[Eoas-seminar] Cansu Duzgun's Ph.D. Seminar

[eoas-seminar at lists.fsu.edu]

Please join us for Cansu Duzgun's Dissertation Defense on Monday, May 18 at 10:00 AM (EDT).



Title: Environmental Influences on Convective Entrainment And Their Implications for Parameterized Vertical Transport



Name: Cansu Duzgun

Date: May 18 (Monday), 10:00 AM – 12:00 PM

Location: EOA 6067

Major Professor: Dr. Henry Fuelberg

Zoom: https://fsu.zoom.us/j/91004687866<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Ffsu.zoom.us%2Fj%2F91004687866&data=05%7C02%7Ceoas-seminar%40lists.fsu.edu%7C06c4d962369948766e7608deb035c5ac%7Ca36450ebdb0642a78d1b026719f701e3%7C0%7C0%7C639141942146100268%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=tcj7%2BWwlmHu3Sve4MbqaJZYv68Q4TJPG8zfAHQqBC9s%3D&reserved=0>

Abstract
Deep convection can produce hazardous weather such as heavy rainfall, hail, and tornadoes, while also redistributing heat and moisture through vertical mass transport. Boundary-layer trace gases are transported to the upper troposphere and lower stratosphere (UTLS) within convective updrafts, influencing UTLS chemistry and the radiation budget. Therefore, understanding the processes that control deep convection is essential for predicting both weather hazards and climate impacts. Updraft strength depends on the storm environment and is modulated by entrainment of environmental air into the convective core. However, the influence of entrainment on storm strength and vertical transport remains poorly understood.
This research evaluated the differences in vertical chemical transport and the role of entrainment using parameterized simulations of two convective regimes from the 2012 Deep Convective Clouds and Chemistry (DC3) field campaign: supercell thunderstorm and mesoscale convective system (MCS). The Weather Research and Forecasting model coupled with Chemistry (WRF-Chem, version 4.3.3) was used for all simulations, with lightning data assimilation applied to improve the timing and location of convection. Supercells were more efficient at transporting carbon monoxide (CO) to the UTLS, with 35.3% of transported CO reaching above the simulated lapse-rate tropopause, compared to 23.4% for the MCS. The subgrid contribution of supercell convective transport to the UTLS was greater than that of the MCS, with supercell contributions ranging from ~60% to 80%, while the MCS contributes ~40%. Although entrainment rates varied by storm type, their impact on vertical transport was consistent, enhanced mid-tropospheric entrainment reduced the amount of CO reaching the stratosphere. A mean fractional entrainment rate exceeding 0.1 km-1 at higher altitudes resulted in less subgrid-scale CO transport to the lower stratosphere in both cases.
Entrainment rate in convective parameterizations remains a major source of uncertainty in model simulations. A satellite-based entrainment dataset was developed to identify the environmental influences on entrainment. Entrainment rates were estimated using a simple entraining-plume model with CO measurements from the Tropospheric Emission Spectrometer and Microwave Limb Sounder (TES/MLS), treating CO as a convective constant. Because the satellite-based entrainment dataset was biased toward near-zero values, a proxy dataset was used for the analyses. The entrainment proxy was defined as the difference between the level of neutral buoyancy (LNB) and the level of maximum detrainment (LMD), with LMD identified from upper-tropospheric CO enhancements and environmental variables derived from reanalysis data. Traditional statistical methods were insufficient to isolate the effects of individual environmental parameters; therefore, machine learning (ML) was employed. Interpretability analyses of both Random Forest (RF) and eXtreme Gradient Boosting (XGB) ML models identified entraining convective available potential energy (ECAPE) as the dominant control on entrainment, followed by 0–6 km wind shear. The relationships captured by the model were not dependent on the surface type (land or ocean). Spatially, lifting condensation level height and low-level environmental lapse rate were more important over the sub-Saharan region than any other region in the 35°N/S domain.
ML-derived insights were used to modify the entrainment formulation in the Kain–Fritsch (KF) scheme and assess its impact on vertical chemical transport. Specifically, 0–6 km wind shear was incorporated with a positive relationship to entrainment, while ECAPE and lifting condensation level (LCL) were included with negative relations. The modified formulation was tested across three convective regimes: an airmass thunderstorm, a supercell, and an MCS. Differences between the original and modified entrainment parameterizations were minimal in the airmass case, consistent with its weaker and less organized structure, whereas the more organized systems showed clearer responses. Comparisons with aircraft observations indicated slight improvements in vertical CO transport, particularly in the MCS case, although overall differences remained relatively small. This study highlights the importance of subgrid-scale convective processes in controlling vertical mass transport and provides a framework for improving entrainment parameterization in convection schemes.



Henry Fuelberg

Professor of Meteorology

Director, Undergraduate Meteorology Program

Florida State University

Office Phone 850-644-6466

Cell Phone     850-385-9448
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fsu.edu/pipermail/eoas-seminar/attachments/20260512/edcfee55/attachment.html>