Mitchell THORPEX Workshop GFS Noah 18Jan06

The new Noah LSM in the NCEP GFS: (implemented 31 May 2005)1) pre-operational tests2) implications for ensemble land states 3) snowpack & streamflow for THORPEX-IPY : 

The new Noah LSM in the NCEP GFS: (implemented 31 May 2005) 1) pre-operational tests 2) implications for ensemble land states 3) snowpack & streamflow for THORPEX-IPY Ken Mitchell NCEP/EMC THORPEX Workshop 17-19 January 2006

Slide2: 

Land-Surface Modeling Takes Root in NCEP Operational Weather and Climate Prediction: The Multi-disciplinary Development of Land-Surface/Hydrology Modeling and Land Data Assimilation at NCEP via Multi-Institutional Partnerships

Slide3: 

Improving Weather and Climate Prediction: Becoming a Complete Earth System Endeavor 1 - ATMOSPHERE: troposphere, stratosphere (GARP) - initial conditions require atmosphere data assimilation 2 - OCEAN: deep ocean, seas, coastal ocean, sea ice (TOGA/CLIVAR) - initial conditions require ocean data assimilation 3 - LAND: soil moisture, snowpack, vegetation, runoff (GEWEX/GAPP) - initial conditions require land data assimilation

GEWEX: Global Energy and Water Cycle Exp.GAPP: GEWEX Americas Prediction Project (NOAA/OGP): 

GEWEX: Global Energy and Water Cycle Exp. GAPP: GEWEX Americas Prediction Project (NOAA/OGP) 1980s TOGA (CLIVAR) - Brought together atmospheric modelers and oceanographers 1990s GEWEX Hydrology Panel GAPP-GCIP, BALTEX, GAME, LBA - Brought together atmospheric modelers and hydrologists

Slide5: 

Partnerships NOAA Office of Global Programs (OGP): GEWEX, GAPP (GCIP) Long term funding of 3-4 members of EMC Land Team WRF Land Working Group JCSDA: Joint Center for Satellite Data Assimilation Community Noah Land-Surface Model (Noah LSM) from OSU to Community Noah LSM (includes WRF “Unified Noah LSM) Noah LSM development and performance Physics, optimal parameters, initial states, validation Uncoupled: 1-d and 3-d 1-d:column model 3-d: N. American LDAS (NLDAS) & Global LDAS (GLDAS): Coupled: N. American Mesoscale Model: NAM/NDAS (Eta-Noah & WRF-Noah) N. American Regional Reanalysis (NARR): Eta-Noah Global Forecast System: GFS/GDAS Climate Forecast System: CFS The long time scales in LSM soil moisture response 1-2 year spin-up in soil moisture states Diversity in the “inherent” soil moisture climatology of land models Land Data Assimilation Systems (LDAS) Coupled, Uncoupled, Quasi-coupled (“Hybrid”)

Slide6: 

Partnerships GAPP WRF

History of the Noah LSM: 

History of the Noah LSM Oregon State University: 1980’s OSU/CAPS LSM Funded by Air Force Geophysics Lab (AFGL) Tested in AFGL MM5 and Global Spectral Model Transitioned to AFWA (AFGWC) late 1980’s Implemented in AFWA AGRMET in 1990 USAF global uncoupled land data assimilation system B. Moore, G. Gayno Transitioned to NCEP NWP models in 1990’s Coined “NOAH” LSM after NCEP, OSU, Air Force and OHD upgrades Transitioned to NCAR in late 1990’s F. Chen: MM5, Transitioned to WRF: NCAR, NCEP, AFWA, FSL

Land Surface Evaporation Treatment : 

Land Surface Evaporation Treatment WHERE: E = total evapotranspiration from combined soil/vegetation Edir = direct evaporation from top soil layer Et = transpiration through plant canopy via root uptake Ec = evaporation from canopy-intercepted precipitation or dew Es = sublimation from snowpack + Es

Noah LSM Testing Sequence: 

Noah LSM Testing Sequence Uncoupled tests 1-d column model 3-d NLDAS and GLDAS Coupled Noah Tests Eta or WRF mesoscale model (NAM) Global Forecast System (GFS) Coupled Forecast System (CFS)

Slide11: 

Soil heat flux improvements in uncoupled testing of the community 1-D NOAH LSM at the flux site of Meyers and Hollinger (1998), 7-day time series of observed (circles) versus modeled soil heat flux for: Top Panel: old (dashed) versus new (solid) thermal conductivity formulation (both without new vegetation effect), during April 1998, Bottom Panel: new thermal conductivity formulation without (dashed) and with (solid) new vegetation effect, during August 1998. Exp: 1-d model dev. Soil heat flux improvements in uncoupled testing of the Community 1-D NOAH LSM soil heat flux (W/m*m) AUGUST 1998 (high vegetation greenness) NEW without veg effect OBS NEW with veg effect Julian day APRIL 1998 (low vegetation greenness, moist soil) soil heat flux (W/m*m) OLD NEW OBS Julian day 300 200 100 -100 0 100 50 -50 0

