Improving Prediction of Marine Low Clouds Using Cloud Droplet Number Concentration in a Convolutional Neural Network DOI Creative Commons
Yang Cao, Yannian Zhu, Minghuai Wang

и другие.

Journal of Geophysical Research Machine Learning and Computation, Год журнала: 2024, Номер 1(4)

Опубликована: Ноя. 30, 2024

Abstract Marine low clouds play a crucial role in cooling the climate, but accurately predicting them remains challenging due to their highly non‐linear response various factors. Previous studies usually overlook effects of cloud droplet number concentration (N d ) and non‐local information target grids. To address these challenges, we introduce convolutional neural network model (CNN Met‐Nd that uses both local includes N as cloud‐controlling factor enhance predictive ability daily cover, albedo, radiative (CRE) for global marine clouds. CNN demonstrates superior performance, explaining over 70% variance three variables scenes 1° × 1°, notable improvement past efforts. also replicates geographical patterns trends from 2003 2022. In contrast, similar without Met struggles predict long‐term properties effectively. Permutation importance analysis further highlights critical Met‐N 's success. Further comparisons with an artificial (ANN model, which same inputs considering spatial dependence, show performance R 2 values CRE being 0.16, 0.12, 0.18 higher, respectively. This incorporating information, at least on scale, into predictions climate parameterizations.

Язык: Английский

Reducing Aerosol Forcing Uncertainty by Combining Models With Satellite and Within‐The‐Atmosphere Observations: A Three‐Way Street DOI Creative Commons
Ralph A. Kahn, Elisabeth Andrews, C. A. Brock

и другие.

Reviews of Geophysics, Год журнала: 2023, Номер 61(2)

Опубликована: Май 4, 2023

Abstract Aerosol forcing uncertainty represents the largest climate overall. Its magnitude has remained virtually undiminished over past 20 years despite considerable advances in understanding most of key contributing elements. Recent work produced modest increases only confidence estimate itself. This review summarizes contributions toward reducing aerosol made by satellite observations, measurements taken within atmosphere, as well modeling and data assimilation. We adopt a more measurement‐oriented perspective than reviews subject assessing strengths limitations each; gaps possible ways to fill them are considered. Currently planned programs supporting advanced, global‐scale surface‐based aerosol, cloud, precursor gas modeling, intensive field campaigns aimed at characterizing underlying physical chemical processes involved, all essential. But addition, new efforts needed: (a) obtain systematic aircraft situ capturing multi‐variate probability distribution functions particle optical, microphysical, properties (and associated estimates), co‐variability with meteorology, for major airmass types; (b) conceive, develop, implement suborbital (aircraft plus surface‐based) program systematically quantifying cloud‐scale microphysics, cloud optical properties, cloud‐related vertical velocities aerosol‐cloud interactions; (c) focus much research on integrating unique measurements, reduce persistent forcing.

Язык: Английский

Процитировано

26
Lightning declines over shipping lanes following regulation of fuel sulfur emissions DOI Creative Commons

Chris K. Wright,

Joel A. Thornton, Lyatt Jaeglé

и другие.

Atmospheric chemistry and physics, Год журнала: 2025, Номер 25(5), С. 2937 - 2946

Опубликована: Март 11, 2025

Abstract. Aerosol interactions with clouds represent a significant uncertainty in our understanding of the Earth system. Deep convective may respond to aerosol perturbations several ways that have proven difficult elucidate observations. Here, we leverage two busiest maritime shipping lanes world, which emit particles and their precursors into an otherwise relatively clean tropical marine boundary layer, make headway on influence deep clouds. The recent 7-fold change allowable fuel sulfur by International Maritime Organization allows us test sensitivity lightning changes ship plume number-size distributions. We find that, across range atmospheric thermodynamic conditions, previously documented enhancement over has fallen 40 %. is therefore at least partially aerosol-mediated, conclusion supported observations droplet number cloud base, show similar decline lane. These results fundamental implications for aerosol–cloud interactions, suggesting are impacted distribution remote environment.

Язык: Английский

Процитировано

2
Aerosol trends dominate over global warming-induced cloud feedback in driving recent changes in marine low clouds DOI Creative Commons
Yang Cao, Hao Wang, Yannian Zhu

и другие.

Research Square (Research Square), Год журнала: 2025, Номер unknown

Опубликована: Фев. 13, 2025

Abstract Over the past two decades, anthropogenic emission reductions and global warming have impacted marine low clouds through aerosol-cloud interactions (ACI) cloud feedback, yet their quantitative contributions remain unclear. This study employs a deep learning model (CNNMet−Nd) Community Earth System Model version 2 (CESM2) to disentangle these effects. CNNMet−Nd reveals that aerosol-driven changes in droplet number concentration dominate near-global shortwave radiative effect (ΔCRE), contributing 0.42 ± 0.08 Wm⁻² per 20 years, compared 0.05 0.37 from feedback. CESM2 effectively reproduces predominant influence of aerosol on ΔCRE by CNNMet−Nd, lending us confidence for stronger estimate effective forcing due ACI (ERFaci) -1.29 since preindustrial era. These findings highlight critical role shaping trends its broader climate implications, especially under ongoing reduction efforts.

Язык: Английский

Процитировано

0
Improving prediction of marine low clouds using cloud droplet number concentration in a convolutional neural network DOI
Yang Cao, Yannian Zhu, Minghuai Wang

и другие.

