Dynamic Connectivity in the Middle Ganga Reaches Impacted by Human Disturbance

Thursday, 18 December 2014
Haridas Mohanta1, Patrice Carbonneau2 and Rajiv Sinha1, (1)Indian Institute of Technology Kanpur, Kanpur, India, (2)University of Durham, Durham, United Kingdom
River systems are often perturbed by human impacts that can be difficult to quantify. In the middle Ganga, Lav Kush barrage at Kanpur was constructed in 2004 for augmenting drinking water supply for Kanpur city. Following this perturbation, the main flow of the river was diverted to the right bank downstream of this barrage. The left bank of the river is now almost dry in the pre-monsoon period. This complex change of flow pattern is difficult to quantify and characterize. However, the resulting changes in flow pathways do broadly fall under the concepts of connectivity. Here we start with a quantitative framework of connectivity based on the Graph Theory. In order to develop a new index of dynamic connectivity in a river network, we examine the channel structure in the vicinity of Lav Kush barrage over a period of 24 years (1990-2014). Based on the graph theory nodes and links, the idea was to extract the channel network from Landsat imagery, and then to construct the nodal structure with connecting links. The results were then used to compute the Hararry index, characteristic path length and the effective path length. Finally, the connectivity indices were analyzed in a multi-scalar framework by using hyperscale graphs which display connectivity indices as a function of both scale and position in the reach (Figure 1). Fig-1a shows undisturbed connectivity with a more heterogeneous nodal structure in the pre-barrage era. Fig-1b indicates a significant loss in connectivity after Lav Kush barrage construction. In the upstream reaches, the channel is almost uniform showing very less nodal density. This lesser nodal density signifies a uniform channel belt thatleads to a homogeneous geomorphic condition. This type of analysis was repeated at seasonal intervals in order to track changes in graph theoretic properties in an attempt to infer changes to the dynamic connectivity structure and their potential relevance to the geomorphic evolution of this river reach.