Monitoring coastal inundation with Synthetic Aperture Radar satellite data

Abstract

Maps representing the presence and absence of surface inundation in the Louisiana coastal zone were created from available satellite scenes acquired by the Japanese Aerospace Exploration Agency's Advanced Land Observing Satellite and by the European Space Agency's Envisat from late 2006 through summer 2009. Detection of aboveground surface flooding relied on the well-documented and distinct signature of decreased backscatter in Synthetic Aperture Radar (SAR), which is indicative of inundated marsh in the Gulf of Mexico. Even though decreases in backscatter were distinctive, the multiplicity of possible interactions between changing flood depths and canopy height yielded complex SAR-based representations of the marshes. Validated by comparison to inland water levels, success of inundation mapping was primarily related to the operational frequencies of the SAR used to perform the mapping. Success of mapping was based on frequency of correspondence between satellite- and ground-based data. Overall, the most successful mapping (83 percent correspondence) was derived from Phased Array type L-band SAR (PALSAR), while mapping derived from C-band Advanced SAR (ASAR) was less successful (≤61 percent correspondence). Exceptions to the low performance of ASAR-based mapping (defined as >76 percent correspondence) occurred when water levels were well below or above ground, occurring over spatially extensive portions of the ASAR scene. When mapping day-to-day coastal inundation extents, results indicate that SAR systems operating at C-band frequencies are not as effective as those operating at L-band frequencies; however, multiple factors not related to frequency also reduced the effectiveness of C-Band in detecting subcanopy inundation. C-band has performed and continues to perform exceedingly well in applications for response to dramatic events and when strategic collections are available; however, L-band seems to be more suitable for day-to-day mapping of coastal inundation.