Large-scale hydrological modelling :physical parameterisation for groundwater recharge

Abstract

There is currently worldwide interest in the effect of human activity on tile global environment, especially the effect of greenhouse gases and land-use change on the global climate, and models are being developed to study both global change and the local effects of global change. The research reported here (funded by CNPq-Brazil) involves the development of GRASP:Groundwater Recharge modelling Approach with a Scaling up Procedure. GRASP has been integrated into the UP (Upscaled Physically-based) macromodel, developed under the UK NERC TIGER programme, which is designed for studying the effects of climate and land-use change on the availability and quality of water resources. The UP macromodel will be coupled to the UK Meteorological. Office's Unified (weather and climate) model to create a state-of-the-art coupled atmospheric/hydrological model. Several important requirements for the design of new large-scale hydrological models are identified in a wide ranging review on GCMs; (General Circulation Models) and physically -based hydrological modelling, and these requirements have been applied in the development of GRASP(and UP). The main requirements are a physical basis, proper treatment of spatial variability, and simplicity. Using the concept of partial analysis, two point-scale models, SM (Soil Moisture content approach) and TF (Transfer Function approach), are developed for recharge, both based on the one-dimensional Richards' equation. SM is a simple two-parameter model relating recharge to water storage in the unsaturated zone, and several unsuccessful attempts are made to link its parameters to physical propcrties. TF is a transfer function model, and is parameterised using the matric potential and unsaturated hydraulic conductivity functions using a new approach developed especially for GRASP. Both SM and TF are verified against numerical solutions of Richards' equation.