Improved simulated diurnal hydrologic cycle in a GCM with super-parameterized clouds

Poster PDF

Description

We show that the Multi-scale Modeling Framework (MMF) approach to climate modeling results in improved simulations of small-scale daily rainfall variability. MMFs are GCMs with cloud super-parameterizations replacing conventional statistical parameterizations. That is, MMFs represent sub-grid cloud and boundary layer processes using an array of nested cloud resolving models in each GCM grid volume. The MMF (SPCAM3) and GCM (CAM3) compared in this study are identical except for this difference. Analysis of the vertically integrated diurnal composite moisture budget provides several new clues to the physical processes involved. Daytime entrainment humidification resolved by the embedded cloud-resolving model (CRM) in the Super-Parameterized Community Atmosphere Model v3.0 (SPCAM3) tempers the amplitude and fixes the timing of the overly vigorous CAM3 diurnal rainfall cycle over land. Diurnal water budget analysis shows that at night, and over the ocean, substantially different representations of diurnal moisture convergence in the two models play a major role in differentiating daily tropical rainfall. In CAM3 a large-scale equatorial planetary wave of diurnal moisture convergence and storage connects the vigorous over-land rainfall cycle to the diurnal rainfall cycle over open ocean thousands of kilometers away. The absence of this wave in SPCAM3 is another, more subtle reason for its improved diurnal rainfall cycle. Only a few MMFs exist. They are untuned, and development is still in its infancy. MMFs are extremely expensive computationally, typically about 200 times more demanding than a conventional GCM. Data from the ARM Climate Research Facility will be invaluable in improving the realism of MMFs.