资源环境科技发展态势分析平台

The present study assesses the added value of high-resolution maps of precipitable water vapor, computed from synthetic aperture radar interferograms , in short-range atmospheric predictability. A large set of images, in different weather conditions, produced by Sentinel-1A in a very well monitored region near the Appalachian Mountains, are assimilated by the Weather Research and Forecast (WRF) model. Results covering more than 2 years of operation indicate a consistent improvement of the water vapor predictability up to a range comparable with the transit time of the air mass in the synthetic aperture radar interferograms footprint, an overall improvement in the forecast of different precipitation events, and better representation of the spatial distribution of precipitation. This result highlights the significant potential for increasing short-range atmospheric predictability from improved high-resolution precipitable water vapor initial data, which can be obtained from new high-resolution all-weather microwave sensors.

Plain Language Summary Weather forecasts will never be perfect because our models are simplified representations of nature and our observations of the atmosphere are inaccurate. In this study we show, nevertheless, that it is possible to improve such forecasts by interpreting the atmospheric signals in spaceborne radar observations of the Earth surface, indicative of the distribution of water vapor. Better and more detailed maps of water vapor are found to lead to better forecasts not just of water vapor but also of precipitation. A two and a half years assessment covering a wide range of weather conditions in a very well monitored region near the Appalachian Mountains, USA, suggests that the proposed methodology has a significant impact in the quality of the forecasts and could easily be implemented.