Detection Threshold Estimates for InSAR Time Series: A Simulation of Tropospheric Delay Approach

Abstract

We present a method for estimating the detection threshold of InSAR time-series products that relies on simulations of both vertical stratification and turbulence mixing components of tropospheric delay. Our simulations take into account case-specific parameters, such as topography and wet delay. We generate the time series of simulated data with given intervals (e.g., 12 and 35 days) for temporal coverages varying between 3 and 10 years. Each simulated acquisition presents the apparent noise due to tropospheric delay, which is constrained by case-specific parameters. As the calculation parameters are randomized, we carry out a large number of simulations and analyze the results statistically and we see that, as temporal coverage increases, the amount of propagated error decreases, presenting an inverse correlation. We validate our method by comparing our results with ERS and Envisat results over Socorro Magma Body, New Mexico. Our case study results indicate that Sentinel-1 can achieve ≈1 mm/yr detection level with regularly sampled data sets that have temporal coverage longer than 5 years.