Abstract
Abstract. This paper presents experimental results from a new headwater research catchment in New Zealand. We made distributed measurements of streamflow, soil moisture and groundwater levels, sampling across a range of aspects, hillslope positions, distances from stream and depths. Our aim was to assess the controls, types and implications of spatial and temporal variability in surface and groundwaters. We found that temporal variability is strongly controlled by the seasonal cycle, for both soil moisture and water table, and for both the mean and extremes of the distributions. The standard deviation of both soil moisture and groundwater values calculated per timestep is larger in winter than in summer, and standard deviations typically peak during rainfall events due to partial saturation of the catchment. Controls on the spatial variability differed between the water stores. Aspect had a strong control on groundwater but not on soil moisture, distance from stream controlled both soil moisture and groundwater. The depth of the soil moisture sensor had little impact in terms of mean water content, but a strong impact on the extreme values, i.e. saturation. Co-measurement of soil moisture and water table level variability allowed us to identify variability components that differed between these water stores e.g. patterns of strong response in soil water content were not the same for groundwater level, and those that were consistent e.g. vertical infiltration of summer rainfall through upper and lower soil depths, or rising near-stream water tables through shallow wells to lower soil depths. Signatures of variability were observed in the streamflow series, showing that understanding variability is important for hydrological prediction. Total catchment variability is composed of multiple variability sources. The dominant variability type changes with catchment wetness conditions according to which water stores are active, and in particular those which are close to a threshold such as field capacity or saturation. Our results suggest that the integrative processes that create emergent catchment behaviour should be understood as the sum of these multiple, time varying components.