Using River Discharge Observations to Understand Precipitation Biases in CMIP6-HighResMIP Simulations

MÜLLER, OMAR V.; VIDALE, PIER LUIGI; VANNIERE, BENOIT; SCHIEMANN, REINHARD; MCGUIRE, PATRICK C.

Conferencia; AGU Fall Meeting 2021; 2021

American Geophysical Union

Previous studies showed that high-resolution global climate models (GCMs) overestimate land precipitation when compared against observation-based data. Particularly, grid point models show a significant increase of precipitation on mountain regions, where the scarcity of gauge stations increase the uncertainty of gridded observations and reanalysis. The goal of this work is to evaluate such precipitation uncertainties indirectly through on river discharge, considering that: a) river flow is an integrator of the water balance at catchment scale, b) an increase of ~10% in land precipitation produces ~28% more runoff when resolution is enhanced from 1° to 0.25°, and c) ~50% of the global runoff is produced in 27% of global land dominated by mountains.We diagnosed the river flow by routing the runoff generated by HighResMIP HadGEM3-GC31 simulations at low and high resolution (~1° and ~0.25° respectively). The river flow is evaluated using a set of 344 monitored catchments distributed around the world. We also infer the global discharge by constraining the simulations with observations following a novel approach that implies bias correction in monitored rivers with two methods, and extension of the correction to the river mouth, and along the coast.Our global discharge estimate is , which is closer to the original high-resolution estimate () than to the low-resolution (). The assessment suggests that high-resolution simulations perform better in mountainous regions, either because the better-defined orography favors the placement of precipitation in the correct catchment, leading to a more accurate distribution of runoff, or the orographic precipitation increases reducing the dry runoff bias of coarse resolution simulations. In summary, we have another good evidence that high resolution must be closer to the real world and current observational estimates must underestimate land precipitation. The improvement of model parameterizations and tuning may reduce the remaining errors in high-resolution simulations. Future efforts will focus on the development of river routing simulations at 0.1° resolution in response to the needs of the new generation of high-resolution models.