Authorea (Authorea), Год журнала: 2024, Номер unknown

Опубликована: Июль 17, 2024

Marine low clouds significantly cool the climate, but predicting these remains challenging: response of to various factors is highly non-linear. Previous studies usually overlook effects cloud droplet number concentration (Nd) and non-local information target grids. To address challenges, we introduce a convolutional neural network model (CNNMet-Nd) that uses both local includes Nd as cloud-controlling factor enhance predictive ability cover, albedo, radiative (CRE) for global marine clouds. CNNMet-Nd demonstrates superior performance, explaining over 70% variance in three variables instantaneous scenes 1°×1°, notable improvement past efforts. also accurately replicates geographical patterns trends from 2003 2022. In contrast, similar without input (CNNMet) fails predict mean properties effectively, underscoring critical role Nd. Further comparisons with an artificial (ANNMet-Nd) model, which same inputs considering spatial dependence, show CNNMet-Nd's performance R2 values CRE being 0.16, 0.11, 0.18 higher, respectively. This highlights importance incorporating into predictions climate parameterizations.

Язык: Английский

Процитировано

2
Aggressive Aerosol Mitigation Policies Reduce Chances of Keeping Global Warming to Below 2C DOI Creative Commons
Robert Wood, Mika Vogt, Isabel L. McCoy

и другие.

Earth s Future, Год журнала: 2024, Номер 12(7)

Опубликована: Июль 1, 2024

Abstract Aerosol increases over the 20th century delayed rate at which Earth warmed as a result of in greenhouse gases (GHGs). Aggressive aerosol mitigation policies arrested radiative forcing from ∼1980 to ∼2010. Recent evidence supports decreases magnitude since then. Using approximate partial perturbation (APRP) method, future shortwave effective changes are isolated other an 18‐member ensemble ScenarioMIP projections phase 6 Coupled Model Intercomparison Project (CMIP6). APRP‐derived near‐term (2020–2050) trends correlated with published model emulation values but 30%–50% weaker. Differences likely explained by location shifts aerosol‐impacting emissions and their resultant influences on susceptible clouds. Despite weaker changes, implementation aggressive cleanup will have major impact global warming rates 2020–2050. forcings used together impulse response estimate temperature trends. Strong GHGs, SSP1‐2.6, prevents exceeding 2C preindustrial strong this scenario probability 2050 near zero without 6% cleanup. When same is applied more GHG (i.e., SSP2‐4.5), than doubles reaching 30% 80%. It thus critical quantify simulate impacts next few decades.

Язык: Английский

Процитировано

1
Aggressive aerosol mitigation policies reduce chances of keeping global warming to below 2C DOI Open Access
Robert Wood, Mika Vogt, Isabel L. McCoy

и другие.

Authorea (Authorea), Год журнала: 2023, Номер unknown

Опубликована: Ноя. 9, 2023

Aerosol increases over the 20th century delayed rate at which Earth warmed as a result of in greenhouse gases (GHGs). Aggressive aerosol mitigation policies arrested radiative forcing from ~1980 to ~2010. Recent evidence supports decreases magnitude since then. Using approximate partial perturbation (APRP) method, future shortwave effective changes are isolated other an 18-member ensemble ScenarioMIP projections phase 6 Coupled Model Intercomparison Project (CMIP6). APRP-derived near-term (2020-2050) trends correlated with published model emulation values but 30-50% weaker. Differences likely explained by location shifts aerosol-impacting emissions and their resultant influences on susceptible clouds. Despite weaker changes, implementation aggressive cleanup will have major impact global warming rates 2020-2050. forcings used together impulse response estimate temperature trends. Strong GHGs, SSP1-2.6, prevents exceeding 2C preindustrial strong this scenario probability 2050 near zero without 6% cleanup. When same is applied more GHG (i.e., SSP2-4.5), than doubles reaching 30% 80%. It thus critical quantify simulate impacts next few decades.

Язык: Английский

Процитировано

1
Improving Prediction of Marine Low Clouds Using Cloud Droplet Number Concentration in a Convolutional Neural Network DOI Creative Commons
Yang Cao, Yannian Zhu, Minghuai Wang

и другие.

Journal of Geophysical Research Machine Learning and Computation, Год журнала: 2024, Номер 1(4)

Опубликована: Ноя. 30, 2024

Abstract Marine low clouds play a crucial role in cooling the climate, but accurately predicting them remains challenging due to their highly non‐linear response various factors. Previous studies usually overlook effects of cloud droplet number concentration (N d ) and non‐local information target grids. To address these challenges, we introduce convolutional neural network model (CNN Met‐Nd that uses both local includes N as cloud‐controlling factor enhance predictive ability daily cover, albedo, radiative (CRE) for global marine clouds. CNN demonstrates superior performance, explaining over 70% variance three variables scenes 1° × 1°, notable improvement past efforts. also replicates geographical patterns trends from 2003 2022. In contrast, similar without Met struggles predict long‐term properties effectively. Permutation importance analysis further highlights critical Met‐N 's success. Further comparisons with an artificial (ANN model, which same inputs considering spatial dependence, show performance R 2 values CRE being 0.16, 0.12, 0.18 higher, respectively. This incorporating information, at least on scale, into predictions climate parameterizations.

Язык: Английский

Процитировано

0