Studies on the Improvement of Mesoscale Numerical Weather Predictions Using Continuous Four-Dimensional Data Assimilation

点击翻译按钮获取中文摘要

Rapid advances in remote sensing technology have made possible nearly continous monitoring of the atmosphere in fine spatial de- tail. Several new observing systems, including wind profilers, NEXRAD doppler radars, automated surface observing stations, computer-assisted aircraft observing systems, and improved satellite instrumentation have either been recently deployed, or will be deployed in the U.S. in the 1990's. At the same time computer facilities with order-of magnitude increases in computational capability have become available to the National Weather Service (NWS). These advances offer opportunites for significant improvements in numerical weather prediction (NWP), exemplified by development of a 30-km resolution mesoscale model (the . model) at the National Meteorological Center (NMC), scheduled for operational use in about 1992. A critical element in the advancement of NWP is the more accurate specification of the initial state, using the full capability of the advanced systems to be deployed. The work proposed combines two mathematical approaches ("nudging" and the adjoint equation method) in the development of a 4-dimensional (spatial and time) data assimilation (4DDA) scheme. It carries through to the testing and evaluation of such a scheme at PSU. The NMC has already expressed interest, contigent upon the PSU research results, in transferring the scheme and testing it in an operational environment.