Investigating Atmospheric Numerical Models using Meteorological and Glider Flight Recorder Data

Abstract

An on-line, glider-pilot meteorological self-briefing system was investigated. Initially, the Colorado State University Regional Atmospheric Modeling System (CSU-RAMS) was connected to the TopTask (TT) flight planning algorithm for Colorado USA and the system had success in predicting long-distance flights. Then, the system was investigated for the northeast USA using meteorological and glider flight recorder data primarily from glider contests. As a result, fundamental problems with the RAMS predictions of surface temperatures and dew points were discovered and minimized by improving the solar-radiation and the surface-flux models. Additionally, coincident RAMS-TT and TOPTHERM-TT predictions for the northeast USA were comparable. This result is encouraging because the RAMS is three-dimensional while TOPTHERM is two-dimensional. Further, using the TOPTHERM-TT system in the northeast USA appears feasible. But, accurate predictions of surface dew-points from both models remained a challenge.