An Intercomparison of the Lifecycle of TTL Cirrus and Top-of-Atmosphere Radiative Effects in the DYAMOND Simulations

Abstract:

Tropical Tropopause Layer (TTL) cirrus have a significant radiative warming effect due to their high altitude, thin optical depths, and vast spatiotemporal extent. In the Tropical Western Pacific (TWP), the slight warming effect of cirrus in the TTL and upper troposphere accumulates over long cirrus lifetimes to effectively cancel out the large cooling effect of deep convective cores at the top-of-atmosphere (TOA). Their high altitude and thin optical depths also make them extremely difficult to observe with radar, lidar or aircraft measurements. A lack of robust observations has limited the study of TTL cirrus; even then, they provide a general starting point for model comparison. Thus, we turn to high resolution Global Storm-Resolving Models (GSRMs) to analyze TTL cirrus. GSRMs use explicit physics to simulate deep convection, a primary source of TTL cirrus. This study uses GSRMs from the DYnamics of the Atmospheric general circulation On Non-hydrostatic Domains (DYAMOND) project. For context, we compare these simulations with radar and lidar retrievals for characteristics such as ice water path, ice water content, and radiation fluxes. The microphysics and dynamics determine the cloud type from deep convective cores to anvil outflow and thin cirrus.

Here, we analyze the lifecycle of TTL cirrus in the DYAMOND simulations compared to available observations. We use a statistical approach through joint albedo-OLR histograms to show the model variation of cloud type distributions. We focus on the TWP but also consider a nearby less convective oceanic region as well as continental convection over the Sahel. Infrequent deep convection brings moisture into the TTL allowing cirrus to form there and persist. Some models generate thicker cirrus or anvils while thin cirrus dominate in other models. While the models generally capture the diurnal cycle of convection, the frequency and radiative properties of thin cirrus vary between the models. Understanding the differences in these models is one important step for reducing uncertainty generated from tropical high cloud feedbacks for predicting future warming.