 polarization reflectance (provided by I. Sano, Kinki Univ.) Difference of SGLI from POLDER 1-km resolution Cloud contamination will be improved than POLDER 1-km scale land cover and geographical influence should be confirmed (applicability of the POLDER ground BPDF) Combined use of the nadir-slant views for aerosol type estimation (influence of IFOV, registration..) Single viewing angle (+45 or -45 along track) Single scattering angle condition (mostly in 60~120 degrees) Sunglint over the ocean (and flood land?) Tau 7

GCOM-C Operation Flow (Atmosphere) Blue: standard products, red: research products and double-line boxes are subject for level-3 production Calibration by JAXA Level-1B 1km/250m 1km/500m/250m 1km VNR-NP IRS SWIR IRS TIR VNR-POL Application Land surface reflectance Suzuki (AORI): model Tachiiri(JAMSTEC) Earth system model Precise geometric correction (land/common) Gridded mosaic data (1km & 1/24 deg) JAXA daytime & nighttime Cloud area/ classification Cloud flag & phase Nakajima (Tokai U.) Cloud detection daytime daytime Cloud properties O2 band analysis Aerosol properties day&night Water cloud mechanical thickness Optical thickness & effective radius of Water cloud Cloud-top temp/ height Ocean & Land aerosol by NUV Land aerosol by Pol Kuji (Nara W.U.) Mukai (Kyoto IU.) JAXA Ice cloud OPT Sekiguchi (Kaiyo U) :Multi pixel method daytime Cloud amount/class Radiation budget analysis Surface longwave radiation Nakajima (Tokai U.) Nakajima (Tokai U.) Ishimoto (MRI) Non-spherical model JAXA Surface shortwave radiation Irie (Chiba Univ.): SKYNET K. Aoki (Toyama Univ.): SKYNET Riedi (Lille, Fr) Cloud by Pol Imasu (AORI):Airborne Val Hayasaka (Tohoku Univ.): BSRN, SKYNET Yamazaki (MRI)): Val

Validation of GCOM-C atmosphere products Match-up analysis with SKYNET, AERONET and other observation groups Uncertainty assessment@pixel is required for model assimilation Empirical approach Validation @ASRVN, AERONET-OC, in-situ observation champagne (TBD).. AERONET comparison in each condition Theoretical analysis Error of satellite sensor calibration Calibration from pre-launch to on-orbit, and vicarious adjustment Error dependency of the algorithm and observation condition Surface reflectance error relating with its brightness, directionality and variability (vegetation) Locality of the aerosol properties (size and absorption, with humidity?) Error sensitivity on the satellite & solar geometries (scattering angle) Contamination by clouds, shadow, snow, and sunglint Area Group Product Release threshold Standard accuracy Target accuracy Atmosphere Cloud Cloud flag/Classification 10% (with whole-sky camera) Incl. below cloud amount Classified cloud fraction 20% (on solar irradiance)*8 15%(on solar irradiance)*8 10%(on solar irradiance)*8 Cloud top temp/height 1K*9 3K/2km (top temp/height)*10 1.5K/1km (temp/height)*10 Water cloud OT/effective radius 10%/30% (CloudOT/radius) *11 100% (as cloud liquid water*13) 50%*12 / 20%*13 Ice cloud optical thickness 30%*11 70%*13 20%*13 aerosol Aerosol over the ocean 0.1(Monthly a_670,865)*14 0.1(scene a_670,865)*14 0.05(scene a_670,865) Land aerosol by near ultra violet 0.15(Monthly a_380)*14 0.15(scene a_380)*14 0.1(scene a_380 ) Aerosol by Polarization 0.15(Monthlya_670,865)*14 0.15(scene a_670,865)*14 0.1(scene a_670,865) *8: Comparison with in-situ observation on monthly 0.1-degree *9: Vicarious val. on sea surface and comparison with objective analysis data *10: Inter comparison with airplane remote sensing on water clouds of middle optical thickness *11: Release threshold is defined by vicarious val with other satellite data (e.g., global monthly statistics in the mid-low latitudes) *12: Comparison with cloud liquid water by in-situ microwave radiometer *13: Comparison with optical thickness by sky-radiometer (the difference can be large due to time-space inconsistence and large error of the ground measurements) *14: Estimated by experience of aerosol products by GLI and POLDER

JAXA Himawari-8 products (planned) Himawari-8 Product development in EORC JAXA Himawari-8 products (planned) Level Product name Grid size Format L1 Reflectance (6 bands) Brightness temperature (10 bands) 0.02~0.05 Equal lat-lon grid (Full disk) 0.01 (around Japan) NetCDF4 L2 Atmosphere Aerosol optical thickness, Angstrom exponent, cloud properties Ocean Sea surface temperature Ocean color (Rw, Chla..)* Land Fire detection, Vegetation index, Snow cover Flux Short-wave radiation, PAR, UV*.. L3 Hourly average Underlined products are produced regularly (* will be open soon) Other products are under investigation JAXA products are distributed by FTP with simple user registration http://www.eorc.jaxa.jp/ptree/

TOA reflectance RGB Solar irradiance (SWR, PAR) Aerosol (AOT) Aerosol (Angstrom exponent) SST Chlorophyll-a concentration

Vicarious calibration coefficients Band 1 Band 2 Band 3 AHI AHI/sim=1.007 AHI/sim=0.999 AHI/sim=1.014 Simulated Simulated Simulated Inter-band calibration based on B4 (0.86m) and B5 (1.6m) The correction coefficients were stable (and nearly 1.0) since 7 July 2015 Rrs470 Rrs510 Rrs640 Chla AHI with corr AHI w/o corr AHI w/o corr AHI w/o corr AHI w/o corr Chla needs the accurate gain correction

Thank for your attention Teruyuki Nakajima EORC/JAXA 2nd Japan-Australia GEO-LEO Applications Workshop Tokyo, Japan Sep. 1st-2nd, 2016