Journal of Advances in Modeling Earth Systems最新文献

Ocean turbulence parameterization has principally been based on process-based approaches, seeking to embed physical principles so that coarser resolution calculations can capture the net influence of smaller scale unresolved processes. More recently there has been an increasing focus on the application of data-driven approaches to this problem. Here we consider the application of online learning to data-driven eddy parameterization, constructing an end-to-end automatically differentiable dynamical solver forced by a neural network (NN), and training the NN based on the dynamics of the combined hybrid system. This approach is applied to the classic barotropic Stommel-Munk gyre problem—a highly idealized configuration which nevertheless includes multiple flow regimes, boundary dynamics, and a separating jet, and therefore presents a challenging test case for the online learning approach. It is found that a NN which is suitably trained can lead to a coarse resolution NN parameterized model which is stable, and has both a reasonable mean state and intrinsic variability. This suggests that online learning is a powerful tool for studying the problem of ocean turbulence parameterization. A test of generalizability with a modified wind forcing shows some positive results. However a test of symmetry preservation demonstrates that the NN parameterized model fails to respect an intrinsic symmetry property of the underlying system.

The accuracy of global forecasting of atmospheric composition is essential for protecting public health and advancing climate research. Carbon monoxide (CO), a key pollutant with indirect greenhouse effects, requires timely and accurate prediction. We propose Duo-AttnOPNets, an operational framework that combines a deep learning-based forecasting module Duo-AttnForeNet with a training-free data assimilation module Duo-AttnVarNet. Duo-AttnForeNet employs CSLSTM blocks—built from ConvLSTM cells and dual attention mechanisms—to effectively capture spatio-temporal dynamics. Duo-AttnVarNet leverages automatic differentiation and GPU acceleration to enable efficient four-dimensional variational (4D-Var) assimilation without extensive manual adjoint coding. We evaluate Duo-AttnOPNets against the Integrated Forecasting System for Atmospheric Composition (C-IFS). Results show that Duo-AttnOPNets achieves comparable or superior accuracy in both forecasting and assimilation, while generating 5-day forecasts within seconds on a single GPU. These findings demonstrate its potential for real-time, scalable, and accurate CO forecasting, marking a promising advance in integrating deep learning with traditional variational methods for operational atmospheric modeling.

Aerosol-cloud interactions (aci) are the leading source of uncertainty in inferring climate sensitivity from the historical record. Earth system models (ESMs) struggle to represent aci because the processes responsible for these phenomena occur at much finer time and space scales than can be resolved by any ESM. Observational constraints provide key benchmarks to test ESMs, but cannot be used alone to fully understand aci processes except in very specific cases where causality is controlled; some degree of modeling is required to infer aci and estimate radiative forcing. Here, we generate and characterize a perturbed parameter ensemble (PPE) in version 3 of the Energy Exascale ESM (E3SMv3). We perturb 25 parameters that govern aci processes over 250 members and integrate the model over present-day and preindustrial aerosol emissions. We find that the process representation in E3SMv3 is flexible and can generate global-mean effective radiative forcings due to aci (ERFaci) ranging from −3.0 to +0.9 W m⁻². The positive ERFaci values simulated by a portion of the PPE are implausible and result from parameter combinations that produce unrealistic top-of-atmosphere energy fluxes. While global-mean cloud droplet number concentration always increases in response to anthropogenic aerosol, cloud liquid water path can both increase and decrease, suggesting that precipitation suppression is not the only aerosol-cloud adjustment represented by E3SMv3. Analysis of which processes control liquid cloud adjustment in the PPE points toward stratiform precipitation processes and aerosol activation, which is consistent with many previous ESMs, as well as the new two-moment convective cloud microphysics in E3SMv3.

