呼和浩特地区降水识别的 Fisher判别研究

Fisher Discriminant Analysis of Precipitation Recognition in Hohhot

利用2018年呼和浩特白塔基地的微波辐射计数据,结合Fisher判别准则,在降水开始前120 min每隔10 min建立一个降水预报判别方程,希望能够找到基于微波辐射计数据的降水预报和人工增雨潜力识别的客观指标,之后利用2017年的数据对方程的预报准确性进行检验,结果表明:(1)降水发生前微波辐射计测得的积分水汽和积分液态水明显增加,对降水的发生有很好的指示意义,可以作为判别方程的前期因子。(2)该判别方法有较好的降水预报能力,且随着降水过程的临近,预报准确率逐渐升高,在降水开始前10 min和20 min,准确率可达到80%左右。

In order to find the objective indexes of precipitation forecast and artificial precipitation potential identification,by using the microwave radiometer data of Hohhot Baita base in 2018,precipitation prediction equation was establish every 10 minutes before precipitation with Fisher discriminant criterion.The prediction accuracy of the equation was tested by using the data of 2017.The results show that:(1)The integral water vapor and integral liquid water measured by microwave radiometer increase significantly before the precipitation,which has a good indication for the occurrence of precipitation,and can be used as the antecedent factor of the discriminant equation.(2)The Fisher discriminant method based on microwave radiometer data has a good ability to forecast precipitation.With the approaching of precipitation process,the accuracy rate gradually increases,which can reach about 80%in 10 to 20minutes before the beginning of precipitation.