Validating runoff of the Community Earth System Model (CESM) CMIP6 simulation against surface observation and reanalysis data

Abstract:

Human activity has had drastic effects on many natural systems and the climate itself, and intense population growth is likely to alter such systems further, particularly the hydrologic cycle. Already, urban flooding and runoff have become a prominent issue for many cities and regions, arising from a combination of altered precipitation patterns, urban growth, development in floodplains, and increases in impervious surfaces. Additionally, climate models consistently project that frequency, severity, and duration of hydroclimatic extremes will increase markedly over this century under climate change. However, there is still uncertainty in accurately representing the hydrologic cycle in climate models. Validation of hydrologic variables modeled by the state-of-the-art climate models provides insights of their improvement in representing hydrologic processes and the benchmark of the uncertainty in models’ future hydroclimatic projections. In this study, we validate the total runoff (including throughflow) from the fully coupled Community Earth System Model (CESM) historical simulations that participate in the CMIP6 against two datasets: UNH-GRDC and MERRA-2. The UNH-GRDC is an observational-based surface runoff dataset derived from a comprehensive global record of streamflow discharge and a water balance model; whereas the MERRA-2 is global reanalysis data from NASA. Preliminary results show spatial consistency between the CESM simulations and the UNH-GRDC composite runoff data. Comparison between MERRA-2 and CESM suggests that CESM calculates higher runoff values than those in presented in MERRA-2. However, there is some spatial consistency in the areas highlighted as major runoff producers, such as the Amazon basin and Southeast Asia. Numeric validation is ongoing to further quantify the discrepancies between CESM and both UNH-GRDC and MERRA-2 datasets.