零漂改正对中国地壳运动观测网络重力数据处理的影响

基于中国地壳运动观测网络2000年流动重力观测数据，分析18台LCR-G型相对重力仪的零漂特性。结果显示，重力仪零漂率在某些时段变化较大，最大达10 μGal/h以上。根据重力仪零漂改正的不同处理方法，利用传统平差模型与改进模型对观测数据进行平差计算，并对结果进行比较。在相同先验中误差的情况下，传统平差模型的后验中误差为39 μGal，重力值平均精度为26.5 μGal，而改进模型的后验中误差为16.5 μGal，重力值平均精度为11.2 μGal，远高于传统平差模型。在数据预处理阶段按闭合观测分段进行零漂改正，可以有效消除重力仪零漂剧烈变化的影响，明显优于将仪器零漂率当作固定参数的计算方法。

Impacts of Zero Drift Correction in CMONOC Data Processing

Based on CMONOC mobile gravity data from 2000, zero drift characteristics of 18 LCR-G relative gravimeters are studied. The results show that zero drift rate of gravimeter may sometime shave a significant change, up to 10 μGal/h. Based on different processing methods of zero drift correction, the results of traditional and improved adjustment models are compared. Given the same priori error, the posterior error of the traditional adjustment model is 39 μGal, and its average accuracy of gravity value is 26.5 μGal; however, for the improved adjustment model, its posterior error is 16.5 μGal and the average accuracy of gravity value is 11.2 μGal. The accuracy of the improved model is much higher than that of the traditional adjustment model. The comparison shows that computing the zero drift correction by circuit in the data preprocessing can effectively eliminate the error of significant change of zero drift. The improved model is superior to traditional adjustment model that assumes the zero drift as a constant.