Evaluation of the tropical upper tropospheric cloudiness simulated by the storm resolving models (SRMs) from DYAMOND project

Abstract

Employing a resolution of 5 km or less, DYAMOND models resolve much of the cloud relevant dynamics and better simulate cloud structure and diurnal cycle of precipitation. Nevertheless, cloud properties in DYAMOND simulations vary significantly despite the fact that they resolve deep convection. We focus on evaluating tropical upper tropospheric (UT) ice clouds in high-resolution DYAMOND simulations as the tropics should particularly benefit from the increased resolution because deep convection controls the tropical UT water budget. We analyse the horizontal distribution of total ice water path (TIWP) and its connection to the convective strength, as indicated by the vertical velocity. While the PDF of tropical vertical velocity simulated by the different models is quite similar, the total ice water path (TIWP) connected with those vertical velocities varies strongly. Tropical total IWP is much improved relative to lower resolution models, such as NWP ICON, and generally underestimated compared to active remote sensing. Precipitation, on the other hand, is overestimated. Using precipitation as an indicator for the strength of tropical convection we can compare the associated changes in UT cloudiness to passive and active remote sensing data. We show that in particular for weak convection the UT IWP is underestimated, which is particularly pronounced in lower resolution NWP simulations using parameterized convection. This shows that while cloud scale dynamics is much improved in the high-resolution simulations the microphysics still lead to a large spread in simulated cloud properties. Uncertainty in cloud microphysics prohibits convergence in simulated high cloud properties.