For analysing diverse water availability and demand options, a suitably flexible modelling framework is necessary. The Water Evaluation And Planning (WEAP) model of the Stockholm Environment Institute is a water demand, priority and preference driven model which provides a modelling framework for assessing sectoral demand, stream flow, reservoir operation, water conservation measures, allocation priorities and project cost-benefit analyses (Yates et al. 2005a; Yates et al. 2005b). It is particularly useful for comparing scenarios of hydrological change (Höllermann et al. 2010; Harma et al. 2012) and has been used for evaluating adaptation options previously (Bonelli et al. 2014; Yates et al. 2015). For this study, the Food and Agriculture Organization (FAO) rainfall-runoff method (Allen et al. 1998) available in the hydrology module of WEAP is used to simulate various hydrological processes including runoff and infiltration. In this method evapotranspiration is determined for irrigated and rainfed crops using crop coefficients (Kc). The remainder of rainfall not consumed by evapotranspiration is simulated as runoff to a river, or distributed between surface runoff to a river and flow to groundwater. A monthly time step has been used for calibration, validation and future scenario analyses to cover the residence time of the study area, during which all flows are assumed to occur. All demand sites are located in the command areas, while WEAP catchments; nodes at which catchment processes occur, cover the entire study area as shown in Fig. 1. While the Mohanpur catchment drains runoff to the Mohanpur gauging station on the Kangsabati river, runoff from the LBMC and RBMC command catchment nodes does not drain to the Kangsabati river. A small dam (anicut) at Mohanpur diverts water into the Midnapore Canal command area. The calibration and validation was carried out for the period 1991–2010, results of which have been described in detail previously (Bhave et al. 2014b, 2014c).
Projections of Climatic and non-Climatic Factors
Projections of climatic change, population change, urban water demand and agricultural water demand have been incorporated into this assessment. Four RCM simulations are used in this study, which represent the most comprehensive ensemble of high resolution future climate change projections available for this region (Mittal et al. 2014; Mathison et al. 2015). These simulations, based on the Special Report on Emissions Scenarios (SRES) A1B scenario were derived by forcing Regional Climate Models (RCMs) REMO and HadRM3 with two Coupled Model Intercomparison Project 3 (CMIP3) Global Climate Models (GCMs) ECHAM5 and HadCM3. This results in four RCM simulations, REMO-ECHAM5, HadRM3-ECHAM5, HadRM3-HadCM3 and REMO-HadCM3. Ability of RCM simulations to reproduce the local temperature and precipitation patterns (Fig. 2) is found to be adequate, as the comparison with Climatic Research Unit (CRU) and Asian Precipitation - Highly-Resolved Observational Data Integration Towards Evaluation of water resources (APHRODITE) data indicates. However, models reproduce spatio-temporal variation for temperature in a better manner than for precipitation, a finding consistent with previous studies in this basin and beyond (Mittal et al. 2014). Individual RCM projections along with their Multi Model Ensemble (MME) mean have been used for forcing the WEAP model, resulting in a total of five future climate simulations.
Agricultural water demand is expected to change in the future due to changing precipitation characteristics and evapotranspiration rates due to the increasing temperature. To address this irrigation requirements are modeled for the four WEAP catchment nodes in the command areas; LBMC, RBMC, Mohanpur and Midnapore Canal. For existing cropping pattern agricultural water demand is expected to increase significantly for the LBMC, RBMC and Mohanpur catchments, while Midnapore Canal Command does not indicate significant change (Supplementary S1). Three major towns exist in the command area; Jhargram, Midnapore and Kharagpur. Their water demand characteristics depend on the specified annual activity level, population and the per capita water use rate. Assumptions about these are based on a report of the Central Public Health and Environmental Engineering Organisation (2005) of the Govt. of India. Increasing population in these three towns in the command area (based on Mahmood and Kundu 2006) leads to increasing water demand (Supplementary S1). The use of projections of both climatic and non-climatic factors provides a more holistic understanding of the water availability and demand characteristics of the Business As Usual (BAU) scenario of the future.
WEAP was run for various scenarios as described below.
