Tropical deep convection produces a wide range of cloud types. Different parts of these cloud systems exert strong positive and negative radiative effects, but the climatological radiative effect of tropical convection is near zero. This balance results from a very specific distribution of convective cloud types, and we are unsure of how it will be affected by climate change. If the radiative balance of tropical convection shifts in the future, it could constitute an important climate feedback.

My work uses a variety of tools to understand the distribution, life cycle, and climate impact of tropical convective clouds. Satellite observations have revealed that the most common anvil clouds in the Tropics are those with the strongest warming effect on climate. Using a radiative transfer model and satellite retrievals of cloud microphysics, we found that these clouds are selectively maintained by radiative heating. Here is the paper in JGR-Atmospheres.

Using satellite measurements and a cloud-resolving model, we have also found that anvil clouds produced during the day have a longer lifetime and a different climate impact than those produced at night. Read that paper here.

More recently, we've studied how convective clouds respond to surface warming in a the RCEMIP ensemble of cloud-resolving models. Our findings suggest that the anvil cloud area feedback is near zero and that the optical depth feedback is slightly positive—which has some important implications for estimates of climate sensitivity. Check out my climate sensitivity symposium talk on this work or our paper in Nature Geoscience.

I have also studied the spatial heterogeneity of cloud phase in mixed-phase clouds around the world. This work was motivated by the need to improve the partitioning of mixed-phase clouds in climate models, since biases in phase partitioning lead to biases in radiative fluxes. Using single-shot profiles from the CALIOP lidar, we quantified cloud phase heterogeneity on a global scale and hypothesizes that the increased availability of ice nucleating particles during the spring leads to more heterogeneous cloud phase at that time of year. Here is the paper in JGR-Atmospheres.

Regions of tropical rainfall such as the ITCZ are generally expected to contract in response to climate change. We examined the hydrological response to warming in idealized mock-Walker simulations, which we expected would be the ideal scenario to study the thermodynamic mechanisms that are hypothesized to drive the contraction of tropical ascent. Surprisingly, our simulations showed a robust expansion of tropical ascent in the presence of both fixed and interactive sea surface temperatures. This unexpected response occurs because of rapid weakening of the large-scale Walker circulation. We trace this weakening to increasing gross moist stability and changes in vertical circulation structure. For details, you can find our preprint here.

Interactions between sea surface temperature (SST), deep convection, and the large-scale tropical circulation shape weather and climate across the globe. In a recent project, I studied these interactions using a cloud-resolving model coupled to an interactive slab ocean. These simulations generated unique, internal oscillations in mean SST, SST gradients, convective organization, and the stratification of the tropical troposphere. You can find our results here.

I am also interested in the vertical structure of the the tropical overturning circulation, in particular the forgotten "congestus" mode of convection. Tropical convection has three distinct modes: deep (thunderstorms), shallow (trade cumulus), and congestus (something in between). Some models simulate all three of these modes, but many do not. Trying to understand intermodel variability in the congestus mode, we found that the strength of congestus convection is affected by the large-scale organization of convection. When convection is tightly clustered into a single region, the congestus mode is strengthened by moisture-radiation feedbacks. Because the congestus mode has significant impacts on mean climate, it is important to understand the interplay between the three modes of convection. Read our JAMES paper here., and stay tuned for future work looking at the vertical structure of the tropical circulation in global storm-resolving models.

I've also examined how surface warming affects the balance between radiative cooling and latent heating in the tropical upper troposphere. Our main finding was that increases in the average amount of cloud ice are much smaller than increases in the rate of radiative cooling. This means that the residence time of cloud ice in the upper troposphere decreases, i.e., that ice is cycled through the atmosphere at a faster rate. Read the paper here.