Validation of long-term precipitation and evolved soil moisture and temperature fields in MAPS

The Mesoscale Analysis and Prediction System (MAPS) is a state-of-the-art coupled model and data assimilation system operating over the conterminous United States (US) and producing grids for the GCIP. MAPS was developed at the NOAA Forecast Systems Laboratory (FSL) where it is run on a real-time, continuous basis. It also has been implemented in a fully operational mode at the National Centers for Environmental Prediction (NCEP) as the Rapid Update Cycle or RUC. The 40-km, 40-level MAPS coupled to soil/vegetation/snow model has been producing Model Output Reduced Data Set (MORDS) grids for GCIP since May 1996. MAPS is unique in that it provides these grids from an ongoing assimilation cycle, including evolution of soil moisture and temperature, and also snow temperature and snow depth if snow exists. The cycling of soil fields has been ongoing since April 1996, so that the MAPS cycle is, in essence, providing seasonal records of these mostly unobserved fields. In May 1997, the analysis interval was shortened from 3 hours to 1 hour, meaning that hourly data such as profiler and surface observations are now being assimilated in their full temporal frequency. From that time onward, the analyzed state of each hourly MAPS forecast consists of the previous 1-hour forecast (first guess) of all fields; atmospheric, multi-level soil, and clouds, are corrected by observations valid in a 1-hour window near the analysis valid time. These observations used in MAPS include those from rawinsondes, surface atmospheric observation stations, commercial aircraft, wind profilers, and geostationary satellites. The main question addressed in this paper is whether a coupled atmospheric/land-surface model, constrained by hourly assimilation of atmospheric observations to follow the evolution of the atmosphere accurately, can provide a realistic evolution of hydrological fields and time-varying soil fields that are not observed over large areas. A prerequisite for success is that the soil/vegetation/snow component of the coupled model, which is constrained only by atmospheric boundary conditions and definition of fields such as vegetation type and fraction and soil type, must be sufficiently robust to avoid drift over long periods of time. Comparisons of MAPS long-term precipitation against the precipitation analysis will be presented at the meeting to validate the accuracy of atmospheric forcing for soil moisture field. Soil moisture and temperature fields at different depths will be verified against station observations.