Creating an historical land cover data set for the Wimmera region, Victoria, Australia from the USGS Landsat archive

The availability of the USGS Landsat archive, which spans nearly 40 years, has presented many opportunities for land cover mapping [1]. This study aimed to develop an approach using the USGS Landsat archive to produce land cover maps which could be used to assess changes in land cover over time. The study area was located in the Wimmera region of north-western Victoria, Australia. Initially, Normalized Vegetation Difference Index (NDVI) was used to generate coarse land cover maps of water, bare ground and low and high vegetation cover. These classes were refined using a Multiple Endmember Spectral Mixing Analysis (MESMA) approach. The production of this baseline data now presents opportunities for further investigation of land management, land cover changes and trends, and links to other biophysical data such as soil characteristics.