【摘要】 为了弥补川渝地区还没有本地化加权平均温度（T_m）模型的不足，提出一种川渝地区大气加权平均温度模型优化方法：选取甘孜、温江、西昌、沙坪坝4个站2017—2019年的探空数据，根据最小二乘法进行线性拟合，构建适用于川渝地区的加权平均温度模型。实验结果表明，本地化模型的平均绝对误差（MAE）和均方根误差（RMSE）均值分别为1.24和1.60 K，均优于贝维斯（Bevis）模型和中国加权平均温度回归模型；在本地化模型基础上加入季节性改正后的三角函数值、测站高程和大地纬度因素，新构建的本地化加权平均温度优化模型较本地化模型能取得更高的精度。提出的本地化加权平均温度模型在川渝地区可有较好的适用性，能够为全球导航卫星系统（GNSS）在当地气象行业的应用提供参考。更多还原

【Abstract】 In order to make up for the lack of localized weighted mean temperature(T_m) model in Sichuan-Chongqing region,the paper proposed an optimization method of weighted mean temperature model in Sichuan-Chongqing region: a T_m model suitable for Sichuan-Chongqing region was constructed with linear fitting according to the least square method based on the radiosonde data of four stations(Ganzi, Wenjiang, Xichang and Shapingba) from 2017 to 2019. Experimental results showed that the mean value of mean absolute error(MAE) and root mean square error(RMSE) of T_m(CY) model would be 1.24 and 1.60 K,respectively, better than that of T_m(Bevis) model and T_m(China) model; furthermore, by adding the seasonal corrected trigonometric function values, station elevation and geodetic latitude factors to the T_m(CY) model, the proposed constructed localized weighted average temperature optimization model could achieve higher accuracy than the localized model. In general,the proposed localized T_m model would have good applicability in Sichuan-Chongqing region, which could provide a reference for the application of local global navigation satellite system(GNSS) meteorology.更多还原