Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation

Abstract Parameterization of subgrid‐scale processes is a major source of uncertainty in global atmospheric model simulations. Global storm‐resolving simulations use a finer grid (less than 5 km) to reduce this uncertainty by explicitly resolving deep convection and details of orography. This study...

Full description

Saved in:
Bibliographic Details
Main Authors: Oliver Watt‐Meyer, Noah D. Brenowitz, Spencer K. Clark, Brian Henn, Anna Kwa, Jeremy McGibbon, W. Andre Perkins, Lucas Harris, Christopher S. Bretherton
Format: Article
Language:English
Published: American Geophysical Union (AGU) 2024-02-01
Series:Journal of Advances in Modeling Earth Systems
Subjects:
Online Access:https://doi.org/10.1029/2023MS003668
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850169932125306880
author Oliver Watt‐Meyer
Noah D. Brenowitz
Spencer K. Clark
Brian Henn
Anna Kwa
Jeremy McGibbon
W. Andre Perkins
Lucas Harris
Christopher S. Bretherton
author_facet Oliver Watt‐Meyer
Noah D. Brenowitz
Spencer K. Clark
Brian Henn
Anna Kwa
Jeremy McGibbon
W. Andre Perkins
Lucas Harris
Christopher S. Bretherton
author_sort Oliver Watt‐Meyer
collection DOAJ
description Abstract Parameterization of subgrid‐scale processes is a major source of uncertainty in global atmospheric model simulations. Global storm‐resolving simulations use a finer grid (less than 5 km) to reduce this uncertainty by explicitly resolving deep convection and details of orography. This study uses machine learning to replace the physical parameterizations of heating and moistening rates, but not wind tendencies, in a coarse‐grid (200 km) global atmosphere model, using training data obtained by spatially coarse‐graining a 40‐day realistic geography global storm‐resolving simulation. The training targets are the three‐dimensional fields of effective heating and moistening rates, including the effect of grid‐scale motions that are resolved but imperfectly simulated by the coarse model. A neural network is trained to predict the time‐dependent heating and moistening rates in each grid column using the coarse‐grained temperature, specific humidity, surface turbulent heat fluxes, cosine of solar zenith angle, land‐sea mask and surface geopotential of that grid column as inputs. The coefficient of determination R2 for offline prediction ranges from 0.4 to 0.8 at most vertical levels and latitudes. Online, we achieve stable 35‐day simulations, with metrics of skill such as the time‐mean pattern of near‐surface temperature and precipitation comparable or slightly better than a baseline simulation with conventional physical parameterizations. However, the structure of tropical circulation and relative humidity in the upper troposphere are unrealistic. Overall, this study shows potential for the replacement of human‐designed parameterizations with data‐driven ones in a realistic setting.
format Article
id doaj-art-e8eeeac9be0a41338c896fabcc34c313
institution OA Journals
issn 1942-2466
language English
publishDate 2024-02-01
publisher American Geophysical Union (AGU)
record_format Article
series Journal of Advances in Modeling Earth Systems
spelling doaj-art-e8eeeac9be0a41338c896fabcc34c3132025-08-20T02:20:37ZengAmerican Geophysical Union (AGU)Journal of Advances in Modeling Earth Systems1942-24662024-02-01162n/an/a10.1029/2023MS003668Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving SimulationOliver Watt‐Meyer0Noah D. Brenowitz1Spencer K. Clark2Brian Henn3Anna Kwa4Jeremy McGibbon5W. Andre Perkins6Lucas Harris7Christopher S. Bretherton8Allen Institute for Artificial Intelligence Seattle WA USANVIDIA Corporation Santa Clara CA USAAllen Institute for Artificial Intelligence Seattle WA USAAllen Institute for Artificial Intelligence Seattle WA USAAllen Institute for Artificial Intelligence Seattle WA USAAllen Institute for Artificial Intelligence Seattle WA USAAllen Institute for Artificial Intelligence Seattle WA USAGeophysical Fluid Dynamics Laboratory NOAA Princeton NJ USAAllen Institute for Artificial Intelligence Seattle WA USAAbstract Parameterization of subgrid‐scale processes is a major source of uncertainty in global atmospheric model simulations. Global storm‐resolving simulations use a finer grid (less than 5 km) to reduce this uncertainty by explicitly resolving deep convection and details of orography. This study uses machine learning to replace the physical parameterizations of heating and moistening rates, but not wind tendencies, in a coarse‐grid (200 km) global atmosphere model, using training data obtained by spatially coarse‐graining a 40‐day realistic geography global storm‐resolving simulation. The training targets are the three‐dimensional fields of effective heating and moistening rates, including the effect of grid‐scale motions that are resolved but imperfectly simulated by the coarse model. A neural network is trained to predict the time‐dependent heating and moistening rates in each grid column using the coarse‐grained temperature, specific humidity, surface turbulent heat fluxes, cosine of solar zenith angle, land‐sea mask and surface geopotential of that grid column as inputs. The coefficient of determination R2 for offline prediction ranges from 0.4 to 0.8 at most vertical levels and latitudes. Online, we achieve stable 35‐day simulations, with metrics of skill such as the time‐mean pattern of near‐surface temperature and precipitation comparable or slightly better than a baseline simulation with conventional physical parameterizations. However, the structure of tropical circulation and relative humidity in the upper troposphere are unrealistic. Overall, this study shows potential for the replacement of human‐designed parameterizations with data‐driven ones in a realistic setting.https://doi.org/10.1029/2023MS003668machine learningparameterizationatmospheric modelingglobal storm‐resolving simulations
spellingShingle Oliver Watt‐Meyer
Noah D. Brenowitz
Spencer K. Clark
Brian Henn
Anna Kwa
Jeremy McGibbon
W. Andre Perkins
Lucas Harris
Christopher S. Bretherton
Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
Journal of Advances in Modeling Earth Systems
machine learning
parameterization
atmospheric modeling
global storm‐resolving simulations
title Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
title_full Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
title_fullStr Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
title_full_unstemmed Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
title_short Neural Network Parameterization of Subgrid‐Scale Physics From a Realistic Geography Global Storm‐Resolving Simulation
title_sort neural network parameterization of subgrid scale physics from a realistic geography global storm resolving simulation
topic machine learning
parameterization
atmospheric modeling
global storm‐resolving simulations
url https://doi.org/10.1029/2023MS003668
work_keys_str_mv AT oliverwattmeyer neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT noahdbrenowitz neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT spencerkclark neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT brianhenn neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT annakwa neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT jeremymcgibbon neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT wandreperkins neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT lucasharris neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation
AT christophersbretherton neuralnetworkparameterizationofsubgridscalephysicsfromarealisticgeographyglobalstormresolvingsimulation