Item

Abstract

The assimilation of observations in limited area models (LAMs) allows to find the best possible estimate of a region’s meteorological state. Water vapor is a crucial constituent in terms of cloud and precipitation formation. Its highly variable nature in space and time is often insufficiently represented in models. This study investigates the improvement of simulated water vapor content within the Weather Research and Forecasting model (WRF) in every season by assimilating temperature, relative humidity, and surface pressure obtained from climate stations, as well as geodetically derived Zenith Total Delay (ZTD) and precipitable water vapor (PWV) data from global navigation satellite system (GNSS) ground stations. In four case studies we analyze the results of high-resolution convection-resolving WRF simulations (2.1 km) between 2016 and 2018 each in every season for a 650 × 670 km domain in the tri-border-area Germany, France and Switzerland. The impact of 3D VAR assimilation of different variables and combinations thereof, background error option, as well as the temporal and spatial resolution of assimilation is evaluated. Both column values and profiles derived from radiosondes are addressed. Best outcome was achieved when assimilating ZTD and synoptic data at an hourly resolution and a spatial thinning distance of 10 km. It is concluded that the careful selection of assimilation options can additionally improve simulation results in every season. Clear effects of assimilation on the water budgets can also be seen.