Comparing approaches for modeling spatially distributed direct recharge in a semi-arid region (Okanagan Basin, Canada)

Abstract

Spatially distributed recharge is compared at two different scales using three different modeling approaches within the semi-arid Okanagan Basin, British Columbia, Canada. Regional recharge was modeled by mapping results for one-dimensional soil columns from the water-balance code HELP (Hydrologic Evaluation of Landfill Performance, V3.80D). The regional model was then compared to two, independently derived, local-scale models to ensure local trends were captured in the regional model, and to compare modeling methods. Average annual recharge, predicted by the regional model, varied from no recharge to 186 mm/yr. For the north Okanagan (Vernon area), regional estimates were compared to Richards’ equation-based MIKE-SHE (V2007) estimates, which showed a significant difference in average annual recharge: 7 mm/yr (MIKE-SHE) and 109 mm/yr (HELP). In the south Okanagan (Oliver area), regional estimates were compared to high-resolution, local HELP estimates. Similar values of average annual recharge were obtained: 34 mm/yr (local) and 42 mm/yr (regional). A comparison with measured actual evapotranspiration data in the north Okanagan, showed HELP over-predicted recharge compared to MIKE-SHE by under-predicting evapotranspiration during summer months. Thus, the use of HELP in semi-arid areas may be limited if accurate estimates of recharge are needed. However, results may give satisfactory groundwater model calibrations results because of high uncertainty in hydraulic properties.