Use of a Real-Time High Resolution CONUS LDAS to initialize land-surface states in operational NWP models: case studies, results and challenges

Baron Advanced Meteorological Systems has been running an operational very-high-resolution (1km) CONUS+ land-data assimilation system for a number of years in real-time. The modeling system uses the community NOAH land-surface model as its core driver, with modifications that improve a number of science processes, including snow-pack evolution, bare-soil evaporation, retention of ponded water and re-infiltration, among many. A significant amount of work has gone into improving input QPE to permit realization of real-time radar-scale precipitation features and removal of radar-based artifacts while at the same time employing bias-correction based on recent NCEP Stage-IV QPE accumulations. Further, improvements in green-vegetation fraction data have now been implemented by switching from the standard monthly average legacy dataset to using MODIS-based estimates of real-time GVF. Software was also developed to allow complete feedback of all needed land-surface states required to initialize either WRF or legacy-MM5 operational meteorological model runs. This talk will briefly describe the most recent improvements in the LDAS but focus mainly on the sensitivities of the meteorological model forecast runs to initialization with the LDAS-generated land-surface states, including soil-temperature, frozen soil water, liquid soil-water, real-time green-vegetation fraction, canopy moisture, and assimilated snow-pack (SWE, snow-density) information. Results from both WRF and MM5 runs will be shown at a variety of spatial and temporal scales and several ongoing challenges will be described.