High-resolution soil moisture data reveal dynamics of preferential flow

Soil moisture data from arrays of vertically-aligned sensors have been used in various ways to detect the occurrence of preferential flow (PF) in the unsaturated zone. Many such data are available at only a few depths, often 5 or fewer, and at fairly long time intervals, often 15 minutes or more. Some soil-moisture networks provide data of substantially greater resolution. One of these, the National Ecological Observatory Network (NEON) in the United States, provides soil moisture data at many locations over 18 ecoregions at 1-minute intervals, at as many as 8 depths, and as deep as 2 m. Evaluated with regard to soil moisture dynamics, such high-resolution data make it possible to go beyond the basic occurrence or nonoccurrence of PF to learn about its dynamic qualities: the magnitude and character of PF within distinct soil horizons, its transformation at layer boundaries, its interactions with soil matrix material, and the depth and duration of its influence. In some cases the rate of change of water content over small depth intervals can permit quantification of fluxes at various positions within the soil profile so that these fluxes can be evaluated with respect to the concurrent intensity and cumulative quantity of water input at land surface.

Investigation of these quantities and qualitative behaviors for identified storm periods at selected NEON locations confirms some of the prevailing expectations about PF, while also revealing new or unexpected features of potential importance. These results provide a strengthened basis for needed improvements in least two types of predictive hydrologic models: (1) for predicting the occurrence of PF in response to site characteristics and varying conditions of soil and weather, and (2) for realistically representing the PF component in general-purpose multi-domain models of flow in the unsaturated zone.