环渤海地区云量的精细化预报方法

A fine method forecasting cloud cover around the Bohai Sea

利用2010-2016年的云观测资料和T639数值预报产品,分析了云量与影响云生成的主要气象因子间的相关性,通过逐步回归方法建立了环渤海地区云量的时间精细化定量预报模型.结果表明,与总云量和低云量相关性显著的气象因子主要是水汽、大气不稳定度和上升运动.其中水汽类因子包括整层相对湿度、整层湿度和大气可降水量;大气不稳定度类因子为K指数和位势不稳定指标;上升运动为850~400 hPa 5层平均垂直速度.通过逐步回归方法建立了环渤海地区总云量和低云量的时间精细化预报模型,总云量和低云量预报方程的复相关系数大部分在0.5以上.回代预报检验结果表明,总云量和低云量绝对误差分别为22%和18%;试预报检验结果表明,总云量预报的绝对误差为23%;个例分析结果表明,新民站和沂源站的预报效果比较理想,大部分时段总云量的预报值及变化趋势与实际观测结果接近,模型预报性能较好,具有一定的实际应用价值.

Based on the cloud observation data,prediction products from the T639 model from 2010 to2016,the correlation between cloud cover and the main meteorological factors affecting cloud formation was analyzed,and time refined forecast models of cloud cover were established by using the multiple stepwise regression.The main results were as follows:water vapor,atmospheric instability and atmospheric vertical velocity were significantly correlated with the total cloud cover and low cloud cover al-most all year round.Humidity of the whole layer,atmospheric precipitable water and relative humidity of the whole layer,which were related to water vapor,were significantly correlated with the cloud cover.The average vertical velocity from 850 to 400 hPa mandatory level was significant,too.The K-index and geopotential instability index related to atmospheric instability was also significant.By the stepwise regression,time refined forecast models of total cloud cover and low cloud cover in the Bohai Sea region were established.The complex correlation coefficients of the total cloud cover and low cloud cover forecast equations were mostly above 0.5.The results of the back generation test showed that the average absolute error of the total cloud cover and low cloud cover was 22%and 18%.The forecast verification of the total cloud cover showed that the average absolute error of the model was 23%.Case studies showed that the forecast effect at the two representative stations of Xinmin and Yiyuan was ideal.The forecast values and trends of total cloud cover were close to the actual observations in most periods.The model has good forecast performance and some practical application value.