Improving the Land Surface Forecast in Navy’s Global Atmospheric Model

Global extended long-range weather forecast is a near-future goal at Navy’s Fleet Numerical Meteorology and Oceanography Center (FNMOC). In an effort to improve the performance of the Navy Global Environmental Model (NAVGEM) operated at FNMOC, and to gain understanding of physics impact on the long-range forecast, the physics package of the U.S. Global Forecast System (GFS) is being evaluated in the framework of NAVGEM. That is GFS physics being transported by NAVGEM Semi-Lagrangian advection, and update-cycled by NAVGEM 4D-variational data assimilation along with the assimilated land-surface data of U.S. Air Force’s Land Information System. The focus of this study is on the improvement of land surface boundary condition that plays an important role in the long-range NWP forecast. The output of free long runs of 10-day GFS physics forecast in a summer and a winter season are evaluated through the comparisons with the output of NAVGEM physics and through the validations with observations. It is found that the GFS physics is able to effectively reduce some of the modeling biases of NAVGEM physics, especially for the land surface temperature and relative humidity that are the major components of surface boundary condition. The correction increases with forecast leads. To understand the interactions between the land surface and upper-air atmosphere, the cloud water, precipitable water, and precipitations of GFS and NAVGEM physics are compared over land and ocean respectively. The comparisons reveal NAVGEM produces much less cloud water of liquid and ice than GFS physics does, which explains to some extend why NAVGEM generates a warming trend in the summer and a cooling trend in the winter at the surface. Also the surface dry bias error in the NAVGEM forecast, particularly in summertime, is found to be partially associated with the loss of precipitable water of NAVGEM, whose global water vapor decreases with time and suffers a 10% loss at the end of each 10-day forecast. However, the GFS physics in NAVGEM shows little surface moisture bias and conserves the global water. This study helps to understand the strength of GFS physics in the land surface forecast and its overall impact on NAVGEM prediction system.