Although fault locking state has been widely acquired with space geodetic observations, the mechanisms controlling fault locking are poorly known. Here, we infer the locking state of the Xiaojiang fault with a viscoelastic deformation model and integrated GNSS data, based on which we probe the mechanism that may control the locking pattern of the fault. Our results reveal four highly locked asperities along the Xiaojiang fault, which are separated by weak locking zones. Relying on the inferred locking degree and spring water temperature along the fault, we construct a model to connect the fault hydraulic conductivity with the cooling process of upwelling water. Model results suggest that high spring water temperature correlates with high hydraulic conductivity. Further, we use an experimental law to infer that the obtained high hydraulic conductivity may associate with high pore fluid pressure, which in turn weakens the fault frictional strength and leads to weak locking.