Birch, Andrew L听1听;听Stallard, Robert F听2听;听Barnard, Holly R听3
1听Institute of Arctic and Alpine Research, Department of Geography 麻豆影院
2听United States Geologic Survey, Institute of Arctic and Alpine Research, Smithsonian Tropical Research Institute
3听Institute of Arctic and Alpine Research, Department of Geography 麻豆影院
While our understanding of the thresholds governing the rainfall-runoff response of catchments in temperate latitudes has increased significantly in recent years, very little work has been done to understand runoff generation thresholds in the humid tropics. Understanding these thresholds is essential to understanding how the hydrologic function of tropical catchments will change under the intensifying threats of land use and climate change. A three month time series (May-August) of streamflow and electrical conductivity (EC) was collected during the 2017 Central American wet season from three small headwater catchments of varying land cover (mature tropical forest, young secondary tropical forest, active cattle pasture) within the Panama Canal Watershed. These data were used to perform continuous baseflow/ event water hydrograph separations using a one component- two end member mixing model, yielding hydrograph separations for 61 individual storm events across the three catchments. Storm hydrograph separation results show that, while old water is the dominant source of runoff during small storm events, new water dominates the hydrograph of larger events across all three land covers. We suggest that the runoff rate at which event water fraction surpassed baseflow to become the dominant source of runoff is indicative of the activation of shallow, near-surface flowpaths, allowing for the fast delivery of event water to the stream. Exceedance of this one to one threshold at each land cover was found to be variably dependent on precipitation characteristics (total event precipitation, maximum precipitation intensity) at the forested sites, with the mature forest requiring less total precipitation and lower maximum precipitation intensities for event water to dominate than at the young secondary forest. At the pasture, only a weak relationship was observed between precipitation forcings and maximum event water fraction, suggesting that soil moisture conditions may dictate the threshold for shallow flow path activation in this land cover. Our results support a conceptual model where precipitation activated shallow lateral flowpaths (preferential flow) dominate storm runoff generation at the forested land covers, while antecedent soil moisture conditions regulate the activation of overland flow as the primary source of storm runoff at the pasture during larger events.