Land Use and Cover Maps from Landsat, Mawas, Central Kalimantan, Indonesia, 1994-2019

There are six files, each representing a five-year interval over the period 1994-2019. An additional file for 2015 was created for accuracy assessment.

A high-quality and low-cloud coverage image from Landsat 5 or Landsat 8 over each 5-year period was selected or composited for the January-August timeframe. Investigators used their knowledge to manually identify training polygons in these images for five LUC classes: peat swamp forest, tall shrubs/ secondary forest, low shrubs/ferns/grass, urban/bare land/open flooded areas, and river. Pixel values of Landsat Tier 1 surface reflectance products and selected indices were extracted for each LUC and used to predict LUC classes across the Mawas study area using the Classification and Regression Trees (CART) method.