The study involved 773 participants in a nationwide online survey, with a subsample of 191 completing a follow-up survey one month later for test-retest analysis. Investigators examined the factor structure through exploratory and confirmatory factor analyses, randomly dividing the main sample into two subsamples for cross-validation. They also assessed measurement invariance across sex and age groups, internal consistency, test-retest reliability, and convergent and discriminant validity. The abstract describes this as a psychometric validation study, not an intervention trial, so the reported results concern measurement properties rather than the effects of a disaster-preparedness program.

Exploratory and confirmatory analyses supported a three-factor structure covering knowledge and management of the external situational environment, management of one’s emotional and psychological response, and management of one’s social environment. The confirmatory factor model showed adequate fit, with a comparative fit index of 0.942 and a root mean square error of approximation of 0.062. The total PPDTS-J23 score had high internal consistency, with alpha equal to 0.956. Alpha values for the three subscales were 0.919, 0.936 and 0.827, while measurement invariance was supported across sex and age groups. Convergent validity was supported, but discriminant validity was only partially supported.

Test-retest reliability was acceptable for the total score, with an intraclass correlation coefficient of 0.780 at the one-month follow-up. The coefficient for the social-environment subscale was lower, at 0.657, below the criterion adopted by the researchers. These findings indicate that the PPDTS-J23 can serve as a reliable measure of psychological preparedness for hydro-meteorological hazards in Japan, while the partially supported discriminant validity and lower reliability for one subscale bound the conclusions. The abstract states that the scale may be used to evaluate disaster-education program outcomes, characterize groups’ preparedness profiles, inform community-based disaster-risk reduction and support mental-health planning. Because the available report is the abstract of an accepted article that remains subject to further edits, it does not provide the full methods or results needed to assess those proposed applications directly.

Subject of Research: Psychometric validation of a Japanese scale for psychological preparedness against hydro-meteorological hazards

Article Title: Development and validation of the Japanese version of the psychological preparedness for disaster threat scale

Article References: Kubo, T., Takamoto, M., Suzuki, M., Takenouchi, K., & Sadamitsu, C. (2026). Development and validation of the Japanese version of the psychological preparedness for disaster threat scale. BMC Psychology. https://doi.org/10.1186/s40359-026-05639-3

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05639-3

Keywords: psychological preparedness, disaster risk, Japan, psychometric validation, survey research, typhoons, flooding, severe storms, measurement invariance, disaster education, Development, validation

The study, conducted by Can Wang of Hubei University of Economics and colleagues at Hubei Minzu University, Postal Savings Bank of China, China Export & Credit Insurance Corporation, and Enshi Polytechnic, draws on comprehensive data from the Chinese A-share market spanning 2008 to 2024. That sixteen-year window captures a period during which typhoons repeatedly struck China’s densely populated and economically vital eastern and southern coastlines, giving the researchers a rich natural laboratory in which to observe how capital markets absorb and transmit physical climate shocks.

To isolate the effect of typhoons from the ordinary noise of market life, the team employed a two-way fixed effects difference-in-differences model, a workhorse of modern empirical economics. The logic is straightforward: compare firms that were hit by a typhoon with comparable firms that were not, before and after the event, while controlling for unchanging characteristics of each firm and for economy-wide shocks that affect everyone at once. If the exposed firms show a sharp rise in stock volatility relative to their unexposed peers, the typhoon itself is the most plausible explanation. The researchers found exactly that pattern: typhoon disasters significantly exacerbated the stock volatility of affected enterprises.

But measuring volatility spillovers between sectors requires heavier statistical machinery. For that task, the authors turned to a framework known as VaR-BEKK-MVGARCH-DUMMY-T. The multivariate GARCH component models how the variance of returns in different sectors evolves over time and, crucially, how shocks to one sector’s volatility feed into another’s. The BEKK structure allows the researchers to estimate the direction of these cross-market transmissions rather than merely noting that two sectors move together. The value-at-risk layer gauges tail risk, the potential for extreme losses, while dummy variables flag the typhoon episodes themselves, and the T designation accommodates the fat-tailed distributions that characterize real financial returns. Together, these tools let the team trace not just whether volatility rose, but where it originated and where it flowed.

