Mechanisms of groundwater recharge and pesticide penetration to a chalk aquifer in southern England

In order to assess the potential for enhanced agrochemical contamination of shallow groundwaters, a field study was established on the Upper Chalk in Hampshire, UK. Two instrumented sites, 380 m apart, were established on a hillslope; one on the interfluve with a deep water table (approximately 18 m depth), the other on the dry valley bottom where the groundwater was close to the surface (approximately 4 m depth). Hourly measurements of water potentials in the unsaturated zone to 3.0 m depth identified very different groundwater recharge processes between the two sites in response to the same storm event. On the interfluve site with the deep water table only matrix flow through the chalk unsaturated zone at 3 m depth was identified. In contrast, at the dry valley bottom with a shallow water table, both rapid preferential flow and matrix flow processes were observed at a 3 m depth. The correlation between groundwater depth measurements and unsaturated profile moisture content measurements demonstrated the importance of the capillary fringe in sustaining a higher moisture content in the unsaturated zone at the shallow groundwater site. The resulting reduced water storage capacity for vertical drainage fluxes meant that little water was required to wet the shallow profile before rapid preferential flow events, demonstrated by rapid water potential responses, occurred. However, where the groundwater was deeper, 'intermediate' storage sites located on chalk surfaces and at chalk ped/block 'contact points' remained empty and unsaturated water potential profiles showed that rainfall pulses were attenuated as these sites absorbed the downward water fluxes. Consequently, preferential events at these deep groundwater sites are rare. The importance of these 'intermediate' storage sites in controlling recharge processes is highlighted. The potential mass load of pesticide transported to the shallow groundwater in preferential events during 1996-7 was determined using bromide tracer studies, water balance calculations and measured groundwater pesticide concentrations. An estimated 0.1% of the applied pesticide reached the shallow groundwater along preferential pathways in 1996-7. Calculations under a worst-case scenario showed that this value did not increase beyond 0.2% of pesticide applied.