Forecasting Surface Area Fluctuations of Urmia Lake by Image Processing Technique

Urmia Lake’s water surface area is among the most important parameters needed for water balance analysis. Periodical measurement of this parameter directly by conventional topography almost seems impossible since it is costly and time – consuming. Such limitations highlight the needs for new approaches to be taken, namely remote sensing technique which could provide a good approximation of lake’s surface area in terms of some other parameters available or at least easily measured. This paper considers development of a new model for lake’s surface area measurement according to available water levels and its comparison with other methods in this field as well as the calculations regarding salt-land formation and coastline changes. High resolution images provided by NASA satellites, Aqua and Terra were collected and passed an image processing stage through MATLAB software for surface area calculations. Finally, the water level and surface area values resulted from the home made code, were put together to reach relationship. The comparison between the results of proposed method and provided data by Eastern Azerbaijan Water Organization and also a similar study indicated that the proposed image processing technique has good performance to estimate the surface area of Urmia Lake. The maximum error between the results of proposed model and a similar study which was used combination method of Cellular Automata and Markov Chain was 5.96 % which indicates the good performance of image processing technique in estimation of surface area of Urmia Lake.