中文摘要：

介绍利用GPS和GRACE观测结果联合反演北美洲的地球模型参数，并检验现有的冰盖模型（ICE-5G and ICE-3G）的预测结果是否与卫星观测结果相符合。反演结果表明，现有的冰盖模型不能正确给出哈德逊海湾西部和东南部两个区域的重力变化中心，预测结果与GRACE卫星观测结果不符合。对ICE-5G冰盖模型的修正试验证实，在哈德逊海湾西部冰盖体积要减小～20％，而在东南部冰盖体积要增大～40％，这样才能使模型计算的垂直形变和重力变化与GPS和GRACE观测结果一致。基于三层地球模型的反演结果表明由卫星观测数据估计的岩石圈有效厚度（～150km，80％置信度）大于由相对海平面数据得到的估值（-90km）；上、下地幔的有效黏性系数分别为3．7×10^20 Pas和1．9×10^21Pas（90％置信度），比由相对海平面数据得到的相应估值小～20％。

英文摘要：

We present a joint inversion analysis of the GPS-derived radial （vertical） deformation and GRACE- derived gravity rates in North America to examine whether the ice-sheet models （ICE-5G and ICE-3G） can fit the satellite based observations. The results show that the existing ice-sheet models can not correctly predict the concentration of gravity change observed by GRACE in two areas of west and southeast of Hudson Bay. Tests of modifying ICE-5G show that a reduction of ice thickness by about 20% in the west area of Hudson Bay and an increase by about 40% in the southeast are required to fit both observed vertical deformation and gravity changes by GPS and GRACE. The results of joint inversion demonstrate that the estimated effective lithosphere thickness （150 km, un- der a statistical confidence level of 80% ） is apparently larger than that valuation （ chouq 90 km） inferred from rel- ative sea-level data, while the upper-and lower-mantle viscosity estimates （3.7 ×10^20 Pa s; 1.9×10^21 Pas; 90% confidence level） are about 20% smaller than those derived from sea-level indicators.