Diapycnal mixing in the ocean interior has diverse numerical representations in current global ocean models. These representations affect the simulated transport and storage of oceanic tracers in ways that remain little studied. Here we present the impacts of three different tidal mixing representations in thousand-year-long simulations with the NEMO global ocean model at one-degree resolution. The first model experiment includes local bottom-intensified mixing at internal tide generation sites and a constant background diffusivity. The second explicitly includes both local and remote tidal mixing, with no background diffusivity. The third experiment is identical to the second but has the added contribution of bottom-trapped (subinertial) internal tides, known to be important in polar regions. The three simulations show broadly similar circulation and stratification but important regional differences. Explicit representation of remote tidal mixing strengthens the Atlantic Meridional Overturning Circulation by up to 1.5 × 10⁶ m³ s⁻¹. Inclusion of bottom-trapped internal tides reduces heat reaching Antarctica by eroding Circumpolar Deep Water at southern high latitudes, and reduces the mean age of the global deep (>2 km) ocean by 10%. The results call for more observational constraints on polar ocean mixing, and point to multi-faceted climatic repercussions of tidal mixing representations.

Parameterizing radiative transfer in climate models means navigating trade-offs between physical accuracy and conceptual clarity. However, currently available schemes sit at the extremes of this spectrum: correlated-k schemes are fast and accurate but rely on lookup tables which obscure the underlying physics and make such schemes difficult to modify, while gray radiation schemes are conceptually straightforward but introduce significant biases in atmospheric circulation. Here we introduce a Simple Spectral Model (SSM) for clear-sky longwave radiative transfer which bridges this “clarity-accuracy” gap. The SSM accomplishes this by representing the spectral structure of $��H_2�$O and C$��O_2�$ absorption using analytic fits at reference conditions, then uses simple functional forms to extend these fits to different atmospheric conditions. This, coupled to a simple, two-stream solver, yields a system of six equations and ten physically meaningful parameters which can solve for clear-sky longwave fluxes given atmospheric profiles of temperature and humidity. When implemented in an idealized aquaplanet GCM, the SSM produces zonal-mean climate states which accurately mimic the results using a benchmark correlated-k code. The SSM also alleviates the significant zonal-mean climate biases associated with using gray radiation, including an improved representation of radiative cooling profiles, tropopause structure, jet dynamics, and Hadley Cell characteristics both in control climates and in response to uniform warming. This work demonstrates that even a simple spectral representation of atmospheric absorption suffices to capture the essential physics of longwave radiative transfer. The SSM promises to be a valuable tool both for idealized climate modeling, and for teaching radiative transfer in the classroom.

Our goal in this study is to characterize the relationship between lower tropospheric environmental humidity and convective mass flux in the tropics. To do so, we have created gridded convective mass flux data sets from five global storm-resolving models (GSRMs). We have three principal findings. First, in humid environments, mass flux increases with height from the surface through the depth of the lower free troposphere, forming a “deep-inflow.” In dry environments, mass flux does not increase with height in the lower free troposphere. Second, mid-tropospheric mass flux increases nonlinearly with increasing lower tropospheric humidity, resembling a widely reported pickup in tropical precipitation. Third, increased lower tropospheric humidity is associated with reduced updraft buoyancy. To interpret these findings, we employ a simple three-equation parcel model with stochastic entrainment. The parcel model suggests that the response of convective mass flux to lower tropospheric humidity is governed by two effects: (a) survival, in which a greater share of entraining parcels ascend rather than detrain with greater humidity; and (b) dilution, in which the average entrainment rate among surviving parcels increases with environmental humidity. Together, survival and dilution account for the three mass flux responses to humidity.

