Evaluation of Wind Energy Simulated by Dynamical Downscaling Methods for Poyang Lake
动力降尺度法对鄱阳湖区风能资源模拟效果分析

A mesoscale weather numerical model, MM5 with high resolution model, Calmet, and the surface and upper air observation date were used for downscaling simulations for wind energy resources over Poyang Lake. Yearlong observation data from four representative wind farms around Poyang Lake were used to analyze the simulation results. We found that dynamic downscaling methods can simulate the spatial distribution of wind power over Poyang Lake, and the simulation results agreed with observed data. The high value area for wind power over Poyang Lake is on the south side of narrow waters and the shores on both sides. Dynamical downscaling methods were also used to simulate monthly changes in wind energy resources over Poyang Lake. Correlation coefficients between the simulation and observation results for monthly average wind speed were 0.773 to 0.897. The simulation and observation results of wind direction and wind power direction of every wind farm were consistent. For each wind farm in the Poyang lack area the dominant wind direction is a northerly wind, the direction of wind power is focused on the north side, and the direction of wind power is consistent. The dominant wind direction and distribution of wind energy direction between simulated and observed results have one to two range deviations. Deviation in the frequency of dominant wind direction is about 5%~15%, and deviation in the frequency of dominant wind power direction is 5%~20%. It is clear from our results that dynamical downscaling methods can be used to simulate the concentration area of wind speed, but the frequency results between simulation and observation do deviate. Correlation coefficients between the simulated and observed results for daily average wind speed are from 0.647 to 0.780, and the relative error between the simulation and observation of annual average wind speed over Poyang Lake is 6.77%.