时空主成分分析(Spatial and Temporal Principal Component Analysis, ST-PCA)是一种将一组相互相关的变量转化为另一组新的不相关的变量的数学变换过程,在地学领域用于提取时空数据在时间和空间维度的动态演化特征。北京平原区地面沉降具有典型的时序演化和空间分布特征,因此本文在利用PS-InSAR技术获取到北京平原区2010-2016年长时序的地面沉降数据的基础上,采用时空主成分分析方法研究了该地区长时间序列中地面沉降的时空特征。研究发现时间主成分分析方法(TPCA)能够揭示地面沉降的趋势特征、波动特征,以及相应的空间分布模式。TPC1反映了研究区内的地面沉降趋势,呈现一种整体为空间分布不均匀沉降的特点；TPC2揭示了地面沉降的季节性波动特征,及其空间分布差异性,发现尤其在沉降速率大于30mm/a的沉降漏斗区呈明显的、南北分化的季节性差异空间分布特征,表现为北部地区夏季沉降量增大,南部地区冬季沉降量增大。空间主成分分析(SPCA)能够揭示地面沉降在时间上的演化规律,将相似时间变化趋势的沉降点聚类到同一个主成分。SPC1反映了大部分地区以沉降为主的特征,地面沉降呈持续的、线性下降的趋势；SPC2与SPC3则发现在轻微沉降区和非沉降区年均沉降量接近于0,但是季节性波动特征明显。通过ST-PCA的研究,本文提取了北京平原区地面沉降的一些时空规律,为北京地面沉降防控提供一定的数据支撑。

Spatial and Temporal Principal Component Analysis (ST-PCA) is a mathematical transformation process that converts a set of interrelated variables into a new set of unrelated variables. It is commonly used in the field of geosciences to extract the dynamic evolution characteristics of spatiotemporal data in time and space dimensions. In this study, based on long-term ground subsidence data obtained from PS-InSAR technology in the Beijing Plain area from 2010 to 2016, the ST-PCA method was used to investigate the spatiotemporal characteristics of ground subsidence in this region. The study revealed that Time Principal Component Analysis (TPCA) can reveal the trend and fluctuation characteristics of ground subsidence, as well as the corresponding spatial distribution patterns. TPC1 reflects the overall spatially uneven subsidence trend in the study area, while TPC2 reveals the seasonal fluctuation characteristics of ground subsidence and its spatial distribution differences. Particularly, significant seasonal differences in subsidence distribution were found in the subsidence funnel areas with a subsidence rate greater than 30mm/a, with increased subsidence in the northern region during summer and increased subsidence in the southern region during winter. Spatial Principal Component Analysis (SPCA) can reveal the temporal evolution patterns of ground subsidence by clustering subsidence points with similar temporal trends into the same principal component. SPC1 reflects the dominant feature of subsidence in most areas, with a continuous linear decrease in ground subsidence. SPC2 and SPC3 indicate that the average annual subsidence is close to zero in areas of slight subsidence and non-subsidence, but with significant seasonal fluctuation characteristics. Through the application of ST-PCA, this study extracted some spatiotemporal patterns of ground subsidence in the Beijing Plain area, providing data support for subsidence prevention and control in Beijing.