中文摘要：

利用湖北宜昌2007年观测的GPS对流层天顶延迟数据，对采用不同水汽反演方法计算的对流层可降水量PW的正确度和精密度进行对比分析。结果显示：不同天顶干延迟计算模型对GPSPW的精密度影响不大，但对其正确度有明显的影响，与探空PW相比，Hopfield模型计算的GPs册的平均偏差最小，Saastamoinen模型的平均偏差次之，而Black模型的平均偏差最大；大气加权平均温度对GPS PW的正确度有重要影响，对其进行本地化订正可以明显减小GPS PW与探空尸形的偏差，但对GPSPW的精密度影响不大；GPS PW与探空PW的相关性受大气水汽含量的影响，当大气水汽含量较低（PW≤65mm）时，两者的相关系数可达0．92，两者的平均偏差为3．8mm，偏差的均方差为6．4mm，而当大气水汽含量较大时，GPS PW与探空PW的偏差会增大，两者的相关系数会变小，这可能与GPS水汽反演方法有关；GPS PW比探空PW偏小，这可能是由两种探测方法的不同所造成的系统偏差。

英文摘要：

Based on the ground-based GPS zenith delay data observed at Yichang in Hubei Province in 2007, precipitable water （PW） is derived with various methods and error analysis on GPS PW is carried out. The results show that the correctness of GPS PW depends on dry zenith delay model, compared with PW obtained from radiosonde data （Sonde PW）, the bias of Hopfield model is the smallest while that of Black model is the biggest, and the bias of Saastamoinen model is between them; moreover, the correctness of GPS PW is also affected by atmospheric weighted mean temperature （Tm）, constructing a local model for Tm can significantly reduce the bias between GPS PW and Sonde PW; however, the precision of GPS PW is hardly affected by dry zenith delay model and Tm. Furthermore, the correlation between GPS PW and Sonde PW is affected by PW, in the case of small PW （PW≤65 mm）, the correlation coefficient between them reaches 0.92 with a bias of 3.8 mm and a mean-root-square error of 6.4 mm, but with the increasing PW, the correlation coefficient reduces while the bias and mean-root-square error increase. In addition, GPS PW is small than Sonde PW on the whole, which might be due to the system error of the two instruments.