HYDROLOGICAL SIMULATION OF A SMALL URBAN UNGAUGED CATCHMENT: A CASE STUDY OF SMART CITY NAVI MUMBAI

Abstract

Land use/ land cover (LULC) is a dynamic and continuous spatio-temporal phenomenon, especially in all the developing countries, showing no sign of stopping, reason being the economic developments and population increase. This study presents a simulation model using Multi-Layer Perceptron integrated with Markov chain, Remote Sensing (RS) and Geographic Information (GIS) for the city of Navi Mumbai, India. Artificial neural networks were used to train on the predictor variables of urban expansion. Navi Mumbai is facing on-going challenges concerning its urban sprawl, due to rapid population increase. In this research we assess the past urban land use transitions in Navi Mumbai between 2000 and 2016, as the major development in Navi Mumbai started after 1992. This is done by a combination of multi-criteria evaluation processes originating transition probabilities that allow a better understanding of the regions urban future by 2030 and 2050, while the transition probabilities are incorporated from the Markov chain model. By the year 2030 about 48% and by 2050, 57% of Navi Mumbai will be urbanized. Major contribution to the urban class is coming from tree cover and barren class. Very less advancement is shown wetland and forest areas, which shows the predicted results are ensuring sustainability. The accuracy rate is 99.98% and calculated with a skill measure of 0.9997. It indicates that model has been trained very efficiently and can be used for understanding the spatio-temporal land use change dynamics of Navi Mumbai. The structure of the model allows for urban growth simulation and it therefore carries scope of being used to anticipate growth for other cities and help government agencies to better understand the consequences of their decisions on urban growth and development. Soil and Water Assessment Tool (SWAT) is a physically based semi - distributed model is implemented to understand the behaviour of urban ungauged catchment of Navi Mumbai, India. Prediction in an ungauged basin (PUBs) has always been a challenging task due to the inability to perform calibration and validation as there is no observed gauge-discharge data. To address this problem well known catchment characteristics similarity technique is adopted. SUFI-2 algorithm is used for calibration (2003-08) and validation (2009-13) of the surface runoff performed on a monthly basis. For the monthly time step the NSE values were 0.76 and 0.86, R2 values were 0.79 and 0.85, PBIAS values were +3.25% and -5.45% RSR values were 0.006 for calibration and validation period respectively. Hence this model can be successfully used for hydrological modelling of an ungauged catchment of Navi Mumbai.