Towards predicting the initiation of overland flow from relatively flat agricultural fields using surface water coverage

On rough agricultural soils, initiation of overland flow is primarily related to the gradual filling of small depressions. As the volume of water ponding in local depressions increases, the connectivity of those depressions increases, and that connectivity permits flow across the field boundaries. Previous studies have aimed at predicting overland flow connectivity by means of depression storage, but this is exceedingly difficult to quantify during an irrigation or rainfall event. There has been little work linking overland flow connectivity through variables that are more readily-measured than depression storage, such as the proportion of the soil surface covered in water (ASW). We propose using ASW as a proxy for hydrological connectivity, which can be measured using proximal remote sensing. A series of overland flow experiments were conducted in two contrasting plots of ~1.5 m2. Outlet discharge and changes in soil surface covered in water were continuously recorded during the experiments. The experiments demonstrated that the rougher soil surface experienced a delayed initiation of overland flow. The results also showed overland flow initiation was characterised by a distinct connectivity threshold and showed a clear response of overland flow to ASW. We further investigated the hydrological connectivity in twenty-one additional agricultural soils with contrasting micro-topographic conditions. Our results suggest that a predictor based on ASW has substantial potential to predict initiation of overland flow. The prediction is sufficiently early that it could be used in modern variable-rate irrigation systems, in combination with a sensor to measure ASW in real time, to prevent substantial flow from fields during irrigation.