What can we obtain from the initial conditions of assimilating wind tower observation to improve near-surface meteorological elements?

Abstract

Accurate initial condition is prerequisite to an accurate prediction/forecast of near-surface meteorological elements and wind power. Although assimilating wind tower observations of wind farms in the initial condition has been proved feasible to improve the prediction, how and why assimilating wind tower observation improves initial condition are unclear. This study aims to improve the prediction by obtaining effective information from the initial conditions of assimilating wind tower observations. Results suggest that assimilating wind tower observations has local impacts on the initial condition, and the improvement of the initial condition is significantly related to the observed near-surface atmospheric stability. The forecast improvement by assimilating wind tower observations depends on two effective information: determining whether the initial condition is improved and the length of the improved initial condition can positively impact the forecast results. When the initial condition improves, forecast improvements can persist for about 12 h and are most distinct in the ultra-short term (0–4 h). Using the two effective information not only can distinctly improve the forecasts of near-surface wind, temperature, relative humidity, and wind power, but also can eliminate the stochastic forecasts. When the initial condition improves by assimilation, enhancing nudging intensity will have more positively impact on the prediction, and the root mean square error and mean absolute error in the ultra-short term can be reduced by 0.39 and 0.44 m s−1 for near-surface wind and 66.64 and 79.09 kW for wind power, respectively.