Project Overview

Expensive, weather-sensitive tests at the Army ranges sometimes must be planned months or years in advance, well beyond the reach of conventional numerical weather prediction (NWP). As a solution to this dilemma, RAL is using the range's observed weather over recent decades (its regional climate, in other words) to make long-range, probabilistic weather forecasts.

Most datasets about past climate are too coarse, by themselves, to be useful for this approach, so RAL is adding detail to these datasets through a method called dynamically downscaling. During downscaling, gridded climate data from coarse sources such as the NCAR-NCEP Reanalysis Project (NNRP) or the North American Regional Reanalysis (NARR) are assimilated into the Advanced Research core of the Weather Research and Forecast (WRF) Model via RAL's own Real-Time Four-Dimensional Data Assimilation (RT-FDDA) system. The WRF-based RT-FDDA then produces a series of very detailed mesoscale re-analyses of past weather. These re-analyses are stored, studied, and translated into statistical descriptions of the range's regional climate. Forecasters at the ranges will be able to access these statistics and provide fast, customized, cost-saving, probabilistic weather predictions to test directors.

Statistical summary of wind at 10 m above ground for a site at the Aberdeen Test Center. The wind rose is created from 3 months of mesoscale reanalyses, during which time the average wind direction was 234 degrees from north, and the average speed was 4 knots. The frequency of winds at certain speeds from specific directions is plotted according to the color bar. The mean wind speed in each directional bin is noted as a black number. The frequency of wind in each bin, no matter the speed, is indicated by the range rings. For example, the single most frequent direction was from the south-southwest (13% of the time) at an average of 5 knots.