The answer was strikingly asymmetric. The study confirms a unidirectional volatility spillover effect running from the non-financial sector to the financial sector during extreme weather events. In other words, when typhoons whipsaw the stock prices of manufacturing, utility, and other real-economy firms, that turbulence propagates into banks, insurers, and other financial institutions, not the other way around. This directionality matters. It suggests that during climate shocks, the financial system behaves less like an independent source of instability and more like a conduit, absorbing and amplifying distress that originates in the productive economy.

The finding is grounded in stakeholder theory, the idea that a firm’s value and risk are shaped by its web of relationships with suppliers, customers, employees, lenders, and investors. A typhoon disrupts those relationships on multiple fronts simultaneously. Factories shut down, shipments are delayed, workers cannot reach their jobs, and customers face shortages. Each of these frictions degrades expected cash flows, and because stock prices embody expectations about those cash flows, the uncertainty surrounding them translates directly into wider price swings. The researchers describe this as a decline in supply chain resilience among non-financial enterprises, one of two distinct pathways through which typhoons amplify volatility risk across industries.

The second pathway operates on the psychology of markets. Typhoons, the study finds, destabilize investor confidence within the financial system by exacerbating widespread uncertainty. Extreme weather events are inherently unpredictable in their scope and consequences, and that unpredictability infects the expectations of traders, fund managers, and analysts. When investors cannot easily assess how badly a storm has damaged a firm’s assets or how long its operations will be impaired, they demand a larger premium for holding the stock, buying and selling more aggressively as information trickles in. The result is heightened volatility not only in directly affected firms but across the financial institutions that hold their debt, underwrite their insurance, and extend their credit.

This dual mechanism, physical disruption of supply chains on one side and erosion of investor confidence on the other, provides what the authors call robust micro-level evidence for cross-market systemic risk contagion triggered by extreme climate events. Systemic risk is usually discussed in the language of banking crises and interbank lending, but this research reframes it in climate terms. A single typhoon, by damaging the real economy, can set off a chain of financial reactions that reaches institutions hundreds or thousands of kilometers from the landfall point. The heterogeneity the study emphasizes, the distinction between financial and non-financial enterprises, is essential to understanding that chain, because the two sectors respond to and transmit climate risk in fundamentally different ways.

The implications extend well beyond China. As climate change intensifies, extreme weather events are projected to become more frequent and more severe, and financial regulators around the world are increasingly demanding that banks and insurers stress-test their portfolios against physical climate risks. This study suggests that such stress tests should model not only the direct exposure of financial institutions to storms, but also their indirect exposure through the real-economy firms they finance. A bank whose loan book is concentrated in typhoon-prone manufacturing regions may face volatility-driven losses even if none of its own branches ever loses power. Similarly, portfolio managers seeking to hedge climate risk cannot treat financial and non-financial holdings as independent, because the study’s evidence shows volatility flowing in one direction between them during disasters.

The research also deepens the theoretical foundations of the rapidly growing field of climate finance, which seeks to understand how climate risk is priced, transmitted, and managed in markets. By documenting heterogeneous industry risk transmission mechanisms with a rigorous identification strategy and a sophisticated volatility model, the study moves the field beyond simple correlations between weather and returns. It demonstrates that the architecture of the economy, who supplies whom, who lends to whom, and who insures whom, determines how a physical shock becomes a financial one. As storms grow stronger and coastlines ever more crowded with industry, understanding that architecture may become one of the most important tasks in modern finance. The study’s authors received support from the Youth Fund Project of Hubei University of Economics and the National Statistical Science Research Project, and the work was published open access, making the detailed evidence available to researchers and regulators worldwide.

