1. Application of Satellite Land Surface Observations in NCEP Models Weizhong Zheng 1,2 and Mike Ek 1 1 NOAA/NCEP/Environmental Modeling Center, College Park, MD 2 IMSG @NOAA/NCEP/EMC, College Park, MD Contributions: Jiarui Dong, Yuhua Wu, Helin Wei and Jesse Meng (NCEP/EMC); Marco Vargas, Zhagyan Jian, Xiwu Zhan, Jicheng Liu, Yuunyue Yu and Ivan Csiszar (NESDIS/STAR); Zhen Song (UMD) The 14th JCSDA Technical Review Meeting & Science Workshop, May 31-June 2, 2016

2. Climate CFS Motivation  Objective: To improve satellite data utilization over land in NCEP forecast models and data assimilation system and then improve the numerical weather prediction (NWP).  Land satellite data assimilation: – Utilization of satellite data sets in the models (e.g., GVF, snow, burning area, albedo, emissivity, LST, radiation, vegetation and soil type) – Assimilation of satellite products (e.g, Soil moisture (SMAP, SMOPS); snow); – Direct radiance assimilation (Tb) (sfc parameters, sfc emissivity and sub-grid scale land surface)  Land-Atmosphere Interaction – I dentify and understand the interaction and feedback between land and atmosphere, and then improve NWP and DA. 2

3. Status of Weekly Real-Time VIIRS GVF  Objective: Use weekly real-time VIIRS GVF data set in the NCEP models to replace the old 5-year mean monthly climatology from AVHRR.  Last year: Sum/Win cases had been tested in GFS and the results were investigated (with weekly real-time VIIRS GVF).  Current status: 2014 Spring case with weekly real-time VIIRS GVF and 2015 Summer case with multi-year mean of VIIRS GVF data. Some tests also included in NAM.  Main achievements from last year. – New GVF showed a good improvement for NWP; – More issues were arisen. 3

4. Weekly Green Vegetation Fraction (GVF) Data sets NCEP Operations: Monthly 0.144-deg (16-km) global climatology of GVF from AVHRR. (Gutman & Ignatov, 1998). Weekly GVF: VIIRS near real-time weekly global 0.036-deg (4- km) GVF (Marco Vargas team). It starts from Sep. 2012 to current. Three data sets: (a)Weekly climatology GVF; (b) Monthly climatology GVF; (c) Near real-time weekly GVF The new GVF data sets can potentially improve the NWP skills, especially during the spring growing season when vegetation has large variations. 4

5. GVF: AVHRR monthly clim and VIIRS monthly mean: JAN and APR Old_Clim (AVHRR) Monthly mean Clim Jan Apr 5

6. GVF: AVHRR monthly clim and VIIRS monthly mean: JUL and OCT Old_Clim (AVHRR) Monthly mean Clim Jul Oct 6

7. T2m bias and rmse over West CONUS Aug. 1—31, 2015 Monthly climatology GVF test: Reduced RMSE (up to 0.35 °C) afternoon and nighttime. 7

8. Precipitation Skill Scores over CONUS: f36-f60 8 Monthly climatology GVF test: Slightly positive impact

9. 9

10. Test Runs in NAM 24 days (2 days each month, the beginning and the middle of each month) in 2014 were chosen for runs with NAM 3 GVFs were used – CLIMO – RGVF1—Xiaoyang Zhang – RGVF2—NESDIS – Surface update program developed by George Gayno was used to interpolate the real time GVF to NAM domain 84 hours simulation was conducted for each run Analysis was conducted over 218 GRID domain Courtesy Yihua Wu 10

11. GVF (%) Albedo (%) Date Comparison of GVF and Albedo for three data sets over 218 Grid_D Courtesy Yihua Wu – NESDIS VIIRS has several weeks behind other two data sets – GVF and Albedo data sets need to be consistent. 11

12. Comparison between VIIRS and other GVF data (AVHRR or 5yr Climatology) VIIRS and 5yr Clim. VIIRS and AVHRR Courtesy Jiang and Vargas monthly weekly 12

