A decentralized approach for modeling organized convection based on thermal populations on microgrids



Neggers, Roel — University of Cologne
Heus, Thijs — Cleveland State University

Area of research:

General Circulation and Single Column Models/Parameterizations

Journal Reference:

Neggers R and P Griewank. 2022. "A decentralized approach for modeling organized convection based on thermal populations on microgrids." Journal of Advances in Modeling Earth Systems, 14(10), e2022MS003042, 10.1029/2022MS003042.


The spatial structure of convective cloud fields plays an important role in Earth's climate, yet is still poorly represented in weather and climate models. This study explores a new way to achieve this goal, by considering spatial organization at the scale of small bubbles of rising air called thermals that together make up convective clouds. Populations of interacting thermals are modeled in a computationally efficient way on a small 2D grid, which is then coupled to a convection scheme. The new framework is tested for observed conditions at ARM's Southern Great Plains (SGP) site. Apart from providing proof of concept, the results give new insights into how the spatial structure of convective cloud populations can affect their vertical transport.


Grid models used for global weather prediction and climate simulation are typically "blind" for the spatial organization in convective cloud fields, because this phenomenon takes place at scales that remain at least partially unresolved at present-day feasible resolutions. The thermal clustering emergent on the microgrid as explored in this study, and its combination with a spectral convection parameterization, has the goal of resolving this longstanding problem. The system is found capable of reproducing spatial organization and scale growth during two observed cases of shallow convection. New insights are also obtained into how organization and convective transport interact. While these discoveries represent progress in addressing the convective organization challenge, they also invite further rigorous testing against ARM data.


In this study a spectral multi-plume Eddy Diffusivity Mass Flux (EDMF) scheme is coupled to a thermal population model on a 2D horizontal “microgrid”. Thermals are considered the smallest building block of convection, with life cycle and movement represented through binomial functions. They interact through two simple rules, reflecting pulsating growth and environmental deformation. Long-lived thermal clusters thus emerge on the microgrid, exhibiting scale growth that represents a simple form of spatial organization and memory. Size distributions of cluster number are diagnosed from this Binomials on Microgrids (BiOMi) model and provided to EDMF. This yields a decentralized transport system, in that the thermal clusters act as independent but interacting nodes that carry information about spatial structure. The BiOMi-EDMF system is tested through single-column model (SCM) experiments at the ARM SGP site, demonstrating satisfactory skill in reproducing two observed cases of continental shallow convection. Metrics expressing self-organization and spatial organization match well with large-eddy simulation results. Spatial organization impacts EDMF transport through the size distribution of cluster number. The rooting of saturated plumes in the sub-cloud mixed layer plays a key role in this process.