Subject of Research: The spillover effects and mechanisms of typhoon disasters on stock volatility in Chinese financial and non-financial enterprises

Article Title: The spillover effects and mechanisms of typhoon disasters on stock volatility: considering the heterogeneity of financial enterprises and non-financial enterprises

Article References: Wang, C., Qin, S., Tan, S., Huang, W., & Tao, K. (2026). The spillover effects and mechanisms of typhoon disasters on stock volatility: considering the heterogeneity of financial enterprises and non-financial enterprises. Humanities and Social Sciences Communications. https://doi.org/10.1057/s41599-026-09394-z

Image Credits: AI Generated

DOI: 10.1057/s41599-026-09394-z

Keywords: typhoons, stock volatility, climate finance, spillover effects, financial sector, non-financial enterprises, supply chain resilience, investor confidence, systemic risk, difference-in-differences, multivariate GARCH, Chinese A-share market

Between July 2019 and December 2023, a team led by Sang-Ho Moon of the Korea Institute of Geoscience and Mineral Resources (KIGAM) collected 617 valid rainwater samples from 24 monitoring stations distributed evenly across the mainland. The stations span the country’s four major river basins: the Han, Geum, Nakdong, and Yeongsan–Seomjin. Sampling intervals ranged from monthly during the wet season to every two or three months in the dry months, and the researchers used commercial Palmex precipitation collectors designed to prevent post-collection evaporation. Tests by the International Atomic Energy Agency have shown that water stored in these devices for nearly a year does not experience significant isotopic alteration, making them ideal for long-duration field campaigns.

The analytical work combined two complementary techniques. Samples from 2019 and 2020 were measured with isotope ratio infrared spectroscopy using wavelength-scanned cavity ring-down spectroscopy, a laser-based method, while samples from 2021 through 2023 were analyzed with continuous-flow isotope ratio mass spectrometry coupled to a high-temperature conversion elemental analyzer. Both approaches were normalized against international reference waters, and analytical uncertainties were kept below ±0.2 per mil for oxygen-18 and ±2 per mil for deuterium. The resulting dataset covers δ18O values from −15.04 to −2.64 per mil and δD values from −112.2 to −9.0 per mil, a wide range that immediately signaled the strong seasonal contrasts shaping Korean rainfall.

The centerpiece of the study is the local meteoric water line, or LMWL, the regression that describes how δD covaries with δ18O at a site. For the full dataset the researchers obtained δD = 8.58 δ18O + 17.13, with a coefficient of determination of 0.94. That line is remarkably close to the one previously reported for Jeju Island, but its slope and intercept exceed those of the global meteoric water line and differ sharply from values reported for China. The reason, the authors argue, lies in winter: a large fraction of cold-season precipitation in South Korea deviates from the global line, inflating both parameters and pushing deuterium excess values above 15 per mil.

Season-specific water lines tell the story even more vividly. Summer precipitation, dominated by the stationary monsoon front and typhoons, yields an LMWL of δD = 8.10 δ18O + 10.48, essentially indistinguishable from the global line. A slope near 8 indicates that vapor formed under near-equilibrium conditions through slow evaporation from the ocean, while the intercept near 10 points to high humidity, roughly 85 percent, and minimal evaporation during rainfall. Winter precipitation, by contrast, produces a line of δD = 7.51 δ18O + 15.27, with a lower slope and a much higher intercept. This combination reflects kinetic fractionation as cold, dry Siberian air masses sweep southeast across the Yellow Sea, rapidly picking up moisture recycled from the sea surface before dumping snow and rain on the peninsula.

Deuterium excess, the second-order isotope parameter that tracks conditions at the vapor source, swings dramatically with the seasons. Values ranged from 4.61 to 28.36 per mil over the study period, staying below 10 per mil during the June-to-September rainy season and above 15 per mil from November to March. Following the classic framework of Merlivat and Jouzel, these numbers translate to formation-time humidities of 85 percent or more in summer and only 65 to 80 percent in winter. The winter signal, the researchers suggest, most plausibly reflects continental recycling over the Yellow Sea, although they caution that the intrinsic signature of the Siberian air mass itself may also contribute, and that interpreting moisture sources from deuterium excess alone carries inherent limitations.

To untangle the geographic controls, the team calculated amount-weighted mean isotope values for each of the 24 stations and subjected them to principal component analysis and multiple regression. The results reveal a textbook continental effect: in the Geum, Han, and Nakdong basins, isotope values become progressively depleted with distance from the coast as air masses rain themselves out moving inland. Latitude matters too. For 21 of the 24 stations, δ18O decreases by about 0.55 per mil per degree of latitude, a gradient closely matching values recorded for continental stations in North America and Europe. Altitude leaves its mark as well, with depletion rates of −0.15 to −0.39 per mil per 100 meters for δ18O in three of the four basins, consistent with global norms.