The validated model is run for the period 2021–2050 using,
Time series of climatic parameters from four RCM simulations and the MME
Evapotranspiration rates based on Penman-Monteith method
Land use/land cover based on the unsupervised classification of the LandSat ETM+ (Enhanced Thematic Mapper) imagery for 2011
Same reservoir operating rules
A BAU scenario without incorporation of any adaptation options is obtained and is the reference scenario for distinguishing the effect of adaptation options.
IFC has been formulated according to guidelines of the National Green India Mission under the National Action Plan on Climate Change (NAPCC) (Govt. of India 2008). The mission envisages afforestation of degraded public lands and increase the canopy cover of forested areas which currently fall under the ‘open forest’ category (< 40 % canopy cover). Accordingly, in the WEAP model land use change is incorporated where open forest is converted to dense forest and barren land is converted to open forest (for details refer to Bhave et al. 2014b, 2014c).
Prior to the evaluation of check dams as adaptation options, potential locations have to be determined and prioritized so as to suit local physiographical characteristics. In a Geographical Information Systems (GIS) environment using the SCS curve number method, key morphometric characteristics such as drainage network, basin geometry, drainage texture, relief and land use characteristics, nine CD locations were obtained. Check dams were applied on 2nd order streams, and assumed to have storage capacity of 0.05x106 m3, no buffer storage, uncontrolled spillage, surface evaporation losses, no irrigation potential and groundwater recharge (assumptions and process described in detail in Bhave et al. 2014c).
The Govt. of West Bengal had proposed an optimal cropping pattern for Kharif (summer monsoon) and Rabi (winter) crops (Govt. of West Bengal 2003) (Supplementary S2) which has been developed as adaptation option CCP. Following Mehta et al. (2013) WEAP catchment nodes in the command areas were given an additional ‘irrigated crop’ land use. Crop water demand at a monthly time step was assigned using reference crop evapotranspiration and crop coefficients based on the Kangsabati Project Performance Evaluation Report (Govt. of West Bengal 2003). When sufficient water is available, the model applies water required for irrigation while taking into account conveyance and transmission losses. The cropping pattern has been distributed evenly across the command area because details regarding spatial distribution are not available. Although groundwater is used for supplemental irrigation in this region, no data is available on the extent of its use. It has been neglected in this study, since reservoir canal based irrigation is considered to be the predominant form of irrigation.
Traditional ponds are usually non-masonry dug ponds having a natural drainage, used for fish cultivation, growing vegetables on its periphery and sometimes for local irrigation. Although spatially distributed on the ground, in WEAP, the storage of these tanks is aggregated for each of the four command area demand nodes. This assumption may impact evaporation rates, due to the difference in volume-area relationship. Hence, evaporation rate for ponds is assumed to be similar to that observed in the Kangsabati reservoir (Govt. of West Bengal 2003; Leemhuis et al. 2009). One pond, with 2000 m3 storage capacity is assumed for every 1 km2 area (de Condappa et al. 2009). The ponds are replenished by August (during monsoon), water is depleted by direct evaporation and groundwater recharge has been neglected. Only three WEAP catchment nodes; LBMC, RBMC and Mohanpur have been modelled with TP. This is because TP located in the Mohanpur catchment have an impact on streamflow at Mohanpur station, which affects water availability for the Midnapore Canal catchment. Therefore, to avoid a compound effect being produced in the Midnapore Canal catchment it has not been provided with TP in this study.
Urban WWR reduces water supply requirement by a factor of ‘1 – reuse rate’. This adaptation option is applied to three major towns in the study area; Midnapore, Kharagpur and Jhargram. Assumptions in this analysis are based on Central Public Health and Environmental Engineering Organisation (2005) and population projections are based on Mahmood and Kundu (2006). A water demand of 95 l/capita/day and population projections for the district of Paschim Midnapore (in which these three towns lie) is assumed. A gradual increase in WWR is assumed with 25 % in 2021, 30 % in 2031 and 35 % in 2041 to reflect the increasing focus on WWR and changing technological and infrastructural characteristics.