A machine learning-based regression technique for prediction of tropospheric phase delay on large-scale Sentinel-1 InSAR time-series
A machine learning-based regression technique for prediction of tropospheric phase delay on large-scale Sentinel-1 InSAR time-series
复制标题
基于机器学习的回归技术，用于预测大规模 Sentinel-1 InSAR 时间序列上的对流层相位延迟
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
H. Nahavandchi
中科院分区：
文献类型：
--
作者：
R. Shamshiri;H. Nahavandchi
大气中温度、气压和相对湿度的时空变化是干涉合成孔径雷达数据误差的最大来源。对流层相位延迟的影响可以通过应用先进的多时相干涉合成孔径雷达（MTI）方法来部分减轻，该方法旨在使用SAR数据的堆叠来恢复速度场和位移时间序列。然而，将MTI方法应用于对流层校正的干涉图上，进一步提高了速度和位移时间序列的精度。干涉图中对流层延迟校正的一种方法是使用外部源，如ERA-Interim模型或全球导航卫星系统（GNSS）。然而，数据的插值是一个很大的挑战，因为我们需要找到一个合适的函数来预测整个干涉图的延迟，这对于大规模的Sentine-1干涉图来说是一个挑战。在这项研究中，我们提出了一种基于机器学习（ML）高斯过程（GP）回归方法的新技术，使用小基线干涉图和GNSS导出的天顶总延迟（ZTD）值的组合来减轻对流层相位延迟。该方法便于校正，因为我们不需要处理寻找用于低分辨率和/或稀疏分布的外部插值的合适函数。
Spatiotemporal variations in temperature, pressure, and relative humidity in the atmosphere produce the biggest source of error in InSAR data. The tropospheric phase delay effects can be partly mitigated by applying the advanced Multi-Temporal InSAR (MTI) methods aiming at retrieval of the velocity field and displacement time-series using a stack of SAR data. However, applying MTI methods on the tropospherically-corrected interferograms further improves the accuracy of velocity and displacement time-series. One way for the tropospheric delay correction in interferograms is using external sources such as ERA-Interim model or the Global Navigation Satellite System (GNSS). However, interpolation of the data is a big challenge, as we need to find a suitable function to predict the delay for the whole interferogram, which is challenging for large-scale Sentine-1 interferograms. In this study, we propose a new technique based on machine learning (ML) Gaussian processes (GP) regression approach using the combination of small-baseline interferograms and GNSS derived zenith total delay (ZTD) values to mitigate tropospheric phase delay. The method facilitates the corrections, as we do not need to deal with finding a suitable function for interpolation of low-resolution and/or sparsely distributed external