The numerical experiment results show that the mean forecast relative error is 1.73% even with a simple strategies library.
数值实验显示,虽然策略库比较简单,但其预测的平均相对误差仅为1.73%。
By selecting sub-ensemble with smaller error, the root mean square error of forecast is reduced by over 10%.
采用最优集合子集预报方式时的臭氧预报均方根误差比原确定性预报低了10%以上。
By selecting sub-ensemble with smaller error, the root mean square error of forecast is reduced by over 10%.
采用最优集合子集预报方式时的臭氧预报均方根误差比原确定性预报低了10%以上。
应用推荐