Evaluation of Spectral/Grid Nudging Methods for Weather Analysis and Forecasting in Kyiv Region with the Use of WRF Mesoscale Meteorological Model

摘要

The possibility of using grid and spectral relaxation methods (nudging) and other options in the mesoscale model WRF for long-term continuous and forecast calculations on fine-resolution grid in the vicinity of Kyiv has been investigated. Results of comparison of selected meteorological parameters with surface measurements are presented. The basic recommendations for selecting the optimal combination of long-term calculation parameters are given. The use of the selected parameters allowed to obtain continuous meteorological fields over a long period (several months), which are well consistent with surface measurements. With grid and surface nudging and update of sea surface temperature for a temperature at a 2 m height in long-term continuous simulation a mean absolute error was MAE = 1.63 °C, and for wind speed at a 10 m height MAE = 1.11 m/s. Application of grid relaxation methods with grid surface nudging also allowed for improved results of calculations in forecast mode. The use of optimal combination allowed decrease of MAE for 2-m temperature from 2.06 °C to 1.85 °C for the first 48 h of forecast and from 2.35 °C to 2.14 °C for the next 49-96 h of forecast. The statistical measures of the quality of forecasted wind fields were also improved.