Conjunctive management strategies of groundwater and surface Water: a case study of meinong reservoir in Taiwan

Water is an essential natural resource for human survival and development. In recent years, global climate change has led to noticeable shifts in rainfall patterns. In Taiwan, the gap between wet and dry years has gradually widened, and rainfall has become shorter in duration but more intense. This has increased the frequency of spring droughts in Taiwan. Hence, there is an urgent need to propose a new water utilization model.

The bucket model is employed to estimate parameters that are challenging to measure, while Bayesian networks are utilized to establish causal relationships among environmental factors. In addition, Bayesian Network is a systematic network based on conditional probability for constructing relationships between factors. It has been shown to capture crucial groundwater flow properties and uncertainties in groundwater systems. This study seeks to alter the previous management strategy, which prioritized the use of surface water during the rainy season. Consequently, two distinct theoretical models were established for comparison. Model 1 gives precedence to surface water usage, whereas Model 2 prioritizes groundwater usage. Compare the remaining water in surface and groundwater before the next rainy season. The results indicate that, under 'high' conditions, the capacity of groundwater and surface water in Model 2 will be 18% greater than in Model 1. This is attributed to groundwater resources flowing to the surface or serving as a source of submarine groundwater discharge due to saturated aquifers. Additionally, Bayesian networks were employed to conduct a sensitivity analysis of factors. The two most influential factors on the target node are rainfall and groundwater inflow and outflow from the outside area.