Three stations refuse to obey the rules, and their misbehavior is scientifically revealing. Geumsan, ringed by tall mountains in the Geum basin, shows the most depleted values in its basin, likely because clouds lose their heavy isotopes over the western lowlands before climbing toward the eastern highs. Kangreung and Pohang, both on the eastern coastal side of the Taebaek mountain range, are strongly enriched and sit far above the national latitude regression, apparently shielded from the continental pattern by the topographic barrier. The Nakdong basin also stands out with an anomalously steep apparent altitude gradient of −1.26 per mil per 100 meters for δ18O, which the authors attribute to its unique exposure to the southeastern sea and to mountain-ringed sub-basins in the north rather than to elevation alone.

The statistical treatment adds quantitative rigor. Principal component analysis showed that three factors explain over 90 percent of the total variance, with oxygen and hydrogen isotope values loading primarily onto latitude and altitude, while longitude loaded separately. Multiple regression confirmed that latitude and altitude are statistically significant predictors, jointly explaining 55 to 66 percent of the isotopic variation, whereas longitude alone was not significant. Notably, the classic amount effect, the tendency for heavier rain to be isotopically lighter, was only weakly expressed nationally; rainfall above roughly 500 millimeters showed no correlation with isotope values, suggesting that air-mass origin, not rainfall quantity, dominates the signal in extreme monsoon events.

Beyond its scientific elegance, the dataset arrives at a practical moment. Because the isotopic composition of shallow groundwater generally mirrors that of local precipitation, the new LMWLs, amount-weighted means, and isoscape maps provide the baseline needed to identify groundwater recharge sources, estimate recharge seasonality, and detect evaporation in soil and shallow aquifer zones across South Korea. As climate change intensifies the monsoon front and shifts typhoon tracks, having a robust four-year isotopic reference for the mainland will allow hydrologists to detect future changes in the water cycle with confidence. For a peninsula squeezed between maritime and continental air masses, the rain has finally been made to speak in full detail.

Subject of Research: Stable isotopic composition of precipitation across mainland South Korea

Article Title: Long-term spatiotemporal assessment of stable isotopic composition of precipitation across South Korea (2019–2023)

Article References: Moon, S.-H., Kim, Y., Moon, H. S., Kim, H. C., Yu, S., & Jung, J. (2026). Long-term spatiotemporal assessment of stable isotopic composition of precipitation across South Korea (2019–2023). Environmental Earth Sciences, 85(16), Article 430. https://doi.org/10.1007/s12665-026-13169-9

Image Credits: AI Generated

DOI: 10.1007/s12665-026-13169-9

Keywords: stable isotopes, precipitation, South Korea, monsoon, local meteoric water line, deuterium excess, hydrology, groundwater recharge, isoscapes, East Asia, typhoons, water cycle

The research team, led by corresponding author Jia Sun of the First Institute of Oceanography under China’s Ministry of Natural Resources, examined tropical cyclone records from 1980 to 2025. The dataset spans the western North Pacific, the basin that generates more tropical cyclones than any other on Earth and that supplies the storms striking eastern China, the Korean Peninsula, and Japan. By tracking where storms formed, the paths they followed, and where they came ashore, the researchers were able to detect a coherent pattern that is difficult to dismiss as random variability. Fewer storms are forming and making landfall at lower latitudes, while a growing share of the basin’s activity is concentrated farther north than it was in earlier decades.

The numbers behind the shift are striking in their geographic clarity. Landfall likelihood decreased south of 20 degrees north latitude but increased north of 30 degrees north, a band that broadly covers parts of eastern China, the Korean Peninsula, and Japan. In practical terms, communities that historically sat at the northern fringe of the tropical cyclone zone are now finding themselves closer to its center of gravity. The storms themselves are also changing their behavior: they tend to form farther to the northwest, follow tracks that are both more landward and more poleward, and move modestly faster along their paths toward the coast.

One of the most consequential details in the study concerns time. The average duration that landfalling cyclones spent over the ocean dropped from about 5.6 days in the 1980 to 2015 period to about 4.7 days in 2016 to 2025. That difference of nearly a full day may sound like a technical footnote, but it has direct implications for emergency management. A tropical cyclone that forms closer to shore and races toward the coast more quickly compresses the window available for forecasting, evacuation planning, and public communication. Coastal cities that once enjoyed several days of lead time before a storm’s arrival may find that window narrowing, even when the storm’s eventual intensity is accurately predicted.

