Development of a Multi-Index Method Based on Landsat Reflectance Data to Map Open Water in a Complex Environment

Mapping surface water extent is important for managing supply agriculture and the environment. Remote sensing technologies, such as Landsat, provide an affordable means of capturing with reasonable spatial temporal coverage suited to this purpose. Many methods are available mapping including modified Normalised Difference Water Index (mNDWI), Fisher’s index (FWI), Observations from Space (WOfS), Tasseled Cap Wetness (TCW). While these can discriminate water, they have their strengths weaknesses, perform at best in different environments, threshold values. This study combines indices by developing rules that applies environment where best. It compares across Murray-Darling Basin (MDB) southeast Australia, assess performance compile a heuristic rule set accurate application MDB. The results found all single well Kappa statistic showing strong agreement, ranging 0.78 WOfS 0.84 TCW (with −0.035), improvement overall output when was selected. mNDWI (using −0.3) works within river channels, while −0.035) wetlands flooded vegetation. FWI 0.63 0, respectively) work remaining areas. Selecting appropriate increases 0.88 pixel accuracy 90.5% dry 94.8%. An independent assessment illustrates benefit using multi-index approach, making it suitable regional-scale multi-temporal analysis.