Macropore domination in runoff generation process: A case study by hydrological modelling in Hilly Watersheds of Koshi River Basin, Himalaya

Abstract

<p>A recent initiative by the hydrologic community identified processes that control hillslope-riparian-stream-groundwater interactions as one of the major unsolved scientific problems in Hydrology. It is a long-time argument among hydrologists whether to eliminate the minor details from field-based costing a lot of time, effort, and resources to understand the hydrological process in watershed scale. The modelling approaches are helpful is these cases by focusing on the dominant controllers and might/might'nt bypassing the implications from minor details. In this work, a conceptual semi-distributed rainfall-runoff model for hilly watersheds is used with satellite-based hydrometeorological inputs to parameterize, and thus understand by calibration and validation, at Koshi River Basin, a partly hilly watershed in Himalaya. The semi-distributed model is operated by dividing the river basin into small grids of around 1km<sup>2</sup>, each representing a micro-watershed. Majority of the model concept is drawn from fill and spill approach from previous literature, observations from plot-scale hillslope experiments, and macropore characterization from dye-tracer experiments, which are upscaled at micro-watershed scale. The parameterization in the rainfall-runoff model includes the daily average variables namely, threshold for runoff generation (<em>T</em>), gradient of runoff generation rate (<em>S</em>), saturated hydraulic conductivity for hillslope aquifers (<em>Ksat</em>), and aquifer thickness limit (<em>D</em>). Variable ranges of these parameters were simulated to find the best values (<em>T</em> = 1±0.25cm; <em>S</em> = 0.6 – 0.1; <em>Ksat</em> ≈ 10<sup>5</sup> – 10<sup>10</sup> times original Ksat; and <em>D </em>= 1m). These ranges resulted in over (NSE = 0.6; R<sup>2</sup> = 0.65) during calibration and validation for daily flow volume at the outlet. In these simulations, the <em>Ksat </em>multiplied with factors at several orders higher scale and producing good NSE values shows domination of preferential pathways in runoff generation process. This might represent a flow similar to that of overland flow affecting the surface runoff volume at river basin scale. This model could be used for water budgeting studies in hilly watersheds where several hillslopes dominated by macropores are present.</p>