Journal of Advances in Modeling Earth Systems最新文献

A solid understanding of the mechanisms behind the presently observed, rapid warming of the northwest North Atlantic Continental Shelf and their biogeochemical impacts is lacking. We hypothesize that a weakening of the Labrador Current System (LCS), especially the shelfbreak jet along the Scotian Shelf, is contributing to these changes and that the future evolution of the LCS will be key to accurate projections. Here we analyze the response of a transient simulation of the high-resolution GFDL Climate Model 2.6 (CM2.6) which realistically simulates the regional circulation but includes only a highly simplified representation of ocean biogeochemistry. Then, we use the CM2.6 to force a medium-complexity regional biogeochemical ocean model, the Atlantic Canada Model, to obtain projections of nutrient availability on the shelf. In the simulation, the shelfbreak jet weakens because of a reduction of the along-shelf pressure gradient caused by a buoyancy gain of the upper water column along the shelf edge. This buoyancy gain is the result of warmer waters along the continental slope. Importantly, we find that the temperature-based criterion used commonly to pinpoint the location of the Gulf Stream is misleading, causing an overestimation of the northward migration of the Gulf Stream. A fixed isotherm may indicate northward movement as a result of basin-wide warming and not necessarily reflect changes in dynamics. The combination of the weakened shelfbreak jet and a lowering of nutrient concentrations in its source water reduce nutrient availability on the northwest North Atlantic shelf by one third by 2100 in the projection analyzed.

All Earth System Models (ESMs) have climatological biases relative to the observed historical climate. The quality of a model and, more importantly, the accuracy of its predictions are often associated with the magnitude and properties of its biases. For more than a decade, new strategies have been developed to empirically reduce such biases in the model components of ESMs during their execution. The present study considers a cyclostationary class of empirical runtime bias corrections to a climate model, referred to here as empirical runtime bias corrections (ERBCs). Such ERBCs are state independent and designed to reduce biases in the climatological annual cycle of the model. We present a new procedure for deriving such ERBCs called Climatological Adaptive Bias Correction (CABCOR). CABCOR is argued to be superior to the standard relaxation approach to defining ERBCs because it requires only a climatological, rather than a multi-year time evolving, observational reference data set. As part of this study, we perform a novel analysis of the relaxation approach in which a mapping is made between the parameter values that define the relaxation and the biases produced by ERBCs in the corrected model. This allows us to identify the optimal bias correction produced by the relaxation approach and to additionally demonstrate that the CABCOR approach can produce bias-corrected models with smaller climatological biases.

Cloud radiative feedback impacts early tropical cyclone (TC) intensification, but limitations in existing diagnostic frameworks make them unsuitable for studying asymmetric or transient radiative heating. We propose a linear Variational Encoder-Decoder (VED) framework to learn the hidden relationship between radiative anomalies and the surface intensification of realistic simulated TCs. The uncertainty of the VED model identifies periods when radiation has more importance for intensification. A close examination of the radiative pattern extracted by the VED model from a 20-member ensemble simulation on Typhoon Haiyan shows that longwave forcing from inner core deep convection and shallow clouds downshear contribute to intensification, with deep convection in the downshear-left quadrant having the most impact overall on the intensification of that TC. Our work demonstrates that machine learning can aid the discovery of thermodynamic-kinematic relationships without relying on axisymmetric or deterministic assumptions, paving the way for the objective discovery of processes leading to TC intensification in realistic conditions.

Terrestrial biosphere models offer a comprehensive view of the global carbon cycle by integrating ecological processes across scales, yet they introduce significant uncertainties in climate and biogeochemical projections due to diverse process representations and parameter variations. For instance, different soil water limitation functions lead to wide productivity ranges across models. To address this, we propose the Differentiable Land Model (DifferLand), a novel hybrid machine learning approach replacing unknown water limitation functions in models with neural networks (NNs) to learn from data. Using automatic differentiation, we calibrated the embedded NN and the physical model parameters against daily observations of evapotranspiration, gross primary productivity, ecosystem respiration, and leaf area index across 16 FLUXNET sites. We evaluated six model configurations where NNs simulate increasingly complex soil water and photosynthesis interactions against test data sets to find the optimal structure-performance tradeoff. Our findings show that a simple hybrid model with a univariate NN effectively captures site-level water and carbon fluxes on a monthly timescale. Across a global aridity gradient, the magnitude of water stress limitation varies, but its functional form consistently converges to a piecewise linear relationship with saturation at high water levels. While models incorporating more interactions between soil water and meteorological drivers better fit observations at finer time scales, they risk overfitting and equifinality issues. Our study demonstrates that hybrid models have great potential in learning unknown parameterizations and testing ecological hypotheses. Nevertheless, careful structure-performance tradeoffs are warranted in light of observational constraints to translate the retrieved relationships into robust process understanding.

Emissions of anthropogenic aerosols are rapidly changing, in amounts, composition and geographical distribution. In East and South Asia in particular, strong aerosol trends combined with high population densities imply high potential vulnerability to climate change. Improved knowledge of how near-term climate and weather influences these changes is urgently needed, to allow for better-informed adaptation strategies. To understand and decompose the local and remote climate impacts of regional aerosol emission changes, we perform a set of Systematic Regional Aerosol Perturbations (SyRAP) using the reduced-complexity climate model FORTE 2.0 (FORTE2). Absorbing and scattering aerosols are perturbed separately, over East Asia and South Asia, to assess their distinct influences on climate. In this paper, we first present an updated version of FORTE2, which includes treatment of aerosol-cloud interactions. We then document and validate the local responses over a range of parameters, showing for instance that removing emissions of absorbing aerosols over both East Asia and South Asia is projected to cause a local drying, alongside a range of more widespread effects. We find that SyRAP-FORTE2 is able to reproduce the responses to Asian aerosol changes documented in the literature, and that it can help us decompose regional climate impacts of aerosols from the two regions. Finally, we show how SyRAP-FORTE2 has regionally linear responses in temperature and precipitation and can be used as input to emulators and tunable simple climate models, and as a ready-made tool for projecting the local and remote effects of near-term changes in Asian aerosol emissions.

