Thesis-Modeling Aerosol Transport for Stratospheric Solar Geoengineering: from Particle to Plume Scale
10 views
Skip to first unread message
Geoengineering News
Jun 13, 2023, 7:26:56 PM6/13/23
to geoengi...@googlegroups.com
https://dash.harvard.edu/handle/1/37375743
We use a Lagrangian trajectory model driven by reanalyzed stratospheric winds and modified to include sedimentation to model the transport of each injected particle (from SAI) in the stratosphere and quantify the sensitivity of particle lifetime to injection locations. From the physical perspective, we analyze how background circulations influence particle transport and lifetime in the stratosphere by considering Brewer-Dobson Circulation, Quasi-Biennial Oscillation, tropopause height, poleward winds, etc. From the engineering perspective, we explore various SAI injection strategies to increase particle lifetime in the stratosphere. For example, we find that an optimal choice of injection locations can increase particle lifetime by 44% at 20 km, compared to without choosing injection locations.
SAI would almost certainly use aircraft for deployment, and these aircraft would produce line-shaped plumes with strong concentration gradients, which are hard for the global Eulerian model to resolve. To help global Eulerian models resolve subgrid plumes in the stratosphere, a Lagrangian plume model, comprising a Lagrangian trajectory model and an adaptive-grid plume model with a sequence of plume cross-section representations (from a highly resolved 2-D grid to a simplified 1-D grid based on a tradeoff between the accuracy and computational cost), is created and embedded into a global Eulerian model (i.e., GEOS-Chem model) to establish a multiscale Plume-in-Grid (PiG) model. We compare this PiG model (with plume model) to the GEOS-Chem model (without plume model) based on a 1-month simulation of continuous inert tracer emissions by aircraft in the stratosphere in several aspects including trace concentration, trace mixing (entropy), nonlinear processes, and computing efficiency. For example, with the plume model, the final injected tracer is more concentrated and approximately 1/3 of the tracer is at concentrations 2-4 orders of magnitude larger compared to without the plume model.
Author
Sun, Hongwei
2023
Citation
Sun, Hongwei. 2023. Modeling Aerosol Transport for Stratospheric Solar Geoengineering: from Particle to Plume Scale. Doctoral dissertation, Harvard University Graduate School of Arts and Sciences.
Abstract
Stratospheric Aerosol Injection (SAI) aims to offset some climate hazards by releasing aerosols into the stratosphere to reflect solar radiation. In this thesis, we outline the use of new modeling methods to simulate aerosol transport in the stratosphere for SAI from particle scale (Lagrangian trajectory model) to plume scale (plume-in-grid model).We use a Lagrangian trajectory model driven by reanalyzed stratospheric winds and modified to include sedimentation to model the transport of each injected particle (from SAI) in the stratosphere and quantify the sensitivity of particle lifetime to injection locations. From the physical perspective, we analyze how background circulations influence particle transport and lifetime in the stratosphere by considering Brewer-Dobson Circulation, Quasi-Biennial Oscillation, tropopause height, poleward winds, etc. From the engineering perspective, we explore various SAI injection strategies to increase particle lifetime in the stratosphere. For example, we find that an optimal choice of injection locations can increase particle lifetime by 44% at 20 km, compared to without choosing injection locations.
SAI would almost certainly use aircraft for deployment, and these aircraft would produce line-shaped plumes with strong concentration gradients, which are hard for the global Eulerian model to resolve. To help global Eulerian models resolve subgrid plumes in the stratosphere, a Lagrangian plume model, comprising a Lagrangian trajectory model and an adaptive-grid plume model with a sequence of plume cross-section representations (from a highly resolved 2-D grid to a simplified 1-D grid based on a tradeoff between the accuracy and computational cost), is created and embedded into a global Eulerian model (i.e., GEOS-Chem model) to establish a multiscale Plume-in-Grid (PiG) model. We compare this PiG model (with plume model) to the GEOS-Chem model (without plume model) based on a 1-month simulation of continuous inert tracer emissions by aircraft in the stratosphere in several aspects including trace concentration, trace mixing (entropy), nonlinear processes, and computing efficiency. For example, with the plume model, the final injected tracer is more concentrated and approximately 1/3 of the tracer is at concentrations 2-4 orders of magnitude larger compared to without the plume model.
Source: Digital Access to Scholarship at Harvard
Reply all
Reply to author
Forward
0 new messages