| dc.description |
Once considered the largest wetland in Central Asia, the Mesopotamian Marshlands of Southeastern Iraq have nearly disappeared. Various hydrological projects by the Iraqi government and dam construction in the region have nearly destroyed these once rich freshwater wetlands by over 90%. With the launching of Operation Iraqi Freedom recent attempts at restoration have been undertaken. To assess these changes and determine the success of restoration attempts this study is using 275-m multiangular MISR imagery to measure vegetation densities and water inundation patterns. By implementing a mapping method called spectral linear unmixing, fractional estimates of vegetation and water have been determined for six images from 2000-2005. Three subsets of the MISR imagery are compared: multispectral + multiangular, multispectral bands only, and multiangular bands only. Spectral linear unmixing, from all three subsets show a change in vegetation coverage from an average of 4490 km2 in 2000, to 9584 km2 in 2005 representing an increase of 114%. Results for water inundation also indicate an average increase of 109% from 3814 km2 in 2000 to 7986 km2 in 2005. To validate the MISR estimates, I compared these results with fractional estimates of vegetation and water derived from high spatial resolution 2-m QuickBird imagery based on data from 2002 and 2005. Each MISR image is rescaled to the 2-m spatial resolution of QuickBird so that a pixel-by-pixel comparison can be made. The validation results for 2002 and 2005 shows the multispectral + multiangular MISR approach identifies 28% of the water that QuickBird detects. Separately, multispectral and multiangular subsets were only able to detect 11% each. When examining vegetation coverage the multispectral subset is able to detect 69% whereas the multispectral + multiangular, and multiangular subsets detect 61% and 21%, respectively. Results presented in this paper are encouraging and suggest that the Mesopotamian Marshlands is recovering as vegetation is establishing and water is once again present. Interseasonal and interannual images may improve results and be more representative of the region. |
|