MODELING SUSTAINABILITY OF THE URMIA LAKE BASIN USING DYNAMIC BAYESIAN NETWORK

Lake Urmia in north-western of Iran is one of the largest permanent hyper-saline lakes in the world and the largest lake in the Middle East. During the past decade, the lake has shrunk significantly and the water level has fallen dramatically. Problems of Urmia Lake has been mainly caused by human intervention in various forms from one hand and natural elements such global warming and drought from another hand. Lake Urmia is unbalanced and unsustainable development outcome of current conditions in the river basin management and uncontrolled exploitation of renewable water resources of the basin. Sustainability is defined as keeping the lake's water level at its ecological level (around 1274 m above sea level). At this level the lake can continue its normal ecological performance. Several factors influence the entering water to the lake including rainfall as well as outgoing water from the lake including evaporation and agriculture water. The research presents a model using Dynamic Bayesian Network to assess the sustainability of the Lake. At first a baseline model is designed to take into account the factors influencing water volume for a given year. The model aims to capture and quantify affecting factors that influence the water level in each year. Then a dynamic model is built which can analyze the water level in a number of consecutive years. In order to verify the model, its results are compared with available data. Then sensitivity analysis is performed. The model can perform different type of scenario analysis by defining the value of some variables and updating the probabilistic value of other variables. Three scenarios are defined and explained for illustrations. The methodology offered in this paper enables policy makers and decision makers to investigate and analyze their policies and actions in order to revive the lake