Guihua Wang of Fudan University, the study’s co-corresponding author, emphasized that this shrinking lead time is a planning problem as much as a forecasting one. Preparedness frameworks in many coastal regions are built on accumulated local experience, which implicitly assumes that future storms will behave like past ones. The new results challenge that assumption directly. If tropical cyclones are forming in different places and traveling along different routes with different speeds, then risk assessments and early warning systems calibrated to historical storm behavior may systematically underestimate the threat facing northern East Asian coastlines, particularly with respect to how much warning time will actually be available.

Notably, the study did not find a significant long-term increase in the maximum strength that tropical cyclones reached. This distinction matters for how the results should be interpreted. The changing pattern of danger along East Asian coasts appears to be driven mainly by where storms travel and where they make landfall, rather than by a steady intensification of the storms themselves. In other words, the hazard is migrating geographically rather than uniformly amplifying. For residents of lower-latitude coastal zones, the news may be comparatively reassuring, at least in terms of landfall frequency. For communities north of 30 degrees north, the opposite applies, because their exposure is growing even though the storms hitting them are not becoming categorically stronger.

What physical mechanism could push an entire basin’s storm tracks poleward? The researchers point to changes in the large-scale winds over the Pacific, which act as steering currents for tropical cyclones. These environmental flows appear to be guiding more storms on westward, poleward, and landward trajectories, effectively funneling them toward the mid-latitude coasts of East Asia. Steering flow is a fundamental control on tropical cyclone motion: a storm is largely carried along by the deep-layer winds surrounding it, so even modest shifts in the position or strength of features like the western Pacific subtropical high can translate into substantial changes in where storms ultimately go. The observed northward migration of landfall locations is consistent with such a large-scale reorganization of the steering environment.

The authors are careful, however, not to attribute the entire pattern to long-term climate change. The observed record reflects a combination of secular climate trends and natural climate variability, including multidecadal oscillations in the Pacific atmosphere-ocean system that can redistribute storm activity on their own. Disentangling the anthropogenic signal from natural swings is one of the hardest problems in tropical cyclone climatology, and the researchers caution that separating these influences will require longer observational records and targeted climate model experiments designed to isolate individual drivers. Until such work is done, the northward shift should be understood as a robust observed trend whose ultimate causes remain under active investigation.

For the tens of millions of people living along the coasts of eastern China, the Korean Peninsula, and Japan, the practical lesson is straightforward: yesterday’s tropical cyclone map may not describe tomorrow’s risk. Building codes, evacuation routes, insurance pricing, and drainage infrastructure are all calibrated to historical landfall frequencies, and a poleward redistribution of hazards can leave those investments mismatched to the actual threat. The study suggests that planners need to consider not only how strong a future cyclone may become, but also where it is increasingly likely to travel, how fast it will move, and how much warning time its shortened ocean journey may leave. As the geography of tropical cyclone danger continues to migrate north, the regions that adapt their risk frameworks to this new reality will be better positioned than those that continue to plan for a storm climate that no longer exists.

Subject of Research: Poleward redistribution of tropical cyclone hazards and landfall patterns over East Asia in the western North Pacific

Article Title: East Asia's tropical cyclone risk is moving steadily north

Article References: East Asia's tropical cyclone risk is moving steadily north. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: tropical cyclones, East Asia, western North Pacific, landfall, poleward shift, coastal hazards, steering winds, early warning systems, climate variability, typhoons, risk assessment, Advances in Atmospheric Sciences

The study addresses a hazard that is often overshadowed by headlines about typhoon winds and flooding. Sediment-related disasters, which include shallow landslides, debris flows, and slope failures, are triggered when rainfall destabilizes hillslopes, sending soil, rock, and woody debris cascading into communities, roads, railways, and critical infrastructure such as power transmission networks. In Japan, where much of the population lives on narrow coastal plains hemmed in by mountains, the distance between a vulnerable slope and a densely populated neighborhood can be measured in hundreds of meters. Understanding how climate change alters the frequency and intensity of the rainfall regimes that trigger these events is therefore not an abstract scientific question but a matter of practical urgency for disaster managers, utilities, and local governments.

