This study proposes an automatic method to parameterize atmospheric optical depths in the Radiative Transfer for TIROS Operational Vertical Sounder (RTTOV) version 13 model. The approach combines statistical inference and Least Absolute Shrinkage and Selection Operator (LASSO) regression to reduce parameters and select relevant gases. Tests with Visible Infrared Imaging Radiometer Suite (VIIRS) channels show reduced computation while preserving accuracy.

This study evaluates the performance of the varying-resolution mesh model, E3SMv2-MPAS, in simulating key Arctic processes—from sea ice distribution and surface properties to the complex, three-dimensional structures of the Atlantic Water layer. We also pinpoint specific areas where the model still struggles, providing valuable directions for future model development. This progress means scientists can now use this tool more confidently to understand how Arctic ocean layers work and change.

We evaluated a new ensemble data assimilation system that uses satellite observations in all weather conditions for global weather forecasts. The results show that including cloud- and precipitation-affected satellite data improves forecasts of moisture, wind, and clouds, especially in the tropics. This work highlights the potential of this new ensemble data assimilation system to enhance global weather forecasts.

CIBUSmod is a biophysical agri-food system model designed to evaluate resource use and environmental impacts on national and sub-national level under future scenarios involving changes in demand and agricultural production systems. It provides a flexible framework that can integrate regionalised data at the spatial resolution and aggregation level available. By enhancing transparency and usability in food systems modelling, CIBUSmod is a valuable tool for exploring agri-food systems transitions.

This paper describes the software package JuWavelet, which implements the continuous wavelet transform, which is a popular tool in the geosciences to analyse wave-like phenomena. The code implements the transform in 1-D, 2-D, and 3-D for both analysis and synthesis, which closes a gap in available open-source software. The mathematics behind the transformation are given, and several examples showcase the capabilities of the software.

We developed the Speciated Isoprene Emission Model with Model of Emissions of Gases and Aerosols from Nature Algorithm for China to improve biogenic emission estimates using updated vegetation data and local measurements. The model predicts summer 2013 emissions of 10.92–11.37 teragrams of carbon. Validation shows our model performs better than the existing models, revealing underestimated isoprene impacts on ozone pollution in eastern China.

Tracers are commonly used in geodynamical models to track various quantities as material moves around. However, methods used to advect them typically do not respect the mass conservation equation, resulting in gaps and bunches in the tracer distribution. Here a method to correct this, based on nudging tracer positions in order to respect mass conservation, is presented. Tests show that it is effective and has a low computational cost.

We introduce GAMCR (Generalized Additive Models for Catchment Responses), a data-driven model that estimates how catchments respond to individual precipitation events. We validate GAMCR on synthetic data and demonstrate its ability to investigate the characteristic runoff responses from real-world hydrologic series. GAMCR provides new data-driven opportunities to understand and compare hydrological behavior across different catchments.

Injection of reflective sulfate aerosols high in the atmosphere is a proposed method to mitigate global warming. Climate simulations with injection are more expensive than standard future projections. We propose a method that dynamically scales the forcing fields based on pre-existing full-complexity data. This opens up possibilities for ensemble generation, new scenarios and higher resolution runs. We show that our method works for multiple model versions, injection scenarios and resolutions.

To understand how well current Earth system models simulate the natural world, we compared the models' outputs against measurements from satellites. Our results show these models struggle to accurately capture trends in variables related to carbon cycle, because the models can’t respond to human and environmental influences. This evaluation is crucial because improving these models will lead to more reliable forecasts of how ecosystems and the climate will change in the future.

Avalanche detection systems are crucial for forecasting, but distinguishing avalanches from other seismic sources remains a challenge. We propose novel autoencoder models to automatically extract features and compare them with engineered seismic features. These features are then used to classify avalanches and noise events. The autoencoder feature classifiers exhibit the highest sensitivity in detecting avalanches, while the engineered seismic classifier performs better overall.

Machine learning is being more widely used across environmental and climate science. This work reviews the use of machine learning in tropospheric ozone research, focusing on three main application areas in which significant progress has been made. Common challenges in using machine learning across the three areas are highlighted, and future directions for the field are indicated.

We developed a soybean model, an ecosystem model for crop yield (namely MATCRO-Soy), integrating crop response toward climate variables. It offers a detailed yield estimation. Parameter tuning in the model used literature and field experiments. The model shows a moderate correlation with observed yields at the global, national, and grid-cell levels. Development of this model enhances crop modeling diversity approaches, particularly in climate change impact studies.

Airborne microplastics are a new air pollutant but are not yet included in most global models. We add them to the UK Earth System Model to show how they move, change, and are removed from air. Smaller microplastics persist for longer and can travel further, even to Antarctica. While their current role in air pollution is small, their presence is expected to grow in future. This work offers a framework to assess future impacts of microplastics on air quality and climate.

Climate model performance at the air–sea interface has long been overlooked across the Southeast Asian seas. We thus assess various regional model physics configurations in this regard. Finding one optimal configuration is challenging: reliable rainfall rarely coincides with correct radiative heating. Simulations of rainfall however yield more dissensus, suggesting that this variable should be prioritized, for which the best results are obtained with the cumulus convection scheme of Tiedtke.

