Improving the agro-hydrological simulation of irrigation in catchments under extreme climatic conditions by using ensembles (icee)

Agricultural irrigation is the largest water usage worldwide. The impact of irrigation on the water balance must be accounted for in hydrological and water resources management studies, e.g., in climate change impact projections. At the regional scale (catchments and irrigation command areas, ca. 100 km² to > 100.000 km²), irrigation can be simulated within agro-hydrological catchment models like SWAT or SWAT+. Previous work showed the potential and shortcomings of SWAT compared to the field scale models. Researchers did not yet put much effort on the uncertainty of the simulated irrigation amounts on both field and regional scale. One of the reasons might be the limited availability of observed irrigation amounts.

This project will investigate parameter uncertainties and structural model uncertainties of agro-hydrological models in simulating the irrigation water demand at field and regional scale. We will use data of long-term field experiments in irrigated crop production from three countries with different climates: Germany (Lower Saxony, humid), India (West-Bengal, monsoon) and USA (Texas, semi-arid). For SWAT, we will investigate parameter uncertainty regarding the simulation of soil water, irrigation and crop growth. At the field scale, we will simulate several agro-hydrological models to capture the uncertainty due to model structure. Based on our findings, we will further develop the soil water, plant growth and irrigation sub-routines in SWAT.

With the aim of improving the prediction of irrigation demand in terms of scheduling date and irrigation amount simulated by models, we will generate a multi-model multi-parameter ensemble (super-ensemble) for the three investigated sites and generate uncertainty ranges for the predictors. The ensemble will be calibrated in order to reduce the number of ensemble members to a smaller number of well performing members (sub-ensemble). Applications of the ensemble are climate change projection in strategic long-term planning and operational advisory for farmers in the short-term. In particular, we will investigate, if the new sub-seasonal forecast ensembles (S2S) provided by ECMWF allow an extended lead-time of predictions of irrigation water demand of up to one month.

In collaboration with:

• Indian Institute of Technology Kharagpur (Prof. M. K. Jha, Prof. D. Swain)
• Texas A&M University, Spatial Sciences Laboratory, Department of Ecosystem Science and Management (Dr. Y. Chen, Prof. R. Srinivasan)
• as A&M University, Spatial Sciences Laboratory, Department of Ecosystem Science and Management (Dr. Y. Chen, Prof. R. Srinivasan)