Slide12: 

old model formulation - cool, moist bias in 2-m T, Td new model formulation – reduced cool, moist bias better ABL representation as well 36-hr OLD 2-meter T=> Td=> NEW 2-meter T=> Td=> Champaign, Illinois Example 3-D test in Eta Mesoscale Model

Noah LSM in the N. American Regional Reanalysis(NARR): 

Noah LSM in the N. American Regional Reanalysis (NARR)

Slide14: 

NARR: Soil moisture availability (percent of saturation) Top 1-meter of soil column 1993 1988 Average during 16-31 July (at 21 GMT)

Slide15: 

NARR: Atmospheric Boundary Layer Depth (m) 1993 1988 Average during 16-31 July (at 21 GMT)

3-D Tests in Uncoupled N. American LDAS(NLDAS): 

3-D Tests in Uncoupled N. American LDAS (NLDAS)

Slide17: 

N-LDAS Design (The Uncoupled Approach) 1. Force models with 4DDA surface meteorology (Eta/EDAS), except use actual observed precipitation (gage-only daily precip analysis disaggregated to hourly by radar product) and hourly downward solar insolation (derived from GOES satellites). 2. Use 4 different land surface models: NOAH (NOAA/NWS/NCEP) MOSAIC (NASA/GSFC) VIC (Princeton U./ U. Washington) Sacramento (NOAA/OHD) 3. Evaluate results with all available observations, including soil moisture, soil temperature, surface fluxes, satellite skin temperature, snow cover and runoff.

NLDAS Simulated River System: 

NLDAS Simulated River System Upstream area [log10(km^2)] Travel time to outlet [days] Large River basins River flow direction mask

Slide19: 

East Fork of White River at Columbus, IN

Slide20: 

Seasonal change of total column soil moisture: 30Apr minus 30Sep, 99, at 23Z

Slide21: 

NLDAS soil moisture Validation across U. Oklahoma Mesonet: Top 40 cm Jan 98 – Sep 99

Slide22: 

COLORADO NORTHEAST NORTHCENTRAL NORTHWEST SNOWMELT SUBLIMATION ALBEDO BASIN AVERAGE PERCENT SNOW COVER The Noah LSM systematic early bias in spring-season snowpack depletion.

Slide23: 

V2.7 Noah LSM: New treatment of evaporation and albedo over snow Validation at 110 SNOTEL sites in Western U.S. (solves the early snowmelt bias)

GFS Implementation31 May 2005: 

GFS Implementation 31 May 2005 NCEP TPB: http://www.emc.ncep.noaa.gov/gc_wmb/Documentation/TPBoct05/T382.TPB.FINAL.htm Increase in horizontal resolution Land surface model upgrade New sea-ice model Enhanced mountain blocking Modified vertical diffusion Analysis upgrades Additional satellite radiance data Enhanced quality control Improved surface emissivity calculations over snow

Slide25: 

GFS: Noah LSM versus legacy OSU LSM 4 soil layers (10,30,60,100 cm) vs. 2 soil layers (10, 190 cm) land surface evaporation: reduced high bias in warm-season vegetation cover: improved properties and seasonality improved seasonal cycle of green vegetation fraction spatially varying root depth (1-2 m) vs. constant 2 m add frozen soil physics (freeze/thaw latent heat, limit infiltration) snowpack physics improvements: greatly reduced early melt bias add snow density state variable (retain SWE) retain some snowmelt in snowpack and allow refreezing refine functions for snow cover fraction and snow albedo add patchy snow cover treatments to snow sublimation, sensible & ground heat flux, skin temp improved numerics/robustness for very shallow snow transpiration: refine soil moisture threshold for stress onset direct soil evaporation: revise dependence on soil moisture smaller ground heat flux bias especially: wet soil, under snowpack, under dense vegetation new functions for soil thermal diffusivity and soil heat capacity

Impact of Noah LSM replacing OSU LSMin GFS: 

Impact of Noah LSM replacing OSU LSM in GFS Key Objectives: Reduce high bias in summer land surface evaporation Reduce high bias in mid-latitude convective precipitation Reduce early bias in snowpack depletion

Slide27: 

09-25 May 2005 17-day mean surface Latent heat flux Operational GFS Parallel GFS test

Slide28: 

Equitable Threat Score Bias: Ops GFS: solid Test GFS: dashed Precipitation Validation Scores: East half of CONUS 60-84 hour GFS fcst from 00Z 12-31 May 2005

Vision for NCEP Land Data Assimilation: 

Vision for NCEP Land Data Assimilation Add members to NCEP Ensemble GFS suite via perturbed land states Produce ensemble river flow forecasts from ensemble GFS runoff forecasts Utilize NASA-NCEP Land Information System (LIS) Ensemble Kalman filter (JCSDA PI research) GLDAS or NLDAS mode Global or regional domain Highly parallelized software architecture and fast ESMF utilized Add CRTM: JCSDA Community Radiative Transfer Model Assimilate microwave satellite estimates: Soil moisture and snowpack