In July 2001, an improved NOAH land-surface model version was implemented in the NCEP operational mesoscale Eta Analysis and Forecast System, with NOAH-LSM upgrades related to soil heat flux, canopy conductance, bare soil evaporation, and cold season processes (snowpack and frozen soil physics). Previous off-line NOAH-LSM tests, and subsequent coupled-Eta-model month-long winter, spring and summer retrospective runs (North American model domain) showed positive impacts (see http://lnx63/NOAHLSM_bundle_dir/NOAHLSM_bundle.html). Additionally, concurrent implementation of a precipitation assimilation system (using real-time hourly radar and automated rain gauge estimates) in the Eta Analysis system improved model precipitation fields, providing improved NOAH-LSM soil moisture fields.

Monthly assessments of the NOAH-LSM coupled to the Eta model continues, e.g. using NCEP Forecast Verification System (FVS) statistics of 2-meter air temperature and relative humidity. An FVS temperature processing error was recently discovered and corrected; there is now better agreement between model and observations, particularly for summertime higher-greenness regions. (See http://www.emc.ncep.noaa.gov/mmb/research/nearsfc/nearsfc.verf.html.) Monthly precipitation plots help follow trends in Eta model versus observed precipitation, and resulting NOAH-LSM soil moisture fields. (See http://www.emc.ncep.noaa.gov/mmb/gcp/h2o/index.html.) Monthly scatter plots (for cloud-free conditions) of GOES satellite-based skin temperature versus Eta model output are consistent with other verification tools. (See http://orbit-net.nesdis.noaa.gov/goes/gcip/html/scatter.html.)

Further NOAH-LSM improvement continues. Soil heat flux under snow currently uses a combined snowpack+upper soil layer thermal conductivity, however for patchy snow-cover (varying between 0 and 1), non-snow-covered ground with greater thermal conductivity was unaccounted for. In our new formulation, conductivity is determined from both the (fractionally weighted) snowpack+soil layer (assuming parallel heat flow) and non-snow-cover (soil only). Soil heat flux is increased, especially under patchy-shallow snow, partially offsetting nighttime snow-surface radiative cooling. By itself, this new formulation was recently tested and will be implemented in the coupled-Eta-model to address a strong nighttime cold bias over snow. Additionally, even for patchy snow, surface moisture flux is attributed entirely to snowpack sublimation (until snow vanishes), resulting in surface moisture flux overprediction since sublimation proceeds at the potential evaporation rate. In our new formulation, surface moisture flux over patchy snow is generally less since it now includes both sublimation and less-than-potential non-snowpack moisture flux (canopy/bare soil evaporation, plant transpiration). This should improve longer-term model runs, e.g. in recent off-line PILPS2e (Sweden) tests, wintertime snowpack had been sublimating away too quickly. These and other changes will be assessed in NOAH-LSM off-line tests, then in coupled-Eta-model retrospective runs.

To unify land-surface parameterization at NCEP, plans include NOAH-LSM testing in all NCEP/EMC regional and global, weather and climate, data assimilation and forecast model systems. For example, in addition to the Eta model, the NOAH-LSM is being used in the NCEP 25-year Regional Reanalysis project, tested in the global MRF model, and will soon be the default option in the mesoscale WRF (Weather Research and Forecasting) model (to be implemented operationally at NCEP in the future). Such unification allows land-surface states, surface characteristics, and model physics parameters to be more appropriately 'exchanged' between the various modeling systems at NCEP (and elsewhere using the NOAH-LSM).