Assessment of the vertical preferential flow characteristics and flow types on a slope in a small headwater catchment

This study investigated preferential flow infiltration patterns at three slope positions (upslope, mid-slope and downslope) of typical slopes in the Miyun Reservoir Basin. The evaluation was conducted by combining the multi-index evaluation method and the macropore flow classification method, based on the profile images obtained from dye tracer experiments. Multi-index analysis shows that preferential flow infiltration depth is 400 mm, and the average uniform infiltration depth is 56.6 mm. The preferential flow fraction quantified as the deviation from a perfectly homogeneous infiltration process on up, middle and down slopes are 56.6%, 74.8% and 67.5% respectively. Middle slope has higher preferential flow level due to higher weights of peak value, stained area and stable infiltration depth. Macropore flow classification results indicate at three slope positions, preferential flow is mainly macropore flow with mixed interaction with soil matrix. It is one of the five flow types and its proportion increases as elevation drops (79.31% on upper, 86.88% on middle and 95.31% on lower slope). With soil layer depth increasing, low-interaction macropore flow proportion rises and its interaction with matrix soil decreases. Correlation analysis reveals that macropore flow with low interaction negatively correlates with soil volumetric water content, porosity, but positively with soil bulk density. Macropore flow with high interaction positively correlates with soil volumetric water content. Matrix flow and fingering positively correlate with silt content. Matrix flow significantly positively correlates with soil saturated hydraulic conductivity. The difference between multi-index method and macropore flow classification lies in analysis scale. Since the calculation scale of the multi-index method encompasses the entire profile and treats the staining characteristics as a whole, it might result in an underestimation of preferential flow outcomes. On the other hand, the preferential flow classification method considering the interaction with soil matrix at pixel scale may lead to an overestimation of the calculated preferential flow results. Combining these two methods can help judge macropore flow characteristics and its interaction with soil matrix more accurately.