利用3个海平面变化冰川均衡调整(GIA)模型———RF3L20、Pau-5-Ra和Pel-5-VM4-R,分别对利用GRACE数据和联合卫星测高及海水温盐度数据解算2004~2010年海水质量变化趋势进行GIA改正。比较改正后的结果得到,用Pel-5-VM4-R模型改正后,以上两种方法得到的海水质量变化趋势较为一致,而用另外两个GIA模型改正后,两种方法得到的结果相差较大。本研究显示,在利用GRACE资料反演海水质量变化趋势时,用Pel-5-VM4-R模型进行GIA改正相对合适,并且在2004~2010年,海水质量变化对全球平均海平面变化贡献为1.5~2.0mm/a。 Based on the GRACE data and the combined satellite altimetry and seawater temperature and salinity data, three GIA models (RF3L20, Pau-5-Ra and Pel-5-VM4-R) GIA corrections for the trend of seawater quality from 2004 to 2010. Compared with the corrected result, the change tendency of seawater quality obtained by the above two methods is consistent after corrected by Pel-5-VM4-R model. When the other two GIA models are corrected, the difference between the two methods is more Big. The present study shows that the GIA correction using the Pel-5-VM4-R model is relatively suitable when using the GRACE data to reverse the trend of seawater quality, and the contribution of seawater quality changes to the global mean sea level change from 2004 to 2010 was 1.5 ~ 2.0mm / a.