JOURNAL ARTICLES
What follows are summaries that have not been peer-reviewed or vetted by the article authors; read the full article for peer-reviewed conclusions. Please note that no single study is ever definitive, and each must be taken in the context of the broader scientific literature.
WoFS-FLASH coupled forecasts for the July 2025 Texas Hill Country Flash Flood Disaster
Bulletin of the American Meteorological Society
Experimental models accurately forecast 2025 Texas floods. The 2025 Texas Hill Country flash floods killed over 120 people. While operational National Weather Service models in use in July 2025 were able to highlight the potential for flash flooding, they were not able to predict precipitation intensity and location at a detailed-enough level to make a highly accurate, basin-specific warning very far in advance for places like Camp Mystic for Girls, where many campers died. Using data from July 2025, researchers ran an experimental, fine-resolution configuration of the NOAA/National Severe Storms Laboratory “Warn on Forecast” system (WoFS) with model runs fed into the “FLASH” flood prediction system. As the authors note, more than 50% of the model runs “would have yielded forecasts of extreme and record-breaking flows 5–7 [hours] in advance, providing strong motivation to develop a coupled WoFS-FLASH for real-time application" which might help save lives in the future.
Future Shifts in Severe Storm Environments Revealed Through Profile-Based Clustering
Journal of Climate
Models project increased U.S. thunderstorm threat. This study uses a climate model and a novel classification method to examine how atmospheric conditions favoring severe thunderstorms may change in a warmer climate. The most extreme of these environments show the largest increases and expand beyond the traditional “Tornado Alley.” Severe weather season is projected to start earlier and become more concentrated, signaling a growing threat to lives and property across a broader portion of the United States.
Performance of a Real-Time North Atlantic Tropical Cyclone Prediction System using the Model for Prediction Across Scales during the 2024 NOAA Hurricane Forecast Improvement Program Real-Time Experiment
Weather and Forecasting
MPAS model proved adept at forecasting Helene, Milton. During a real-time NOAA experiment in 2024, twice-daily five-day forecasts from the NOAA Global Systems Laboratory using the Model for Prediction Across Scales (MPAS) were able to predict tropical cyclone (TC) track, hazards, structure, and intensity with skill “comparable to or exceeding” NOAA’s operational forecasting models in most cases, especially for landfalling TCs Francine, Helene, and Milton. This raises the possibility of operational MPAS-based systems.
Policy-Driven Land Deformation in Coastal Megacities Quantifying Coupled Subsidence–Uplift Dynamics and Infrastructure Risks via a Viscoplastic Framework
Earth Interactions
Restoring groundwater under sinking cities leads to new hazards. Some coastal Chinese megacities are sinking due to over-extraction of the groundwater beneath them; local authorities have attempted to address this by restricting groundwater extraction and/or diverting water. Long-term monitoring in Tianjin reveals that even though these efforts have successfully restored groundwater, the ground itself responds in a non-uniform way: some areas lift while others remain compressed. The authors have developed a model to help planners assess the risks to infrastructure from this newly uneven ground.
Record-Breaking Atmospheric River Drives April 2024 Extreme Precipitation in the United Arab Emirates and Surrounding Gulf Region
Bulletin of the American Meteorological Society
Atmospheric river amplified catastrophic UAE flooding. In April 2024, the United Arab Emirates experienced catastrophic flooding, with more than 170% of the typical annual rainfall in just 72 hours. While much of the discussion has focused on local storms, this paper emphasizes the role of a larger-scale “record-breaking moisture transport corridor” (atmospheric river) in triggering the thunderstorms and flooding, and notes the vulnerability of arid regions like the Gulf to future extreme precipitation.
SCOAPE-II: A 2024 Multiplatform Measurement Campaign off the US Gulf Coast to Assess Oil and Gas Emissions on the Outer Continental Shelf
Bulletin of the American Meteorological Society
Ships and planes beat satellites for Gulf of Mexico oil and gas emissions monitoring. The 2024 SCOAPE-II experiment used ship- and aircraft-borne instruments to analyze air pollution emissions from oil and natural gas (ONG) plants on the U.S. Gulf Coast and compare them with satellite data. Ship and aircraft observations captured dozens of persistent methane and nitrogen dioxide plumes from ONG plants, and proved more accurate at methane measurement than satellites. Satellite data from the European TROPOMI instrument were more accurate than data from NASA’s TEMPO instrument at measuring nitrogen dioxide (a newer TEMPO data format is still being analyzed).
The Impact of Lakes in North American Regional Climate Simulations
Journal of Applied Meteorology and Climatology
In simulations, lakes significantly impact North American climates and favor cyclone development. A modeling study finds that lakes cool regional surface temperatures (up to 2.5 kelvin) and increase precipitation rates as much as 25% (although the Great Lakes may actually reduce summer precipitation). Lakes can also enhance climate change and affect the cyclonic/anticyclonic activity near them, with most lakes favoring the development of cyclones over anticyclones. These findings highlight the importance of accurate lake modeling for understanding and predicting North American climate.
Spatiotemporal Variations in Warm Season Heat Extremes in the Midwestern United States, 1959-2020
Journal of Applied Meteorology and Climatology
Extreme heat stress events increasing faster than temperatures across the Midwest. Heat stress extremes are intensifying across the U.S. Midwest at a rate not apparent from temperature increases alone. A study of trends in wet-bulb globe temperature (WBGT), which incorporates physiologically important factors like humidity, wind, and solar radiation, finds extreme events (including those requiring event cancellations) increasing in frequency and affecting larger areas.
