Validation of three land cover maps utilizing Space Shuttle photography and Landsat Thematic Mapper imagery

Astronaut acquired photography (AAP) can be used as a quantitative reference source in land cover map validation. Three land cover maps were selected for validation with AAP and Landsat Thematic Mapper (TM): (1) United States Geological Survey Level II, (2) International Geosphere Biosphere Programme DISCover and (3) Olson Global Ecosystem. The area validated extended from California to Louisiana, encompassing all of the southwestern United States (Figure 8). At one hundred and sixteen sample points randomly distributed across the southwestern United States, the dominant type of vegetation was established using Landsat Thematic Mapper data for comparison to astronaut acquired photography. Photointerpretation and a supervised classification algorithm were applied to assess the dominant type of vegetation at sample points savanna, evergreen needle leaf forest, shrubland, grassland, or a cropland/vegetation mosaic. The comparison of results from Thematic Mapper and astronaut acquired photography were accomplished via confusion matrices, a Khat and Z statistic. The results showed that there was almost no significant difference between utilizing Thematic Mapper data versus astronaut acquired photography as a source of reference data for map validation. The only significant difference occurred when using the classified Thematic Mapper imagery as a reference data source. When that reference source was used, this resulted in the largest number of misclassified land cover classes. Based on these results, several recommendations were made to streamline the efficiency of the Johnson Space Center photographic astronaut-training program. These recommendations include a web-based form to standardize and prioritize photographic requests, case of interpretation maps, and additional ancillary images hot-linked to the locational software used to identify areas of interest. These recommendations will not only increase efficiency but also improve the accuracy with which astronauts acquire photography in the future.