The WaterSip software locates regions where precipitation comes from. WaterSip evaluates of the water budget of the air masses, providing information on the conditions during evaporation, transport, and arrival at the target area. WaterSip can be easily configured and writes gridded output files. Guidance is given on where uncertainties arise using a case study, and best practices are recommended. This manuscript supports the comparison of different methods to find precipitation sources.

We developed a maize version of a process-based crop model coupled to a land-surface model by incorporating photosynthesis for C4 plants and maize-specific parameters. The model was calibrated with field data and literature, and it was extensively validated with global reference yields. The model effectively captured interannual yield variability in global and county-level yield data, demonstrating its potential for assessing the climate impacts on maize production.

The study explores Maritime Antarctica's geology, shaped by periglacial forces, using pioneering gamma-spectrometric and magnetic surveys on igneous rocks due to limited Antarctic surveys. Machine learning predicts radionuclide and magnetic content based on terrain features, linking their distribution to landscape processes, morphometrics, lithology, and pedogeomorphology. Inaccuracies arise due to complex periglacial processes and landscape complexities.

We document the forecasts of the composition of the stratosphere by the Copernicus Atmosphere Monitoring Service. The model's predictions are compared with satellite measurements over a recent period, during polar ozone depletion events, and after the Mount Pinatubo volcanic eruption. The system performs well for sulfate aerosols, ozone and several other key gases but not as well for several nitrogen-containing gases. Chemical processes in aerosols and polar clouds should be improved.

The future climate is very uncertain due to the large dispersion in projections from numerical models. Observational constraints (OCs) decrease this uncertainty using real-world observations. The article proposes a new rigorous statistical OC model that provides updated estimates of confidence intervals as used in Intergovernmental Panel on Climate Change (IPCC) reports. It allows the use of multiple observations simultaneously and proposes an innovative and proper illustration of this OC approach.

We present the Chemical Mechanism Integrator (Cminor) v1.0, a tool to predict concentrations of chemical compounds undergoing arbitrary reactions. Cminor is an advanced, open-source solver to model either combustion chemistry, or atmospheric chemistry and its direct influence on condensation of cloud droplets and the subsequent processing of aerosol. It uses the superdroplet idea, making it particularly feasible for coupling with such models, which is part of future work.

We present the development of the Unified Inputs (of initial and boundary conditions) for WRF (Weather Research and Forecasting)-Chem (UI-WRF-Chem) framework to support the Multi-Angle Imager for Aerosols (MAIA) satellite mission. Major updates include improving dust size distribution in the chemical boundary conditions, updating land surface properties using recent available satellite data and enhancing the representation of soil NO_x emissions. We demonstrate subsequent model improvements over several of the MAIA target areas.

We developed a methodology for comparison of simulation results by a dynamic global vegetation model (DGVM). Using this methodology, we reveal systematic differences between high- and low-resolution DGVM simulations caused by under-representation of climate variability in the low-resolution data and poor representation of shore lines and inland water bodies. In a study area covering European Union, the differences in aggregated output variables were found to be 2.8%–7.3%.

Snow microstructure controls snowpack properties, but most land surface models overlook this factor. To support future satellite missions, we created a new land surface model based on the Crocus scheme that simulates snow microstructure. Key improvements include better snow albedo representation, enhanced Arctic snow modeling, and improved forest module to capture Canada's diverse snow conditions. Results demonstrate improved simulations of snow density and melt across large regions of Canada.

Matter is a new computer model that simulates granular media like sand, snow, and soil. These materials can behave like both solids and fluids, making their modeling difficult. Matter addresses this with a unified framework, using a numerical solver called MPM. Able to capture cohesion, density variations and complex terrains, it's particularly relevant for snow avalanches or landslides. Matter runs efficiently on standard computers, making advanced simulations more accessible.

Earth system models consist of many components, coupled in time and space. Standard coupling algorithms introduce a numerical error, which one can compute with iterative coupling methods. We use such a method for a coupled model of a single vertical column of the atmosphere, ocean, and sea ice. We find that coupling errors in the atmosphere and at the ice surface can be substantial and that discontinuous physics parameterizations lead to convergence issues of the iteration.

We developed a computer program that can read chemical formulas and identify key features in thousands of organic compounds. This helps scientists estimate how easily these compounds evaporate, which is important for understanding air pollution and climate. We tested the program using real-world data and found it to be highly accurate. Our work makes it faster and easier to study the behavior of many complex chemicals in the atmosphere.

Geodynamical simulations cover a wide spatial and temporal range and are crucial to understand and assess the evolution of the Earth system. To enable computationally efficient modeling approaches that can account for potentially unknown subsurface properties, we present a surrogate modeling technique. This technique combines physics-based and machine-learning techniques to enable reliable predictions of geodynamical applications, as we illustrate for the case study of the Alpine Region.

We evaluated a high-resolution numerical weather prediction model in a small, semi-arid U.S. city using dense ground-based measurements. While the forecasts demonstrated good skill for temperature and humidity, they consistently overestimated wind and underestimated nighttime cooling, with inaccurate heat advection predictions. The results highlight the need for improved urban representation in forecast models to better support heat warning systems for small cities.