13. Status of Soil Moisture Assimilation  Objective: Assimilate SMOPS/SMAP SM data and improve NWP.  Last year: Some cases tested for SMOPS and SMAP SM.  Current status: More validation using more in situ SM measurements; New NASA SMAP soil moisture product test.  Main achievements from last year. – Validation indicated that NESDIS SMOPS blended SM /SMAP SM quality needs further improvement, in order to assimilate in the NCEP models. – NASA SMAP SM product was tested and the initial result showed a positive impact to improve NWP in the NCEP model. – Satellite-based land DA in GFS/CFS Ops systems are under going. 13

14. Station 2093 Solid: In-situ Diamonds: AMSR2 SM Number of days: 268 Corr. coef: 0.840 Bias: -0.042 AMSR2 SM: Phillipsburg, Kansas Station 2129 Solid: In-situ Diamonds: AMSR2 SM Number of days: 257 Corr. Coef.: 0.354 Bias: -0.131 AMSR2 SM: Milford, Utah Number of Stations: 150 Corr. Coef: 0.545 Bias: 0.021 RMSE: 0.038 AMSR2 SM vs SCAN Ground Obs: Statistics Courtesy J. Liu and X. Zhan NESDIS SMOPS: SMOS, ASCAT-A, WindSat In late 2016, it includes: AMSR2, SMAP, GPM/GMI, etc. 14

15. NASA SMAP Soil Moisture (re-scaled) 1-31 AUG 2015 Averaged: 1-31 AUG10 AUG 2015 15

16. Surface temperature and its RMSE South Plains Aug2015 16 SMAP test: Slightly reduced bias; and consistently reduced RMSE (up to 0.2 °C) afternoon and nighttime.

17. Surface temp. and its RMSE N. Plains/Mid West Aug2015 17

18. Surface DPT and its RMSE South Plains Aug2015 18

19. Temperature fits to Obs: Bias and RMSE at fh48 & fh60 CONUS 19

20. Satellite-based Land Data Assimilation in NWS GFS/CFS Operational Systems Use NASA Land Information System (LIS) to serve as a global Land Data Assimilation System (LDAS) for both GFS and CFS. LIS EnKF-based Land Data Assimilation tool used to assimilate soil moisture from the NESDIS global Soil Moisture Operational Product System (SMOPS), snow cover area (SCA) from operational NESDIS Interactive Multisensor Snow and Ice Mapping System (IMS) and AFWA snow depth (SNODEP) products. 1.Build NCEP’s GFS/CFS-LDAS by incorporating the NASA Land Information System (LIS) into NCEP’s GFS/CFS (left figure) 2.Offline tests of the existing EnKF-based land data assimilation capabilities in LIS driven by the operational GFS/CFS. 3.Coupled land data assimilation tests and evaluation against the operational system. NGGPS Project: Land Data Assimilation NASA (LIS) Michael Ek, Jiarui Dong, Weizhong Zheng (NCEP/EMC) Christa Peters-Lidard, Grey Nearing (NASA/GSFC) Jiarui.Dong@noaa.gov Courtesy Jiarui Dong 20

21. 21 Demonstration of LIS land data assimilation of AFWA Snow Depth 12/31/2014 00Z 01/01/2015 00Z 03/01/2015 00Z Open Loop Direct Insertion 02/01/2015 00Z Control Run 07/01/2014 00Z GFS/GDAS Jiarui.Dong@noaa.gov Courtesy Jiarui Dong 21

22. Comparison between EnKF and DI Jiarui.Dong@noaa.gov Courtesy Jiarui Dong EnKF DI 22

23. Status of Land-Atmosphere Interaction  Objective: Identify and understand the interaction and feedback between land and atmosphere, and then improve NWP and DA.  Last year: Continuously investigated the cause of surface temperature biases in the NCEP models which affect DA too.  Current status: Together with field exp., improve physical schemes.  Main achievements: – GFS excessive cooling and decoupling identified and fixed – Improvement of GFS surface temperatures forecast. – Improvement of brightness temperature simulation in GSI. 23

24. Why is important to study land-atmosphere interaction for data assimilation (DA) Direct radiance assimilation: Requiring a forward radiative transfer model (RTM) to calculate Tb with input of model atmospheric profiles and surface parameters. For surface-sensitive channels, Tb simulation largely depends on: (a) Sfc parameters such as LST and soil moisture; (b) Sfc emissivity (IR/MW) Thus, in order to improve radiance DA, we have to improve sfc emissivity calculation and sfc parameters simulation. Our efforts on land DA must include these two tasks. 24

