A Multi-sensor Approach for Hydrological Monitoring of Wetlands: Altimetry-InSAR (ICESat-2/Sentinel-1) Integration Method Development over the South Florida Everglades

Wetlands deliver critical ecosystem services, including serving as habitats for diverse species (some of which are vulnerable or endangered), facilitating nutrient cycling, storing and sequestering carbon, and offering recreational opportunities. However, over the past century, wetlands have experienced significant loss, degradation, and stress due to anthropogenic influences such as water diversion, agricultural expansion, and urbanization, as well as natural processes like sea-level rise and climate change. This underscores the urgency to protect, conserve, and restore the remaining wetlands worldwide. A fundamental component of wetland conservation, management, and restoration is the monitoring of their hydrological systems, as wetland ecosystems are inherently dependent on water availability. Hydrological monitoring is commonly conducted using stage (water level) stations, which provide high temporal resolution data but suffer from limited spatial resolution, as these stations are often distributed several kilometers apart or more. Additionally, many wetlands remain ungauged or are monitored with only a sparse network of stage stations due to logistical constraints.

Space-based remote sensing technologies offer an effective alternative, providing high spatial resolution measurements of wetland water levels and their temporal changes. These techniques include Synthetic Aperture Radar (SAR), optical imagery, and radar and laser altimetry. SAR observations yield two independent observables—amplitude and phase—each sensitive to distinct hydrological parameters. Radar and laser altimetry missions deliver centimeter-level accuracy in water-level measurements along satellite tracks.

To overcome the limitations of individual monitoring methods, we developed a novel, space-based, multi-sensor approach to estimate absolute water level changes in wetlands by integrating ICESat-2 laser altimetry and Sentinel-1 InSAR data. This approach employs ICESat-2 absolute water levels to calibrate Sentinel-1 InSAR-derived relative water level changes, generating high spatial resolution (50–200 m) maps of absolute water level changes across entire wetland areas. We applied this methodology to the South Florida Everglades, a natural laboratory characterized by significant wetland variability and abundant ground-based hydrological data. The analysis utilized all ICESat-2 observations for the region, comprising 202 ground tracks collected between October 2018 and December 2023. Additionally, we processed 146 Sentinel-1 interferograms from the same period using the Alaska Satellite Facility's Hybrid Pluggable Processing Pipeline. This yielded 103 water level change maps with temporal intervals ranging from 12 to 364 days. Validation against gauge data revealed a root mean square error (RMSE) of 15.4 cm for the absolute water level change estimates. Error sources included uncertainties in ICESat-2 observations, InSAR measurements, and the EDEN interpolation scheme. To further investigate error contributions, residuals were decomposed into short- and long-wavelength components. Short-wavelength errors, primarily attributed to InSAR data, captured localized variations, while long-wavelength errors, associated with ICESat-2 data, reflected broad-scale biases. By removing long-wavelength biases, we achieved an RMSE of 7.8 cm, demonstrating the potential for high-accuracy wetland water level monitoring using the integrated multi-sensor approach.