To quantify the hazard, the researchers adopted a framework widely used in Japanese operational landslide assessment: sediment-related disaster potential, or SDP. Rather than modeling slope mechanics directly, SDP is built from two complementary measures of effective rainfall, meaning the cumulative precipitation that actually contributes to slope instability after losses to infiltration, evaporation, and other hydrological processes. The first measure uses a short accumulation window of 1.5 hours, capturing the kind of burst-like, high-intensity downpours that can rapidly saturate a shallow soil layer and trigger sudden slope failure. The second uses a long window of 72 hours, capturing the slower buildup of antecedent wetness that primes slopes for failure even when the triggering rain is less intense. Together, these two metrics encode the two classic pathways to landslide initiation: a sudden punch of water and a prolonged soaking.

The climate evidence underpinning the analysis comes from an unusually rich simulation resource. The team drew on large-ensemble climate simulations produced under Japan’s national modeling programs, in which global atmospheric models were run many times with slightly perturbed initial conditions to generate thousands of simulated years of weather, both for the historical climate and for a future climate under continued warming. From this ensemble, the researchers used dynamical downscaling to 5-kilometer horizontal resolution, a grid fine enough to resolve the mountainous topography and coastal geometry that shape Japan’s precipitation. At this scale, the models can represent orographic enhancement of rainfall, the interaction of typhoons with complex terrain, and the localized bands of heavy rain that coarser global models simply cannot capture. Large ensembles are essential for this kind of work because extreme events are by definition rare, and only by pooling thousands of simulated years can researchers distinguish a robust climate-change signal from natural variability.

When the team compared future and present climates, the headline result was a broad increase in sediment-related disaster potential across Japan. Warmer air holds more water vapor, roughly seven percent more per degree Celsius of warming, and this thermodynamic amplification intensifies extreme precipitation in ways that are now well established across the climate literature. But the study’s most valuable contribution lies in its treatment of the spatial heterogeneity of that increase. The projected changes in SDP were not uniform: some regions and some rainfall regimes showed far larger increases than others, and the pattern of change differed between the short-duration and long-duration metrics. A single national number would have hidden precisely the information that regional adaptation planners need most.

To untangle that heterogeneity, the researchers turned to self-organizing maps, an unsupervised neural-network technique originally introduced by the Finnish scientist Teuvo Kohonen. Self-organizing maps take high-dimensional data, in this case daily surface atmospheric circulation fields such as sea-level pressure patterns over and around Japan, and project them onto a two-dimensional grid of nodes, each representing a characteristic weather pattern, or WP. Similar circulation configurations cluster onto neighboring nodes, allowing researchers to compress the endless variety of daily weather into a manageable catalog of archetypal patterns. The method has become a standard tool in climate science for linking large-scale circulation to local extremes, because it makes no prior assumption about which patterns matter and lets the data reveal the natural modes of atmospheric variability.

Applying this technique to both present-day and future simulations, the team identified the dominant weather patterns associated with high sediment-related disaster potential and then evaluated how sensitive each pattern is to climate forcing. The analysis revealed that specific weather patterns act as the key regulators of SDP variability, and, crucially, that different patterns dominate in different parts of the country. Typhoon-type weather patterns, in which a tropical cyclone approaches or makes landfall, contribute predominantly to short-duration heavy rainfall in western Japan, where the steep terrain of Kyushu, Shikoku, and western Honshu sits directly in the path of many storm tracks. Frontal rain-type patterns, associated with the stationary Baiu rain front that stalls over Japan in early summer, dominate the long-duration precipitation that drives elevated SDP in eastern Japan, where prolonged rains can soak slopes for days.

This weather-pattern attribution has direct mechanistic implications. If typhoon-type patterns become more effective at generating short-duration extremes, western Japan faces a future in which flash-triggered slope failures grow more likely, demanding rapid-response warning systems and slope-stabilization measures tuned to intense, fast-onset events. If frontal patterns deliver longer, wetter episodes in eastern Japan, the relevant adaptation there involves antecedent-wetness monitoring, cumulative-rainfall thresholds, and patrol schedules for infrastructure such as power transmission equipment, which is precisely the operational context in which the SDP methodology was originally developed by the research group. The study thus converts a broad climate projection into region-specific guidance: the same national warming signal produces different disaster mechanisms on either side of the archipelago, and adaptation strategies should be designed accordingly.

