Uncertainties in agricultural water supply under climate change: Aidoghmoush basin, Iran

Abstract

The planning of future supplies of agricultural water is beset by uncertainties stemming from inaccuracies in general circulation models, assumed greenhouse gases emissions scenarios (GHGESs), imperfect models employed for estimating reservoir inflows and approximate methods for estimating agricultural water demand. The uncertainty of providing agricultural water under climate change was assessed, relying on simulations involving baseline (1971–2000) and future periods (2040–2069 and 2070–2099). Climatic variables were simulated using six atmosphere–ocean general circulation models (AOGCMs) driven by GHGESs A2 and B2 in the Aidoghmoush basin, Iran. Projection of reservoir inflow was performed using the IHACRES model and artificial neural network (ANNs). Agricultural water demand was calculated using the FAO–Penman–Monteith and Hargreaves-Samani (HS) methods. Eight modelling scenarios were considered based on combinations of AOGCMs, GHGESs, reservoir inflow and agricultural water demand projections. Reservoir operation rules were calculated with a particle swarm optimisation algorithm. The results show that agricultural water demand will increase in future periods compared with the baseline period. The operation rule derived from the combination of the HS and ANN models (under GHGES A2) showed the best performance in 2040–2069 by achieving the highest reliability (93%) of water supply. The operation rule derived from the combination of HS and ANN models (under GHGES B2) achieved the highest reliability (95%) of water supply in 2070–2099. The results provide adjusted reservoir operation rules under uncertainty caused by climate change and related impacts on water resources management.