Flood monitoring with SAR in Rupununi Savannah, Guyana

Despite their high environmental, social and economic value, wetland habitats show the highest rate of land degradation on a global scale. Road construction is among the main drivers for tropical wetland degradation. Cost-effective methods for continuous monitor- ing and evaluation are needed to provide quantitative data for conservation efforts. In our study we used freely available Earth Observation time series to evaluate impacts from a new road construction project on the hydrology of Rupununi wetlands in Guyana. We used the Cumulative Sum (CUSUM) statistical method on Sentinel-1 GRD satellite im- ages in Google Earth Engine platform to compare changes of flood occurrence and flood extent two years before/after road construction. We found two locations that experience notable changes in flood dynamics and illustrate blockages of flood flow after road con- struction. One of the locations had 55% more new flooded areas on one side of the road in the first year after road construction and 49% in the second. The other location showed a corresponding increase of flood occurrence on one side (31% and 30%) with decrease on the other (42% and 38%) for the first and second year respectively. Both locations coincide with locations hypothesized by previous field research, as potential barriers of flow. Our findings reinforce the need for construction of culverts to allow the natural flow of creeks, crucial for fish migrations. The main limitation faced in our study was the lack of previously produced hydrological model data, showing creeks and rivers, flow direction and connectivity of separate wetland patches. Overall, the use of SAR remote sensing and CUSUM analyses, demonstrated in our study, has the potential to fill the gap in tropical wetlands monitoring, as it is a continuous, distant, and low-cost monitoring method.