The Improvement Experiments On Near Surface Wind Forecasting With WEF-3DAR

Near-surface wind is the difficulty in the meteorological forecast and engineering design because of the strong instantaneous degeneration and intermittent, especially under the condition of the complex weather and terrain, which is affected by topography and underlying surface thermodynamic effects and complex turbulent flow process of PBL (Planet Boundary Layer). In this paper, a series of numerical forecast experiments, such as selection of the optimal PBL parameterization scheme, WRF-3DVAR assimilation and CALMET downscaling, have been taken to study the improvement way of near-surface wind forecast. The forecast effects of assimilating conventional data and ATOVS radiances on near-surface wind forecast are focused in this study to reveal the impact of ATOVS on near-surface wind forecast.Firstly, the simulation performances of near-surface wind with different PBL schemes in WRF model and the possible reason of near-surface wind simulation error are analyzed. The results show that, YSU scheme simulates the variation of near-surface wind in spring better than others over northeast China, semi-humid area; The stability conditions of near-surface have important influence on the vertical distribution of wind speed and turbulence, which affects the accuracy of the near-surface wind simulation; YSU scheme have large simulation error for the stability of actual near-surface, which results in large simulation error of near-surface wind, especially during the night and day-night alternation periods.Secondly, WRF-3DVAR system is used to investigate the impact of assimilating both conventional data and ATOVS radiances or assimilating conventional data only on near-surface wind. The results show that, the forecast performances of initial field and the forecast field in lower atmosphere are improved obviously after assimilating both conventional data and ATOVS radiances or assimilating conventional data only, which improve the forecast effects of near-surface wind, assimilation results capture the spatial distribution of wind speed in research region and periods; The detailed structure of local weather system is better reflected by WRF-3DVAR assimilating ATOVS radiances, which has the positive effects on near-surface wind forecast, however, relative to the results of assimilating conventional data only, the improvement effects of assimilating ATOVS are limited, assimilating conventional data is the effective way to improve the near-surface wind forecast.Finally, the forecast improvement of a strong wind process is analyzed with WRF-3DVAR system, and the small scale diagnosis model, CALMET, is coupled in WRF-3DVAR system to further discuss the downscaling improvement way of near-surface prediction. The results indicate that, CALMET provides more elaborate micro topography, which produce a positive effects on near-surface wind forecasting; Among the three numerical experiments, such as no assimilation, assimilation and CALMET downscaling, the results of forecasted strong wind with CALMET are the optimal; WRF-3DVAR system assimilating both conventional observations and ATOVS radiances improve the initial filed of WRF model, which provides a favorable thermal and dynamic structure for this strong wind process and contributes to the formation of the strong wind; Relative to the result of without any assimilation, the variations of near-surface wind are forecasted more objective and reasonable after assimilation, especially during the windy start stage and end stage.