利用傅里叶函数构建全球气压气温改进模型

An improved-GPT2w model using Fourier function in Yangtze River Delta region

针对全球气压气温(GPT2w)模型各气象参数存在明显周期性误差的问题,该文以2015—2017年长三角地区的7个探空站资料作为参考,分析GPT2w模型各气象参数残差的季节性周期变化,并利用多阶傅里叶函数建立一种新的GPT2w改进模型。分析结果表明:与探空数据相比,GPT2w模型气压、温度、水汽压和加权平均温度的RMS均值分别为8.87 hPa、8.80 K、8.80 hPa和7.30 K;GPT2w改进模型气压、温度、水汽压和加权平均温度的RMS均值分别为3.65 hPa、2.76 K、2.47 hPa和2.48 K,相较于GPT2w模型总体精度有大幅改进,且与GPT3模型相比,其精度亦有明显改进。总体而言,利用傅里叶函数的GPT2w改进模型可在长三角地区获取精度优于GPT3模型的气象参数。

Aiming at the problem that the meteorological parameters obtained from Global Pressure and Temperature(GPT2w)model have obvious periodic errors,based on the reference data from 7 radiosondes in the Yangtze River Delta region during 2015—2017,the seasonal variations of parameter deviations of GPT2w model were analyzed in this paper.Then a new improved-GPT2w model was established using multi-order Fourier function.The results show that the average RMS of atmospheric pressure,temperature,vapor pressure and weighted mean temperature(Tm)of GPT2w model are 8.87 hPa,8.80 K,8.80 hPa and 7.30 K respectively compared to the radiosonde-derived data.The average RMS of atmospheric pressure,temperature,vapor pressure and Tmof the improved-GPT2w model based on Fourier function are 3.65 hPa,2.76 K,2.47 hPa and 2.48 K respectively.It means that the precision of improved-GPT2w model has been greatly improved compared to the GPT2w model and GPT3 model.In general,more precise meteorological parameters can be obtained by the improved-GPT2w model,which is even better than GPT3 model in the Yangtze River Delta region.