Evaluation of water-saving methodologies for rice farming using a distributed long-term rainfall-runoff model incorporated agricultural ponds detected by multi-temporal SAR images in the Ibague Watershed, Colombia

Abstract

A distributed long-term rainfall-runoff model was developed to quantify the watershed scale impacts of water-saving measures in the Ibague watershed, Colombia. The water-saving effects were evaluated by scenario analyses that incorporated the introduction of a water-saving irrigation method (early stopping) to improve irrigation efficiency with respect to redundant paddy irrigation water. In addition to water stopping, the impacts of a new water-saving rice genotype are analyzed. To calculate rainfall-runoff, the Sugawara tank model was utilized. The tank model quantified runoff discharges from forest, upland, paddy field, and urban areas by incorporating each land type into every mesh for a distributed rainfall-runoff model. This quantified the rainfall-runoff characteristics for each land use. Agricultural ponds and irrigation canals were allocated to each corresponding mesh to model irrigation management practices in the watershed. Agricultural ponds, which have a significant effect on the water balance at the watershed scale, were detected by performing cluster analysis on the seasonal transition of the backscatter coefficient obtained from multi-temporal Sentinel-1 SAR images. Quantitative analyses of the impacts of water-conservation under various scenarios indicated that the early-stopping irrigation method reduced water consumption by 9.4%, compared to the conventional methods. The scenarios, which assumed the introduction of a new water-saving rice genotype, assumed regular irrigation intervals of one or two days and indicated 24.1% and 48.2% reductions, for water consumption, respectively, compared to conventional irrigation methods. Scenarios that implemented a new water-saving rice genotype, as well as the early-stopping irrigation method showed reductions of 30.7% and 53.2% in water consumption compared to the conventional irrigation method, for the one and two day irrigation periods respectively.