LANDFIRE (LF) creates image composites to better map disturbance and vegetation. Improving image composite methods is ongoing and investigations into cloud processing of Landsat data is in progress.
Image compositing to create a clear satellite image without clouds or missing data is an important first step to mapping vegetation and detecting change for annual disturbance. Compositing began with the best pixel method, but during LF 2016 Remap transitioned to the percentile composite method. The percentile composite method uses a range of Landsat scenes (dates) for all bands to calculate the 50th percentile of each band’s pixel values. This method reduces the no data values in the final image and provides a representative pixel value for the date range of interest. Currently, LF scientists are exploring accessing and processing Landsat image composites in the cloud. This method would use distributed cluster processing to avoid downloading Landsat scenes and would potentially speed up compositing.
Last updated: January 2023