PNNL

The Science                   

Mesoscale convective systems (MCSs)—large cumulonimbus clouds frequently observed over the U.S. Great Plains—are major contributors to heavy precipitation, flooding, and strong winds. However, current climate models are unable to accurately represent MCSs partly because of difficulty simulating the large-scale environments that affect them. Using machine learning, a team from the U.S. Department of Energy’s Pacific Northwest National Laboratory identified four types of environments with favorable circulation patterns for spring MCSs to form. They also identified four types of environments associated with summer MCSs, with circulation patterns that are favorable in two but unfavorable in the others. They found that MCSs can still develop in unfavorable circulation conditions because with abundant moisture and high temperature during summer, MCSs can develop without large-scale forcing. The development of MCSs in unfavorable circulation conditions makes for a more formidable challenge for modeling and predicting summer MCSs than spring MCSs and deserves further investigation.

The Impact

This study provides insights into different environments where MCSs form and how predictable MCSs are in spring versus summer. Understanding the large-scale environments associated with MCSs is important for modeling current MCS and predicting future MCS under environmental changes.

Summary

To identify the different environments associated with MCSs, researchers used a machine learning technique called self-organizing map (SOM) applied to an MCS database and an atmospheric analysis. They found that during spring, there were four SOM types of favorable large-scale environments with two important circulation features: frontal systems providing a lifting mechanism and an enhanced low-level jet transporting atypical moisture from the Gulf of Mexico. During summer, two favorable large-scale environments for MCSs also had frontal characteristics and an enhanced low-level jet, although they were both shifted northward compared to spring. However, summer MCSs were also associated with two other types of large-scale environments featuring enhanced upper-level ridges, which induce subsidence, a broad descending motion in the upper atmosphere that suppresses convection. Analysis suggests that with summer’s abundant moisture and warmer temperatures near the surface, MCSs can still develop in the presence of smaller-scale lifting despite the large-scale subsidence from the upper-level ridges. In both seasons, MCS precipitation amount, area, and intensity are much larger in the MCSs related to frontal systems with a lifting mechanism.

To determine how predictable MCSs are based on the large-scale environment alone, a large-scale index was constructed using pattern correlation between the large-scale environments and the favorable SOM types. The large-scale index accurately estimated the MCS number, precipitation rate, and coverage area in spring, but summer proved more challenging when MCSs can also develop in unfavorable large-scale environments. The low predictability of summer MCSs deserves further investigation.

PNNL Contact

L. Ruby Leung, Pacific Northwest National Laboratory, Ruby.Leung@pnnl.gov

Funding

All authors were supported by the Regional and Global Climate Modeling Program of the U.S. Department of Energy Office of Science Biological and Environmental Research Program.