Convectively driven wind variability in connection to wind biases in the ECMWF operational weather model

Despite playing a key role in the atmospheric circulation, the representation of momentum transport by moist convection (cumulus clouds) has been largely overlooked by the model development community over the past decade, at least compared with diabatic and radiative effects of clouds. In particular, how shallow convection may influence surface and boundary layer winds is not thoroughly investigated. In this talk, we discuss the role of convective momentum transport (CMT) in setting low-level wind speed and its variability and evaluate its role in long-standing wind biases in the ECMWF IFS model.

We use high-frequency wind profiling measurements and high-resolution large-eddy simulations to inform our understanding of convectively driven wind variability. We do this at two locations: in the trades, using wind lidar and radiosonde measurements from the Barbados Cloud Observatory and the intensive EUREC4A field campaign, and over the Netherlands, using an observationally constrained reanalysis wind dataset and large-eddy simulation hindcasts.

At both locations we use the data and model output to investigate whether CMT can be responsible for a missing drag near the surface in the IFS model. Namely, at short leadtimes, the model produces stronger than observed easterly/westerly flow near the surface, while “a missing drag” produces weaker than observed wind turning. Consequently, the meridional overturning circulation in both the tropics and midlatitudes is weaker in the IFS and in ERA-Interim and ERA5 reanalysis products.

Comparing simulated and IFS wind tendencies at selected grid points at the above locations, and by turning off the process of CMT by shallow convection in the model, we gain insight in the role of CMT in explaining wind biases. We find that CMT alone does not explain a missing drag near the surface. CMT often acts to accelerate winds near the surface. But CMT plays a role in communicating biases in cloud base wind speeds towards the surface. In the trades, a strong jet near cloud base is determined by thermal wind and a strong flux of zonal momentum through cloud base, where “cumulus friction” minimizes. Near this jet, the presence of (counter-gradient) turbulent momentum fluxes produces most of the drag. Implications of these findings for CMT parameterization are discussed.