Samantha Makana Turbeville

Stratospheric aerosol injection (SAI) modeling experiments are often designed to keep global mean surface air temperature (GMSAT) close to a reference target, such as early‐century SSP2‐4.5. However, meeting this target does not necessarily restore regional climates or the seasonal and spatial patterns of key variables. We evaluate multiple SAI strategies (ARISE‐SAI‐1.0/1.5, G6‐1.5K‐SAI, and G6‐1.5K‐HiLLA) across four Earth system models (CESM2‐WACCM6, UKESM1, E3SMv3, and MIROC) using a dimensionless, variability‐scaled distance‐to‐target metric applied to the monthly climatological seasonal cycle. We decompose this distance into mean offset, seasonal‐cycle amplitude, and phase/pattern contributions, and use matched baseline‐against‐baseline null distributions to estimate the finite‐sampling floor. SAI substantially reduces temperature distance‐to‐target relative to late‐century SSP2‐4.5, but the majority of SAI model–scenario cases exceed the 97.5th percentile of their matched finite‐sampling null distributions, indicating that residual temperature departures remain detectable even when large‐scale temperature targets are met. These residuals are more strongly associated with differences in seasonal timing than with mean offsets. In contrast, most precipitation distances do not exceed their matched null thresholds, so some apparent precipitation residuals cannot be distinguished from finite‐sampling variability. This framework enables consistent comparisons across intervention strategies, variables, and models without regridding, and identifies where meeting GMSAT targets does or does not restore regional and seasonal climate.

This report details a contribution of Energy Exascale Earth System Model version 3 (E3SMv3) simulations to a proposed model intercomparison project funded and organized by Reflective, a nonprofit group studying the possible global impacts of SAI. Using annual feedback control and seasonally-variable injection sites to achieve a desired global near-surface temperature target, the simulated SAI campaign indicates robust maintenance of the 2020-2039 climatic state over 60 years into the future.

Recent advances in computer modeling have spurred the production of several global storm‐resolving models (GSRMs), which explicitly represent atmospheric circulations from convective to global scales. As a result, GSRMs simulate the formation and evolution of tropical cirrus clouds more physically than typical global climate models/general circulation models (GCMs) which use parameterizations to represent deep convection. We analyze the output from nine GSRMs from the DYAMOND initiative, focusing on the second phase of DYAMOND that simulated a period in January–February 2020. This paper is the third in a series investigating tropical cirrus clouds in GSRMs using DYAMOND model output for an intercomparison. In the tropics, models capture the mean outgoing longwave radiation within −5 to 14 W m⁻² of observed climatology, though most models have more convective precipitation over the 40‐day simulation period than observed. While the models represent large‐scale tropical convection with some fidelity, large regional differences in cloud properties and top‐of‐atmosphere radiation fluxes exist. We focus on a region within the Tropical Western Pacific to study the small‐scale features available with the high spatiotemporal resolution of GSRMs. Most models that participated in both phases of DYAMOND capture the seasonal differences between the two phases, yet each model exhibits unique cloud populations that are persistent across seasons. GSRMs even simulate the notoriously difficult‐to‐observe tropical tropopause layer (TTL) cirrus, providing a novel perspective on TTL cirrus even though the models have different cloud characteristics over the short 40‐days simulation.

The recent advances in computer modeling have spurred the production of several global storm-resolving models (GSRMs), which resolve scales from deep convection to large-scale and global circulations. As a result, GSRMs may simulate the formation and evolution of tropical cirrus clouds more physically than typical GCMs which use parameterizations to represent deep convection. We use nine GSRMs from the DYnamics of the Atmospheric general circulation Modelled On Non-hydrostatic Domains (DYAMOND) initiative (Stevens et al., 2019), with a focus on the second phase of DYAMOND that simulated a period in Jan.-Feb. 2020. In the tropics, models capture the mean OLR within -5 to 14 W/m2 of observed climatology, though most models have a higher OLR and more convective precipitation over the 40-day simulation period than observed. While the models represent large-scale tropical convection with some fidelity, large regional differences in cloud properties and TOA radiation fluxes exist. We focus on a region within the Tropical Western Pacific (TWP) to study the local and small-scale features available with the high spatiotemporal resolution of GSRMs. When including models that participated in both phases of DYAMOND and simulated boreal summer and winter, most models capture the seasonal differences between the two phases. Each model simulates unique cloud characteristics that are persistent across seasons. GSRMs even simulate the notoriously difficult to observe thin TTL cirrus so GSRMs may be a useful tool to enhance our knowledge of TTL cirrus even though the TTL varies greatly between models over the short 40-day simulation.

