A geo-informatics approach for estimating water resources management components and their interrelationships
A remote sensing based geo-informatics approach was developed to estimate water resources management (WRM) components across a large irrigation scheme in the Indus Basin of Pakistan. The approach provides a generalized framework for estimating a range of key water management variables and provides a management tool for the sustainable operation of similar schemes globally. A focus on the use of satellite data allowed for the quantification of relationships across a range of spatial and temporal scales. Variables including actual and crop evapotranspiration, net and gross irrigation, net and gross groundwater use, groundwater recharge, net groundwater recharge, were estimated and then their interrelationships explored across the Hakra Canal command area. Spatially distributed remotely sensed estimates of actual evapotranspiration (ETa) rates were determined using the Surface Energy Balance System (SEBS) model and evaluated against ground-based evaporation calculated from the advection-aridity method. Analysis of ETa simulations across two cropping season, referred to as Kharif and Rabi, yielded Pearson correlation (R) values of 0.69 and 0.84, Nash-Sutcliffe criterion (NSE) of 0.28 and 0.63, percentage bias of −3.85% and 10.6% and root mean squared error (RMSE) of 10.6 mm and 12.21 mm for each season, respectively. For the period of study between 2008 and 2014, it was estimated that an average of 0.63 mm day−1 water was supplied through canal irrigation against a crop water demand of 3.81 mm day−1. Approximately 1.86 mm day−1 groundwater abstraction was estimated in the region, which contributed to fulfil the gap between crop water demand and canal water supply. Importantly, the combined canal, groundwater and rainfall sources of water only met 70% of the crop water requirements. As such, the difference between recharge and discharge showed that groundwater depletion was around −115 mm year−1 during the six year study period. Analysis indicated that monthly changes in ETa were strongly correlated (R = 0.94) with groundwater abstraction and rainfall, with the strength of this relationship significantly (p < 0.01 and 0.05) impacted by cropping seasons and land use practices. Similarly, the net groundwater recharge showed a good positive correlation (R) of 0.72 with rainfall during Kharif, and a correlation of 0.75 with canal irrigation during Rabi, at a significance level of p < 0.01. Overall, the results provide insight into the interrelationships between key WRM components and the variation of these through time, offering information to improve the management and strategic planning of available water resources in this region.