Integrating InSAR coherence and SAR amplitude to unravel the surface change processes operating during extreme rainstorm events in the Atacama Desert.

Devoting more efforts to understand how arid landscapes respond to extreme rainfall events, given the expected increase in storm frequency in the future due to global warming projections, is of great relevance and therefore needs to be addressed. While local studies of recent storm impacts in drylands have proven to be useful, our understanding of global impacts at local-and-regional-scales over longer time-scales is now more qualitative than quantitative.Deciphering the effects of erosion runoff processes operating during extreme rainstorm events requires developing practical measuring approaches that assist understanding the temporal and spatial extent of erosion and sediment pathways in the ephemeral drainage networks of bare lands. The advent of Synthetic Aperture Radar (SAR) satellite missions with, for example, the Sentinel 1 constellation from the ESA, has provided a great number of images that can be used to map the areal and temporal extent of erosion during rainstorm events. As a result, we are now able to unravel surface runoff erosion operating in arid areas using InSAR coherence change detection following, for example, the work of Cabré et al. (2020, 2023). Interferometric SAR (InSAR) coherence can be used to decipher the sediment entrainment areas and identify channels and drainages disturbed by the passage of floods. However, the coherence remains a dimensionless parameter with no physical meaning of surface change. Thus, it cannot be used yet to estimate surface change processes in an automatic basis. For this reason, we have explored the areas with surface change identified in InSAR coherence images using SAR amplitude and field calibration data. In the identified surface change areas we have performed grain-size measurements to prove that sediment grain-size diameter (e.g., D84, D50) in ephemeral channels is well correlated (R=0.93 and 0.72, respectively) with SAR amplitude values and therefore can be used to (i) unravel the downstream variations in grain-size by providing valley-floor grain-size maps and, (ii) identify fluvial features (e.g., longitudinal bars) preserved within the ephemeral channels after the passage of a flood. The latter can be of wide application to monitor ungauged ephemeral channels in arid areas worldwide and provide insights about the dryland sedimentary system dynamics during extreme storm events.