Mapping the temporal and spatial changes in crescent dunes using an interferometric synthetic aperture radar temporal decorrelation model

Abstract The interferometric synthetic aperture radar temporal decorrelation model is a powerful tool for describing surface changes over time. Recently, this model has been used in several dune stability studies. However, so far, the combined effects of spatial and thermal decorrelation on the estimation of temporal decorrelation in dune regions have not been studied. In this paper, we propose a temporal decorrelation model for a dune area based on the consideration of the effects of spatial and thermal decorrelation, and we validate the model using the sand drift potential (DP) and precipitation, by which the potential of this model to describe surface changes in crescent dune area is evaluated. Furthermore, we test the proposed model by mapping changes of stability in crescent dunes on both sides of road in the Taklamakan Desert between May 22, 2017, and May 29, 2018. The results show that, first, spatial decorrelation in the dune area can be ignored, but thermal decorrelation cannot. Second, the temporal decorrelation gradually decreases with increasing synthetic sand drift potential (RDP, a vector synthesis of DP), but its sensitivity to different levels of RDP is different. When the RDP is at a relatively low level (between 0 and 50), even small changes in it can be identified. However, when the RDP is at a relatively high level (greater than 50), even relatively large changes in it are difficult to distinguish. Third, when the RDP is low, this model can clearly distinguish the intrusion of dunes of different scales into a highway protection system.