Iraqi marshes were considered the largest wetland in the Middle East and characterized by varied environment (such as river, lakes and vegetation covers). During the last decades the Iraqi marshes subjected to many artificial and natural changes. These changes have impacts on its ecosystem (land cover, vegetation and aquatic environment). The aim of this study is detect environmental changes in the Iraqi marsh for the period 1973-2004, by using multi-temporal and multispectral images with digital image techniques. These techniques include digital mosaic, geometric correction, image enhancement, image classification; supervised (by using maximum likelihood method) and unsupervised (by using Isodata method), Normalized Difference Vegetation Index (NDVI) for the images (1973, 1990, 2000, 2003, 2004) and computed surface radiant temperature for the images (1990, 2000, 2003, 2004). Series of Landsat images shows that the Iraqi marshes are desiccated vegetation cover and water decrease during the period 1990- 2003. While 2003 shows reflooding the area. The results show that NDVI is good indicator for vegetation area and surface temperature. There is negative correlation between NDVI and surface temperature.