中国区域大气加权平均温度的时空变化及模型

大气加权平均温度T_m是地基GPS水汽遥感的关键参数，决定了水汽反演的精度。利用2008—2011年全国123个探空站点资料，分析了T_m与其影响要素纬度、海拔、地面气温、水汽压及大气压之间的关系，结果表明:T_m与纬度和海拔随季节变化呈周期性负相关，与地面温度和水汽压的自然对数呈正相关，与大气压呈负相关；T_m的空间变化具有纬度地带性和明显的气候分布特征，其变异程度在空间分布上有显著差别，不同地理位置的T_m受季节性影响程度不一，T_m也具有明显的年际周期性变化，年内T_m的每日变化符合二次函数分布规律。按照全国、气候分区和季节分区方法，分别建立了T_m单因子和多因子回归模型，并利用2012年1—5月数据对所建模型进行验证，T_m的估算误差能满足GPS水汽遥感2%的精度，模型普遍适用于我国地基GPS水汽遥感中T_m的估算。

The Spatial and Temporal Variations of Weighted Mean Atmospheric Temperature and Its Models in China

Atmospheric water vapor plays an important role in the high-energy thermodynamics of the atmosphere and the generation of storm systems. Water vapor remote sensing can provide a detailed primary parameter within meteorological prediction and climate models. Ground-based GPS can obtain continuously precipitable water vapor with high spatial and temporal resolution since regional GPS networks are established widely all over the globe. Weighted mean atmospheric temperature (T_m) is a key parameter in retrieval of atmospheric precipitable water vapor from ground-based GPS measurements, and the precision of P_W retrieved by ground-based GPS is proportional to the accuracy of estimated T_m. Using radiosonde data of 123 stations in China from 2008 to 2011, the relationship of T_m is analyzed with its affecting factors, latitude, altitude, surface temperature (T_s), partial pressure of water vapor (e) and atmospheric pressure (P₀). Results show that T_m has a negative periodic correlativity with latitude and altitude as season changes, it has a good positive correlativity with surface temperature and partial pressure of water vapor transformed by natural logarithm, and it also has a negative one with atmospheric pressure. Furthermore, the spatial and temporal variations of T_m is discovered. The spatial variation of T_m displays the distinct latitudinal and climatic features and its spatial heterogeneity is evidently different in different regions, so T_m in different latitudinal zone is dissimilarly affected by seasonal climate change. For the temporal variation, T_m displays the prominent inter-annual variation and its diurnal variation is accord with the quadratic function. Given that, the regression models of T_m based on single factor T_s multi-factor T_s, e and P₀ are deduced respectively corresponding to all areas in China, climatology divisions and seasonal divisions, and then these models are validated by radiosonde data from January to May in 2012. The results show that estimated T_m would achieve an uncertainty of 2% for precipitable water vaper retrieved from GPS measurements, and these models are suitable to estimate T_m for the retrieval from ground-based GPS measurements in China.