利用高原夏季温度(代用)、年太阳黑子相对数、西宁年平均气温,以及青藏铁路青海段夏季温度制作1990-2000年预测,得出最优子集预测的定性和定量评分高于周期外推,但低于均生函数。周期外推、均生函数、最优子集定性趋势预测的平均评分分别为60%~67%、80%和67.5%,而定量预测误差≤1.0℃的平均评分分别为40%~57%、77.5%~82.5%和70%~72.5%。

Using the summer temperature(substitute data) of Qinghai-Xizang Plateau,sunspot number,and annual average temperature at Xining,Qinghai,as well as summer temperature of Qinghai section along Qinghai-Xizang railway,we obtained the summer temperature prediction in 1991-2000 for that reagion.The qualitative and quantitative scores of the optimum subset is higher than that using periodic extrapolation,but lower than that using mean generating function.Averge score of periodic extrapolation,mean generating function,and optimum subset forecast is 60%~67%,80%,and 67.5%,respectively.For quantitative forecast with error ≤1.0℃ is 40%~57%,77.5%~82.5%,and 70%~72.5%,respectively.