Weather forecast applications are ubiquitous on smartphones and occupy much of our time. However, the weather forecast is based on temporally stale information as the forecast is based upon a model that is typically run every six hours. Further, the weather forecast is based on global-scale models featuring coarse horizontal resolutions. For instance, current models used by forecasting entities may employ a horizontal grid resolution of approximately 50-100 kilometers (km) for a forecast covering an area of 1,000 km. Thus, current weather forecasts do not capture the influences of the local underlying topography and land use. Using global-scale models alone may suffice in non-dynamic weather conditions, but will typically be inadequate if weather conditions are changing rapidly or if the local area features significant terrain and/or dramatic land-use transitions.
The Army has a high need for current and accurate weather forecasts and to that point, Army researchers have developed a system configured to generate a weather forecast optimized for a local geographic zone – the one you are in. The system uses a high-resolution (both horizontally and vertically) weather model with the best possible three-dimensional picture of the atmosphere for model initialization, and built-in high-resolution land use and topography datasets. This weather data application is configured to receive information initially from the Meteorological Assimilation Data Ingest System (MADIS) or the National Centers for Environmental Prediction (NCEP). Data is fed into the Weather Research and Forecasting Preprocessing System (WPS) and the Weather Research and Forecasting Data Assimilation (WRFDA) software running on the handheld device. The first weather source provides surface reports, upper-air reports, and large-scale initialization data which covers a large geographic area. The user can adjust the geographic scale for differently sized zones surrounding the device. The handheld device generates a forecast using information from the weather source and the geographic input. This is supplemented by sensor information on the device which detects the current air temperature, pressure, and relative humidity. Combined and processed, this information provides an enhanced analysis and forecast for the selected zone. The user can rapidly update the weather forecast for their location since the weather model is running on their phone. Feeding updated data into the model amounts to tapping for updated sensor data and retrieval of Internet-sourced weather data.
- Localized weather inclusive of real-time ground conditions
- This system is enabled by the cross-compiling of WPS, WRFDA, and WRF code on a Linux system and executed on the Android OS
- US application number 20170090069 available for license
- Potential for collaboration with Army researchers