【摘要】 同区域不同期SAR数据时间的大气延迟差异是影响SAR干涉测量的最主要因素之一。以北京主城区作为研究区域,综合地形和气象要素,将BP神经网络技术用于GPS对流层延迟内插,并用于InSAR大气延迟改正研究。实验证明,经与GPS沉降量相比较,该方法优于克里金插值法,大气延迟双差改正结果用于InSAR大气效应改善,可有效提高InSAR处理精度。更多还原

【Abstract】 Atmospheric delay differences of SAR time data of one region in different time periods influences the interference measurement of SAR majorly.The main urban area in Beijing is selected as the study area,combining with terrain and meteorological element,and BP neural network technology is used for GPS tropospheric delay interpolation and InSAR tropospheric delay correction study.It shows that the predicted results of terrain and meteorological elements are better than Kriging interpolation method which the interpolation accuracy is suitable for tropospheric delay correction.The double difference correction of tropospheric delay is used to remove InSAR atmospheric effect.Comparison between InSAR deformation and GPS deformation shows that GPS atmosphreic delay can improve the accuracy of InSAR.更多还原