Agrivoltaics, combining agriculture with photovoltaic systems, offers a promising solution to address land-use conflict between food and energy production. However, the complexities of agrivoltaics and its effects on the water-energy-carbon interactions remain poorly understood. In this study, we developed a process-based agrivoltaic model within the Community Land model 5 to assess the impacts of agrivoltaics on water, energy, and carbon cycles. The model was validated using data from agrivoltaic sites in Illinois and Colorado, generally capturing spatiotemporal variations in light conditions, soil moisture, and biomass carbon. Simulation results suggest that agrivoltaics significantly impact water, energy, and carbon budgets at the patch and system levels for maize and soybean in Illinois and grass in Colorado (2000–2014). Our findings show that the impacts of agrivoltaics vary by climate conditions and plant types. In dry climates, rainfall redistribution and shading from agrivoltaics conserve soil moisture and enhance evapotranspiration, promoting greater carbon assimilation and soil carbon storage for C₃ grass. Conversely, in wetter regions, reduced solar radiation from shading becomes the dominant factor, lowering carbon assimilation and sequestration for maize and soybean. These results suggest that agrivoltaics can help mitigate drought impacts in arid environments. Our analysis of land equivalent ratios across different photovoltaic ground coverage ratios (PV GCR) shows that a medium PV GCR (60%) under “AgPV” deployment, where PV and plants share the same land, maximizes land-use efficiency at the study sites. Our modeling study supports informed decision-making to promote sustainable management of water, energy, and food resources amid environmental change.

A new signal simulator for Doppler velocity, derived for W-band cloud profiling radar onboard EarthCARE, is developed and implemented into the Cloud Feedback Model Intercomparison Project Observation Simulator Package version 2 (COSP2). The simulator converts the vertical motion of hydrometeors and cumulus mass flux in global climate models (GCMs) into Doppler velocity signals, providing statistics on Doppler velocity and its spectrum width in the form of Contoured Frequency by tEmperature Diagram (CFED) or by Altitude Diagram (CFAD). To account for the different treatments of vertical air motion in stratiform and convective clouds within GCMs, their statistics are processed separately. The simulator was tested on the MIROC6 GCM and compared with ground-based radar measurements. The results showed consistency in ice particle growth and melting between the model and the observations. However, the droplet fall speed in the model was quantitatively underestimated, revealing a bias in the cloud microphysics of MIROC6. The combined use of this simulator with calculation of cloud optical depth in COSP2 also allows for the investigation of Doppler velocity characteristics as a function of cloud type. The developed simulator enabled COSP2 to generate model diagnostics of cloud particles and cumulus vertical air motions, facilitating future global comparisons with EarthCARE data. The enhanced capabilities of COSP2 thus will add value to model evaluation through the combined use of multiple simulators and multi-sensor synergistic observations provided by EarthCARE.

Motivated by the need to improve the representation of small-scale surface heterogeneity in Earth System Models (ESMs), new algorithms have been introduced to discretize ESM computational units (CUs) into a variable number of subgrid topographic units for improving model simulations with minimal increase in computational demand. The algorithms can be applied to structured (regular grid) and unstructured (e.g., watersheds) CUs to derive topography-based subgrid units (TGUs). This study evaluates the capability of the TGUs to capture surface heterogeneity within grid- versus watershed-based CUs. For this purpose, TGUs are derived for the grid- and watershed-based CUs at four equivalent spatial scales (1°, 0.5°, 0.25°, and 0.125° for grid-based and Hydrologic Unit Code levels HUC07, HUC08, HUC09, and HUC10 for watershed-based) over the CONUS domain. Statistical metrics are computed at the CU and TGU levels at each spatial scale for comparison. Results show that compared to the grid-based TGUs, the watershed-based TGUs are superior in capturing spatial heterogeneity associated with topographic slope, land cover, and surface hydrometeorology, despite their similar capability in capturing topographic elevation. This improved capability of the watershed-based TGUs resulting from the combined effects of the CU and TGU level discretization is consistently found across all spatial scales examined. At the finest spatial scales (0.125° and HUC10), the watershed-based TGUs better capture the observed precipitation, temperature, and snow water equivalent than the grid-based TGUs at 94%, 84%, and 72% of the SNOwpack TELemetry sites, respectively, highlighting the potential advantage of the watershed-based TGUs for improving accuracy and realism in ESM simulations.