Infiltration, Macroporosity, and Mesoporosity Distributions on Two Forested Watersheds

Abstract

Macro- and mesopore processes substantially control the subsurface flow in forested watersheds. Limited field scale information is available on spatial variability of macropore infiltration and associated porosity of the hydrologically active macro- and mesopores. The double-ring and tension infiltrometer methods were employed at 37 and 39 locations, respectively, on two contrasting forested watersheds. The spatial variability of infiltration under ponded-flow (macropore-flow) and under 2-, 5-, and 14-cm water tension (mesopore-flow) conditions was determined. The frequency distributions were tested for lognormality with the Shapiro-Wilk ω-statistic and with isopleth probability analysis, and the spatial dependence of these data was tested with semivariogram analysis. The infiltration rates were found to be lognormally distributed, with the mesopore infiltration rates as variable (coefficients of variation of 102–184%) as the macropore infiltration rates (coefficient of variation of 107%). Macropore flow constituted 85% of the ponded flux; however, the mesopore fluxes were large (≈2 × 10−5 m s−1) and were considered sufficient to infiltrate rainfall without macropores filling and contributing to the flow. The large measured infiltration rates were associated with exceedingly small macroporosities of 0.0003 m3 m−3 for Walker Branch Watershed and 0.0002 m3 m−3 for Melton Branch Watershed, respectively. The reduction in infiltration rate with increased tension was described as an exponential function of the diameter of the largest effective pore. Semivariograms revealed no spatial dependence for separation distances >4 m at Walker Branch Watershed and weak spatial dependence for separation distances < 15 m at Melton Branch Watershed. Infiltration into these forested watersheds can be considered a stochastic process for hydrologic modeling.