Impact of Iranian permanent GPS network precipitable water estimates on numerical weather prediction

The aim of this study is to assess the impact of continuous and precise ground-based GPS water vapor estimates as a by-product of Iranian Permanent GPS Network (IPGN) geodetic data processing, together with conventional surface and upper air meteorological data on the short range prediction of rainfall and surface moisture fields, including 2 m relative humidity and Precipitable Water Vapor (PWV) over north of Iran. The Weather Research and Forecasting (WRF) model and its Four-Dimentional Variational Data Assimilation (4DVAR) system is used to determine the impact of data assimilation on simulation of three heavy rainfall cases that occurred over the northern part of Iran. All three rainfall cases considered in this study are associated with a shallow and cold high pressure located over Russia that extends towards the southern Caspian Sea. The results of numerical experiments showed that the assimilation of ground-based GPS PWV data, on average, improves simulation of precipitation, PWV and near surface relative humidity, even though the skill declines after 24-h simulation. It is found that inclusion of GPS PWV improved the predicted accumulated precipitation in day-1 of the model simulations for February and November cases up to 7 percent while there was almost no positive impact in September case. Results suggest that incorporation of observations in initial conditions of the WRF gives generally a slight improvement in 2 m relative humidity forecasts when compared with the control experiment without assimilation. Assimilation of GPS PWV in February and September cases reduces, on average, 0.8 mm the Mean Absolute Error (MAE) of the PWV model during 12-h forecast period. Overall, best results in terms of MAEs were achieved when GPS water vapor estimations were used along with conventional surface and upper air radiosonde data.