The USGS, in cooperation with the U.S. Bureau of Land Management (BLM), created a series of geospatial products of the Scotts Creek Watershed in Lake County, California, using National Agriculture Imagery Program (NAIP) imagery from 2018, 2020 and 2022. The imagery was downloaded from United States Department of Agriculture (USDA) - Natural Resources Conservation Service (NRCS) Geospatial Data Gateway (https://datagateway.nrcs.usda.gov). The NAIP imagery from 2018, 2020 and 2022 was classified using Random Forest Modeling to produce land cover maps with three main classifications – bare, vegetation, and shadows. A total of 600 independent reference points were used in the accuracy assessment. The overall accuracy for all classes for each dataset is 98%. See attached ScottsCreek_20XX_AccuracyAssessment.csv files (contained within each LandCoverMap_associated_files_20XX.zip for each year respectively) for details. A preview image of the land cover map for 2018 is attached to this data release as an example (see LandCoverMap_RF_ScottsCreekWatershed_USGS2022_CC0.png).
The percentage of bare, vegetation and shadow pixels were calculated for the complete watershed and each individual NHDPlus2.1 catchment basins (slightly modified to support hydrological modeling). These metrics can be used to quantify bare and vegetated areas and detect and quantify vegetation changes over time. Users should be aware of the inherent errors in remote sensing products.