Signifying of the Urmia Lake Changes using Objected- Oriented Image Processing Techniques

The main aim of the present study was to map of Urmia Lake water surface diminishing rates over a long-term period, and demonstrating the most recent surface of emerged salty alluvial plains. For this purpose, Landsat TM, ETM+ and OLI imageries, observing from 1984 to 2015, were progressively processed to generate most of the thematic models in tempo-spatial context. All multi-date satellite data were subset and progressively preprocessed by ERDAS software based on the accurate topographical maps and the precise referenced Ground Control Points. Two classification projects were then implemented in this research based on preprocessed satellite imageries. Primary, by applying an ISOData function all Urmia Lake time series images have been classified to extract water surface classes. Subsequently, the 2015 year images have been segmented to derive remaining water surfaces and other landcover maps, particularly salty surfaces, by applying an object-oriented method in eCognition software. Revealed models demonstrate several long-term meaningful persuaded fluctuations and considerable periodical changes on the Urmia Lake water surfaces, mostly observed during the last decade. These great variations have occurred as the result of 15 meters decrease in height of water in the lake and about at least 70 percent deceasing on water capacity.