Land subsidence using absolute and relative gravimetry: a case study in central Taiwan

Abstract

We experiment with absolute and relative gravimetry to determine land subsidence. A gravity network in the Yunlin County of central Taiwan is established to determine gravity variations that are largely due to land subsidence. Every 6 months, gravity values at two absolute gravity stations were measured by a FG5 gravimeter and those at 7 relative stations by a Graviton-EG and a Scintrex CG-5 relative gravimeter. A weighted constraint network adjustment was carried out by holding fixed gravity values at the absolute stations. Correction models for temporal gravity changes are developed and applied to raw gravity measurements. The adjusted gravity values oscillate, but in general increase with time, showing signature of land subsidence. An empirical gravity-to-height admittance factor is determined using gravity change (from FG5) and height change (from levelling) at the two absolute gravity stations. At most gravity stations, there is good agreement between subsidence rates from gravimetry and levelling. Some large discrepancies (> 1 cm/year) also exist and are caused by uncertainties in relative gravity accuracy, hydrological effects and the admittance factor. This study suggests that gravimetry has the potential to determine land subsidence to cm accuracy and is more efficient and economic than tools such as levelling and GPS.