25. Ops GFS or GFSX: Rapidly cooling up to 15 °C during 3hr; About 13 degrees of cold bias at 00Z, 25 Jan. GFS/GFSX T2m @ MRB Matinsburg RGNL, WV 00Z 01/24/2016 Cycle T2m @ KMRB 25

26. CTL: Rapidly cooling more than 15 °C during 3hr; EXP: Substantially improved GFS Test: T2m @ MRB Matinsburg RGNL, WV 26

27. CTL: Large difference between T1 and T2m (or Tskin) during a period of nighttime on 1/25. EXP: Substantially improved not only T2m, but also Tskin and T1. Rapidly cooling: Decoupled Improvement T1: Temperature at the lowest model level (Blue); T2m: Red; Tskin: Black GFS Test: T1, T2m and Tskin @ MRB GFS: CTL GFS: EXP 27

28. Simulated Brightness Temperature with Ops GFS METOPA/HIRS Ch 8METOPA/AMSU-A Ch 1 Tb: Large cold bias over the area where GFS LST has excessive cold bias. 28

29. Future Plan  Assimilation/utilization of satellite data sets in the NCEP models will be continuously investigated and tested.  More tests will be performed for the GVF and soil moisture data sets after they are further improved and updated.  As many parameters (e.g. soil/vegetation type) are updated in the NCEP models, future work will be performed to modify GSI/CRTM accordingly (e.g. IR/MW emissivity calculation, etc.). 29

30. Future Plan: Update of vegetation type (13 SiB to 20 IGBP) 1: broadleaf-evergreen trees! 2: broadleaf-deciduous trees 3: broadleaf and needleleaf trees! 4: needleleaf-evergreen trees 5: needleleaf-deciduous trees (larch) 6: broadleaf trees with groundcover 7: groundcover only (perennial) 8: broadleaf shrubs with perennial groundcover 9: broadleaf shrubs with bare soil 10: dwarf trees and shrubs with groundcover (tundra) 11: bare soil 12: cultivations (the same parameters as for type 7) 13: glacial ice 1:Evergreen Needleleaf Forest 2:Evergreen Broadleaf Forest 3:Deciduous Needleleaf Forest 4:Deciduous Broadleaf Forest 5:Mixed Forests 6:Closed Shrublands 7:Open Shrublands 8:Woody Savannas 9:Savannas 10:Grasslands 11:Permanent wetlands 12:Croplands 13:Urban and Built-Up 14:Cropland/natural vegetation mosaic 15:Snow and Ice 16:Barren or Sparsely Vegetated 17:Water 18:Wooded Tundra 19:Mixed Tundra 20:Bare Ground Tundra 30 Courtesy Helin Wei

31. 1loamy sand 2 silty clay loam 3 light clay 4 sandy loam 5 sandy clay 6clay loam 7sandy clay loam 8 loam 9loamy sand 1: sand 2: loamy sand 3: sandy loam 4: silt loam 5: silt 6:loam 7:sandy clay loam 8:silty clay loam 9:clay loam 10:sandy clay 11: silty clay 12: clay 13: organic material 14: water 15: bedrock 16: other (land-ice) 17: playa 18: lava 19: white sand Future Plan: Update of soil type (9 Zolber to 19 STASGO) 31 Courtesy Helin Wei

32. 32 Summary and Future Plan  Several satellite data sets developed recently (e.g., GVF, snow, burning area, albedo, radiation, soil and vegetation type) have been tested in the NCEP models. The results show good improvements, compared with the current data sets; However, some data sets need further verification with ground measurements, and consistence of all these data sets is required.  The SMOPS and SMAP soil moisture data assimilation in GFS have been continuously examined; More tests will be done after NESDIS developed the blended SMOPS including SMAP.  Large cold bias of surface temperatures in GFS was identified and substantially reduced by the proposed solution, which resulted in improvement of Tb in GSI/CRTM;  With update of NCEP models (e.g. soil and veg type), future work will be performed in modifying GSI/CRTM accordingly. 32