Pervasive cirrus clouds in the upper troposphere and tropical tropopause layer (TTL) influence the climate by altering the top‐of‐atmosphere radiation balance and stratospheric water vapor budget. These cirrus are often associated with deep convection, which global climate models must parameterize and struggle to accurately simulate. By comparing high‐resolution global storm‐resolving models from the Dynamics of the Atmospheric general circulation Modeled On Non‐hydrostatic Domains (DYAMOND) intercomparison that explicitly simulate deep convection to satellite observations, we assess how well these models simulate deep convection, convectively generated cirrus, and deep convective injection of water into the TTL over representative tropical land and ocean regions. The DYAMOND models simulate deep convective precipitation, organization, and cloud structure fairly well over land and ocean regions, but with clear intermodel differences. All models produce frequent overshooting convection whose strongest updrafts humidify the TTL and are its main source of frozen water. Intermodel differences in cloud properties and convective injection exceed differences between land and ocean regions in each model. We argue that, with further improvements, global storm‐resolving models can better represent tropical cirrus and deep convection in present and future climates than coarser‐resolution climate models. To realize this potential, they must use available observations to perfect their ice microphysics and dynamical flow solvers.

Cirrus clouds of various thicknesses and radiative characteristics extend over much of the tropics, especially around deep convection. They are difficult to observe due to their high altitude and sometimes small optical depths. They are also difficult to simulate in conventional global climate models, which have coarse grid spacings and simplified parameterizations of deep convection and cirrus formation. We investigate the representation of tropical cirrus in global storm‐resolving models (GSRMs), which have higher spatial resolution and explicit convection and could more accurately represent cirrus cloud processes. This study uses GSRMs from the DYnamics of the Atmospheric general circulation Modeled On Non‐hydrostatic Domains (DYAMOND) project. The aggregate life cycle of tropical cirrus is analyzed using joint albedo and outgoing longwave radiation (OLR) histograms to assess the fidelity of models in capturing the observed cirrus cloud populations over representative tropical ocean and land regions. The proportions of optically thick deep convection, anvils, and cirrus vary across models and are portrayed in the vertical distribution of cloud cover and top‐of‐atmosphere radiative fluxes. Model differences in cirrus populations, likely driven by subgrid processes such as ice microphysics, dominate over regional differences between convectively active tropical land and ocean locations.

Pervasive cirrus clouds in the upper troposphere and tropical tropopause layer (TTL) influence the climate by altering the top-of-atmosphere radiation balance and stratospheric water vapor budget. These cirrus are often associated with deep convection, which global climate models must parameterize and struggle to accurately simulate. By comparing high-resolution global storm-resolving models from the Dynamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) intercomparison that explicitly simulate deep convection to satellite observations, we assess how well these models simulate deep convection, convectively generated cirrus, and deep convective injection of water into the TTL over representative tropical land and ocean regions. The DYAMOND models simulate deep convective precipitation, organization, and cloud structure fairly well over land and ocean regions, but with clear intermodel differences. All models produce frequent overshooting convection whose strongest updrafts humidify the TTL and are its main source of frozen water. Inter-model differences in cloud properties and convective injection exceed differences between land and ocean regions in each model. We argue that global storm-resolving models can better represent tropical cirrus and deep convection in present and future climates than coarser-resolution climate models. To realize this potential, they must use available observations to perfect their ice microphysics and dynamical flow solvers.

Cirrus clouds of various thicknesses and radiative characteristics extend over much of the tropics, especially around deep convection. They can be difficult to observe due to their high altitude and sometimes small optical depths. They are also difficult to simulate in conventional global climate models, which have coarse grid spacings and simplified parameterizations of deep convection and cirrus formation. We investigate the representation of tropical cirrus in global storm-resolving models (GSRMs), which have higher spatial resolution and explicit convection and could more accurately represent cirrus cloud processes. This study uses GSRMs from the DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) project. The aggregate life cycle of tropical cirrus is analyzed using joint albedo and outgoing longwave radiation (OLR) histograms to assess the fidelity of models in capturing the observed cirrus cloud populations over representative tropical ocean and land regions. The proportions of optically-thick deep convection, anvils, and cirrus vary across models and are reflected in the vertical distribution of cloud cover and top-of-atmosphere radiative fluxes. Model differences in cirrus populations, likely driven by subgrid processes such as ice microphysics, dominate over regional differences between convectively-active tropical land and ocean locations.

Cirrus clouds of various thicknesses and radiative characteristics extend over much of the tropics, especially around deep convection. They can be difficult to observe due to their high altitude and sometimes small optical depths. They are also difficult to simulate in conventional global climate models, which have coarse grid spacings and simplified parameterizations of deep convection and cirrus formation. We investigate the representation of tropical cirrus in global storm-resolving models (GSRMs), which have higher spatial resolution and explicit convection and could more accurately represent cirrus cloud processes. This study uses GSRMs from the DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) project. The aggregate life cycle of tropical cirrus is analyzed using joint albedo and outgoing longwave radiation (OLR) histograms to assess the fidelity of models in capturing the observed cirrus cloud populations over representative tropical ocean and land regions. The proportions of optically-thick deep convection, anvils, and cirrus vary across models and are reflected in the vertical distribution of cloud cover and top-of-atmosphere radiative fluxes. Model differences in cirrus populations, likely driven by sub-grid processes such as ice microphysics, dominate over regional differences between convectively-active tropical land and ocean locations.