Stochastic water balance model for rainfall recharge quantification in Ruataniwha Basin, New Zealand

Abstract

Quantification of groundwater recharge is important for water resources management. Different methods can be used to estimate groundwater recharge. The most suitable approach depends on site characteristics. Water balance model was used in this study to quantify groundwater recharge from rainfall in Ruataniwha Basin, Hawke’s Bay, New Zealand. Because it is a closed basin, this method was determined to be more suitable than any other method. The basin is fed by surface water flow and rainfall, without any lateral groundwater flow. Records of surface water inflow and outflow are available with a certain degree of reliability. To cope with uncertainty in different components of water balance, different sampling methods were used. The Mean Value Latin Hypercube Sampling (MVLHS) was used and compared with Latin Hypercube Sampling (LHS), and Monte Carlo Simulation (MCS). It was found the groundwater recharge from rainfall is about 415 million m3/year. Results of this study have revealed that MVLHS converges faster and with lower variance than LHS and MCS.