The Paris 2024 Olympics Research Demonstration Project
Bulletin of the American Meteorological Society
Using the Paris Olympics to study urban weather and sports decision-making. The World Meteorological Organization’s multinational Paris 2024 Olympics Research Demonstration Project (Paris2024 RDP) used the Paris 2024 Olympics to study weather hazards and improve modeling at the fine geographic scales needed to capture urban weather. These efforts yielded insights on Paris’s effect on thunderstorms, modeling local air pollution transport during heat waves, modeling heat stress at fine scales within cities, the sociology of calling off events due to weather, developing weather protocols for huge public sporting events like the 2024 Olympics’ “Marathon for All,” and more.
Asian Dust Emissions under Climate Change
Journal of Climate
East Asian dust responds to vegetation and climate factors. A machine-learning study of dust pollution in China and Mongolia finds that in more vegetated areas, vegetation declines are the main driver of dust activity, and vegetation restoration reduces dust. In areas with persistently bare vegetation, climate factors like soil moisture are the main driver of dust pollution. Revegetation is a vital strategy for mitigating dust, especially given recent vegetation losses in Mongolia.
Numerical Simulation of the Impact of Lake Size on the Lake Breeze over Bosten Lake, China
Journal of Applied Meteorology and Climatology
Bigger lake, bigger breeze. A modeling experiment of the breeze created by Bosten Lake in China found that, when the size of the lake in the model was increased, the breeze from the lake accelerated and traveled a greater distance from the lake. When the modeled lake was contracted in size, the breeze grew weaker and reached less far across the land surface.
CNN-based Surface Temperature Forecasts with Ensemble Numerical Weather Prediction
Monthly Weather Review
Convolutional neural network helps improve temperature forecasts. Medium-range (up to around 5 days in advance) temperature forecasts typically rely on low-resolution numerical weather prediction (NWP) models, which are prone to error. This study integrates a convolutional neural network (CNN) with an ensemble of low-resolution NWP models to produce high-resolution surface temperature forecasts, providing a practical and scalable solution for improving forecast accuracy in offices with limited resources.
Unveiling Future Individual and Compound Heat and Air Pollution Extremes in China: insights for mitigation
Bulletin of the American Meteorological Society
Increasing, climate-dependent heat and pollution extremes in China. Heat waves and air pollution pose major health risks in China, but long-range CMIP6 climate models lack the daily detail needed to examine trends in compound heat-pollution extreme events. The authors use machine learning to downscale monthly data and create daily pollution estimates. Using these estimates, they find that even if strong emissions controls are put in place, harmful compound heat-ozone extremes will persist in a warmer climate. Weak controls lead to dramatic increases in air pollution events. Effective risk management requires addressing both air pollution and climate warming.
Generative Adversarial Networks for Deep Learning Downscaling of Temperature, Humidity, and Precipitation
Artificial Intelligence for the Earth Systems
Deep-learning may help improve local climate projections. Climate predictions are usually large-scale and low-resolution to capture overall patterns. Effectively “downscaling” that information, to capture and predict more local extremes and improve preparedness, is computationally intensive. To provide a more efficient alternative, the authors train a deep-learning model (a “generative adversarial network” or GAN) on downscaling data and find that it performs well for many downscaling tasks.
Revealing Atmospheric Von Kármán Vortex Street Past a Severe Thunderstorm
Monthly Weather Review
When thunderstorms act like mountains. A von Kármán vortex street (VKVS) is a type of turbulent air vortex that forms on the lee side of topographic obstacles like mountains as they interrupt air flow. By examining a 2015 event over the Mediterranean, the paper highlights evidence that such structures can also form downwind of a strong thunderstorm, which acts as an atmospheric obstacle.
Implementing a National Framework for Climate Services: Understanding progress, challenges and future opportunities
Bulletin of the American Meteorological Society
Africa, Caribbean lead progress on National Frameworks for Climate Services. This paper examines global progress toward implementing World Meteorological Organization-recommended National Frameworks for Climate Services to “inform decision-making in climate sensitive socioeconomic sectors.” More than 70 countries have initiated this process, with extensive progress in Africa and the Caribbean (other countries, such as the United States, already had equivalent frameworks).
Winter Warmth: Quantifying Historical Changes in Very Warm Winter Days Across The United States
Journal of Applied Meteorology and Climatology
Very warm winter days on the rise. Exceptionally warm winter days can disrupt snowpack and water resources, plant growth cycles, and more. The authors find that “very warm” days (in the 95th percentile or higher for local winter days) in December-February have increased across the contiguous United States since the 1960s, and in some regions they are increasing faster than the average rate of temperature change.
Assessing and Refining the Heat Index for Sub-daily Heat Conditions
Journal of Applied Meteorology and Climatology
Suggested improvements to the Heat Index for extreme and dry heat risk. The author highlights two potential issues with NOAA’s Heat Index (HI, a measure of potential heat stress) when used to measure heat risk at sub-daily timescales, and suggests solutions. 1) Due to the way HI is calculated, the HI value can be less than the actual air temperature under hot and dry conditions. Keeping HI values at greater than or equal to the real temperature may avoid under-diagnosing the severity of dry heat. 2) The categorization of “Extreme Danger” is set at an HI of 125 °F, a condition that is almost never met. To increase HI’s usefulness, the author suggests five danger categories instead of four, with level one starting at 80 °F, level 2 at 90 °F, level 3 at 95 °F, level 4 at 100 °F, and level 5 at 105°F.
You can view all research published in AMS Journals at journals.ametsoc.org .
Bulletin of the American Meteorological Society