In many areas of India, increasing groundwater use has led to depleted aquifers. Rainwater harvesting (RWH), the small scale collection and storage of runoff to augment groundwater stores, is seen as a solution to the deepening groundwater crisis in India. However while the social and economic gains of RWH have been highlighted, there has not yet been a thorough attempt to evaluate the impacts of RWH on larger catchment hydrological balances. The thesis here will endeavour to address this research gap through a case study of the 476 km2 ungauged semi-arid Arvari River catchment in the state of Rajasthan. Over 366 RWH structures have been built in this catchment since 1985 by the community and the local non-government organisation (NGO), Tarun Bharat Sangh (TBS).
The local effects of RWH structures and general catchment characteristics were determined through field investigations during the monsoon seasons of 2007 and 2008. The analysis described large variability in both climatic patterns and recharge estimates. Potential recharge estimates from seven RWH storages, of three different sizes and in six landscape positions, were calculated using the water balance method, which were compared with recharge estimates from water level rises in twenty-nine dug wells using the water table fluctuation method. The average daily potential recharge from RWH structures is between 12 – 52 mm/day, while recharge reaching the groundwater was between 3 – 7 mm/day. The large difference between recharge estimates could be explained through soil storage, and a large lateral transmissivity in the aquifer. Approximately 7% of rainfall is recharged by RWH in the catchment, which is similar in both the comparatively wet and dry years of the field analysis. This is because the capacity of an individual structure to induce recharge is related to structure size and capacity, catchment runoff characteristics and underlying geology. Due to the large annual fluctuations in groundwater levels, the field study results suggest that RWH has a large impact on the groundwater supply, and that there is a large lateral flow of groundwater in the area.
The results inferred from the field analysis were then applied to a conceptual water balance model to study catchment-scale impacts of RWH. An existing model was not used because of the paucity of data, and the need to incorporate an effective representation of RWH function and impact. The model works on a daily time step and is divided into subbasins. Within the subbasin hydrological response units (HRUs) describe the different land use/soil combinations associated with the Arvari River catchment, including irrigated agriculture.
Sustainability indices, related to water from groundwater and rainfall for irrigated agriculture demand, were used to compare scenarios of management simulated in the conceptual model. The analysis shows that as RWH area increases, it reaches a limiting capacity from where developing additional RWH area does not increase the benefit to groundwater stores, but substantially reduces streamflow. This limiting capacity was also seen at the local-scale, where cumulative potential recharge from an individual RWH structure reaches a maximum daily recharge rate. These results could have important implications for RWH development, but require further research. The analysis highlighted the important link between irrigation area and RWH area. If the irrigation area is increased at the optimal level of RWH, where the sustainability indices were greatest, the resilience of the system actually decreased. Nevertheless RWH in a system increased the overall sustainability of the water demand for irrigated agriculture, compared to a system without RWH. Also RWH provided a slight buffer in the groundwater store when drought occurred.
While RWH addresses the supply-side issues of groundwater operation, the institutions that form rules for groundwater use must also be considered, because of the link between irrigation area and RWH. The Arvari River Parliament, the community-based group in the case study area, was examined according to Ostrom’s factors for collective action. It was found that the major limitation for the effectiveness of this group was the minimal information available about the aquifer characteristics.