This study showcases a global, heterogeneously coupled total water level system wherein salinity and temperature outputs from a coarser-resolution ($��\sim �$12 km) ocean general circulation model are used to calculate density-driven terms within a global, higher-resolution ($��\sim �$2.5 km) depth-averaged total water level model. We demonstrate that the inclusion of baroclinic forcing in the barotropic model requires modification of the internal wave drag term to prevent excess degradation of tidal results compared to the barotropic model. By scaling the internal tide dissipation by an easy to calculate dissipation ratio, the resulting heterogeneously coupled model has complex root mean square errors (RMSE) of 2.27 cm in the deep ocean and 12.16 cm in shallow waters for the $��M_2�$ tidal constituent. While this represents a 10%–20% deterioration as compared to the barotropic model, the improvements in total water level prediction more than offset this degradation. Global median RMSE compared to observations of total water levels, 30-day sea levels, and non-tidal residuals improve by 1.86 (18.5%), 2.55 (42.5%), and 0.36 (5.3%) cm respectively. The drastic improvement in model performance highlights the importance of including density-driven effects within global hydrodynamic models and will help to improve the results of both hindcasts and forecasts in modeling extreme and nuisance flooding. With only an 11% increase in model run time compared to the fully barotropic total water level model, this approach paves the way for high resolution coastal water level and flood models to be used alongside climate models, improving operational forecasting of total water levels.

Implicit time-stepping for advection is applied locally in space and time where Courant numbers are large, but standard explicit time-stepping is used for the remaining solution which is typically the majority. This adaptively implicit advection scheme facilitates efficient and robust integrations with long time-steps while having negligible impact on the overall accuracy, and achieving monotonicity and local conservation on general meshes. A novel and important aspect for the efficiency of the approach is that only one iteration is needed each time the linear equation solver is called for solving the advection equation. The demonstration in this paper uses the second-order Runge-Kutta implicit/explicit time integration in combination with a second/third-order finite-volume spatial discretization and is tested using deformation flow tracer advection on the sphere and a fully compressible model for atmospheric flows. Tracers are advected over the poles of highly anisotropic latitude-longitude grids with very large Courant numbers and on quasi-uniform hexagonal and cubed-sphere meshes with the same algorithm. Buoyant flow simulations with strong local updrafts also benefit from adaptively implicit advection. Stably stratified compressible flow simulations require a stable combination of implicit treatment of gravity and acoustic waves as well as advection in order to achieve long time-steps.

Bias correction (BC) of the cloud-affected infrared (IR) radiances is one of the most difficult challenges in the all-sky data assimilation. This study introduces an offline nonlinear bias correction model based on the machine learning (ML) technology of Random Forest to enhance the impacts of Fengyun-4A Advanced Geostationary Radiation Imager (AGRI) all-sky radiance data assimilation. The effects of the developed model were comprehensively evaluated through sensitivity experiments based on the NoBC, BC and modified BC schemes for two super typhoon cases. Among them, the modified BC scheme is designed to extract the features of cloud-affected systematic biases, which are more prevalent in the all-sky IR radiance assimilation. Results showed that the modified BC scheme outperforms other schemes in terms of removing the cloud-impacted systematic bias while retaining the useful meteorological signal. Whereas, those biases were improperly corrected by the original BC scheme when the inputs of a grid point were handled by the ML model site by site without the feature extraction, leading to a non-Gaussian error distribution. Assimilating those better-corrected IR radiances in the modified BC experiments would lead to a greater improvement in the analysis of the humidity and cloud ice. Based on the improved initial condition, the positive effects of the modified BC scheme are also evident in the forecasts of atmospheric variables and typhoon systems.

Reactive Transport Models (RTMs) are essential tools for understanding and predicting intertwined ecohydrological and biogeochemical processes on land and in rivers. While traditional RTMs have focused primarily on subsurface processes, recent watershed-scale RTMs have integrated ecohydrological and biogeochemical interactions between surface and subsurface. These emergent, watershed-scale RTMs are often spatially explicit and require extensive data, computational power, and computational expertise. There is however a pressing need to create parsimonious models that require minimal data and are accessible to scientists with limited computational background. To that end, we have developed BioRT-HBV 1.0, a watershed-scale, hydro-biogeochemical RTM that builds upon the widely used, bucket-type HBV model known for its simplicity and minimal data requirements. BioRT-HBV uses the conceptual structure and hydrology output of HBV to simulate processes including advective solute transport and biogeochemical reactions that depend on reaction thermodynamics and kinetics. These reactions include, for example, chemical weathering, soil respiration, and nutrient transformation. The model uses time series of weather (air temperature, precipitation, and potential evapotranspiration) and initial biogeochemical conditions of subsurface water, soils, and rocks as input, and output times series of reaction rates and solute concentrations in subsurface waters and rivers. This paper presents the model structure and governing equations and demonstrates its utility with examples simulating carbon and nitrogen processes in a headwater catchment. As shown in the examples, BioRT-HBV can be used to illuminate the dynamics of biogeochemical reactions in the invisible, arduous-to-measure subsurface, and their influence on the observed stream or river chemistry and solute export. With its parsimonious structure and easy-to-use graphical user interface, BioRT-HBV can be a useful research tool for users without in-depth computational training. It can additionally serve as an educational tool that promotes pollination of ideas across disciplines and foster a diverse, equal, and inclusive user community.