Developing a Methodology for Model Intercomparison and Its Application to Improve Simulated Streamflow by Land Surface Models

Abstract Runoff generated by land surface models (LSMs) is extensively used to predict future river discharge under global warming. However, the structural bias of LSMs, precipitation climate model, and other factors could cause runoff be biased. A model intercomparison study can help understand LSM behavior. Traditional discover output variation evaluate performance, but explaining reason for difference challenging. This developed a novel method identify reasons disparities suggest improvements. Consequently, we investigated impacts settings adopting another in one until it mimic similar features its output. Hence, process called “emulation model.” We employed two LSMs [Simple Biosphere with Urban Canopy (SiBUC) Meteorological Research Institute Simple (MRI-SiB)] Thai River basin. SiBUC produced higher than MRI-SiB, development MRI-SiB emulation revealed this variation. The differences characteristics affected streamflow estimation. For instance, peak was faster observed dry year. Conversely, there tendency underestimate flow estimated during transition from wet seasons. Incorporating alleviate shortcomings each model. Overall, proposed strengths weaknesses enhance reproducibility hydrological Significance Statement aims develop new methodology Understanding why behave differently important enhancing reliability prediction. Our findings guide what affects runoff-based estimation, which will reduce uncertainty flood drought predictions.