Abstract:

Based on the Weather Research Forecast (WRF) Model and its 3D-Var (three-dimensional variational assimilation) system, the numerical prediction experiments of water vapor data from aircraft, automatic station, conventional sounding and GPS during a short-time heavy rainfall in Jilin province on July 13, 2017, were carried out to study the influence of variant conventional observation assimilation on the initial field and forecast field of the model.The results show that the assimilation of conventional observation can improve the initial humidity field and temperature field, increase the humidity in the central part of Jilin province at the initial moment, and strengthen the temperature difference between the north and south sides of the real rain belt, and better predict the evolution process of the physical quantity.The influence of different conventional observation data assimilation on the prediction of humidity field, vertical velocity, wind field, and convective effective potential energy appears discrepantly.Assimilation of aircraft data increases regional overlaps between the strong vertical movement and the moisture, relative humidity, and wind shear in the precipitation area, being conducive to the intersection and transport of warm and cold airmass.Assimilation of automatic station data can increase the vapor in the precipitation center and southeast Jilin province, while the temperature field is mainly weakened.The assimilation of GPS vapor data has the most significant influence on the central wind shear forecast of precipitation, but has a less significant effect on the temperature field adjustment.The prediction of vertical velocity and wind field shows deviations due to the assimilation of radiosonde data.The TS scores, missing report rates, and empty report rates of 24-hour accumulated precipitation are better than those of unassimilated precipitation, and the TS scores of light rain, moderate rain, and heavy rainstorm are significantly improved.In general, the forecast of precipitation from the assimilated aircraft data is the closest to the real situation.Higher TS scores appear in the prediction of heavy and the above rain, especially in the area and intensity of 100-milli meter precipitation.

Key words: WRF-3VDAR, Data assimilation, Short-time heavy precipitation, Numerical weather forecasting