Assessing hydrological connectivity development by using a photogrammetric technique with relative surface connection function (RSCf) in a plot-scale experiment

Abstract

Understanding the development of hydrological connectivity (HC) is important for improving the prediction of overland flow generation and/or erosion processes in modeling all scales from plot to watershed. In this study we aim to reveal the links between the dynamics of HC and the evolution of the microtopographical features at the plot scale and to understand how the development of HC is affected by environmental factors (e.g., rainfall intensity and slope). We collected a loess soil from a small catchment of Loess Plateau of China, known for its high erosion susceptibility, and carried out sequential rainfall simulations with intensities of 25, 50, 70, and 86 mm h⁻¹, and different slope surface gradients of 5°, 15°, and 25°. Hydrological conductivity development was assessed by means of a photogrammetric technique, a semivariogram analysis (in terms of parameters γ and l delivering information on microdepressions), and a simplified hydrograph and its derived relative surface connection function (RSCf), in terms of parameters C₀, C_thr, and D_thr, representing the runoff generation process. The results suggest that slope played a predominant role in affecting the dynamics of the structural aspect of HC. Parameter γ after events (posterior-γ) showed an increasing trend as slope increased. The C₀, as expected, was directly controlled by the rainfall intensity, while the C_thr was indirectly influenced by both rainfall intensity and slope gradient as result of the increase in Δγ. The increase in γ before events (i.e., prior-γ) exhibited a restricting role on the development of the functional aspect of HC on both moderate and steep slopes, i.e., the higher the prior-γ, the less developed the parameters of the functional HC. By relating the parameters C₀, C_thr, and D_thr to the easily measured prior-γ and accounting for the microscale HC dynamic information, it is possible to improve the predictions of overland flow and soil loss generation.