Leveraging the Global Fiducial Library (GFL) to Enhance Dynamic Surface Water Extent (DSWE) Product Evaluation.

The Global Fiducials Program (GFP) was established to facilitate monitoring and scientific understanding of earth processes and environmental issues around the world. Scientists from a broad range of disciplines nominate sites of scientific and environmental importance. And imagery for selected sites are processed and made openly available. The goal for Dynamic Surface Water Extent or `DSWE', is the generation and distribution of long-term records of inundation of the land surface at the highest spatial and temporal frequencies possible from moderate-resolution environmental monitoring satellites such as Landsat and Sentinel-2. Starting with Landsat Thematic Mapper class instruments, openly available DSWE data span more than a 35-year period for some locations. Such systematically collected and openly available value-added data allow a broad range of scientific and resource management analyses from local to global scale. To promote best use of these data however, an understanding of their uncertainty is required. While commercial satellite imagery can be used to assess DSWE uncertainty for relatively recent, individual periods of time, the imagery from the Global Fiducials Library (GFL) provide low-cost opportunities to assess DSWE accuracy across a variety of environments with greater length of record, that is, across multiple Landsat platforms. We used automated processing of GFL imagery to generate DSWE evaluation data and compared those results to DSWE output for a range of physiographic and land cover settings sampled by the GFP. Because both datasets are openly available at no cost to the user, these analyses of environmental data records and the scientific application of time-series remote sensed imagery can be replicated, modified, and/or extended by anyone with access to the internet and geospatial processing resources.