Abstract: For the problem of inaccurate analysis of land subsidence caused by single monitoring method, the “air-surface-underground” monitoring system of land subsidence is adopted with platform and parameter collaboration in this paper. Then a fusion model for aerial and surface measurements is constructed and solved with the least squares method, to obtain high-precision subsidence data. Finally, these methods of visualization and spatiotemporal statistics are used to integrate air/surface and underground data to obtain the causes of subsidence. These techniques are applied to analyze the local area in Tianjin, and a three-dimensional monitoring system for land subsidence is constructed, combing InSAR, leveling, and layered markers. These indexes of maximum error and mean square error are used to elevate the fusion result, with 20% increase of mean square error. The conclusions that groundwater exploitation is the main cause of severe subsidence from both qualitative and quantitative analysis. It is suggested that the method proposed in this article are effective in improving the monitoring reliability of land subsidence, achieving accurate subsidence causes. The results are expected to provide a valuable reference for comprehensive prevention and control of land subsidence.

Key words: land subsidence, three-dimensional monitoring, data fusion, layered label monitoring, subsidence induced factors