The research also demonstrates the power of pairing very high-resolution ensembles with pattern-based analysis. Because the 5-kilometer downscaling resolves the terrain that governs where rain actually falls, the weather-pattern statistics derived from the simulations can be interpreted physically rather than treated as statistical artifacts. And because the underlying global ensemble spans thousands of simulated years, the team could assess how frequently each disaster-relevant weather pattern occurs in present and future climates and how the rainfall intensity within each pattern shifts with warming, separating the effects of changing circulation from the effects of a moister atmosphere. This dual attribution, of pattern frequency and pattern intensity, is exactly the kind of decomposition needed to build confidence in regional hazard projections.

For Japan, a nation that has repeatedly experienced catastrophic landslides in events such as the 2018 western Japan floods, the message is that the hazard landscape is shifting in regionally distinctive ways, and that the tools exist to anticipate the shift. The d4PDF ensemble and the 5-kilometer downscaling dataset used in the study are publicly documented and available, allowing other researchers, and potentially operational agencies, to extend the analysis to different warming scenarios, different hazard metrics, or other mountainous regions facing similar threats. As the authors note, the findings provide critical insights for developing region-specific climate adaptation strategies for sediment-related disaster management. In a warming world, the rain that brings down a hillside in Kyushu and the rain that brings down a hillside near Tokyo will arrive under different skies, and protecting the people below those slopes will require plans that know the difference.

Subject of Research: Climate change impacts on rainfall-induced landslide and sediment-disaster potential in Japan

Article Title: Climate change impacts on rainfall-induced sediment-related disaster potential in Japan: a weather-pattern analysis using large-ensemble climate simulations

Article References: Ohba, M., Nomura, M., Kitano, Y., & Hattori, Y. (2026). Climate change impacts on rainfall-induced sediment-related disaster potential in Japan: a weather-pattern analysis using large-ensemble climate simulations. Climate Dynamics, 64(11), Article 457. https://doi.org/10.1007/s00382-026-08417-4

Image Credits: AI Generated

DOI: 10.1007/s00382-026-08417-4

Keywords: sediment-related disasters, landslides, extreme rainfall, climate change, Japan, weather patterns, self-organizing maps, d4PDF, regional climate modeling, typhoons, Baiu rain front, disaster adaptation

The core problem the researchers tackled is one of scale and cost. Numerical weather prediction models simulate the atmosphere by solving equations of fluid dynamics and thermodynamics on a grid, and the finer that grid becomes, the more computationally punishing the exercise. Increasing spatial resolution leads to a quadratic or even cubic growth in the number of grid points, translating into orders-of-magnitude more floating-point operations. Even the European Centre for Medium-Range Weather Forecasts, in a landmark experiment, needed a substantial fraction of the Summit supercomputer at Oak Ridge National Laboratory to run a global 1 kilometer simulation at a pace of roughly one simulated week per day. Operational forecasting centers simply cannot afford to run kilometer-scale global models within the time windows that forecasts demand.

The practical workaround has long been downscaling: taking the output of a coarser global model and deriving finer regional detail from it. Dynamical downscaling embeds a high-resolution regional model within the global one, but inherits the same computational burden. Statistical downscaling, by contrast, learns the mapping between low-resolution and high-resolution meteorological fields from historical data, offering simpler implementation, lower cost, and often higher accuracy. Deep learning has supercharged this approach by borrowing techniques from computer vision, where super-resolution models routinely turn blurry images into sharp ones. The catch, as the authors note, is that meteorological downscaling must guarantee accuracy and physical consistency, not merely visual plausibility.

CorrDiff, originally developed by NVIDIA researchers, takes a two-step approach that elegantly sidesteps a fundamental obstacle in generative downscaling. Because low-resolution and high-resolution atmospheric data occupy very different statistical distributions, training a diffusion model directly on the translation between them is difficult. CorrDiff therefore first trains a deterministic regression model, a UNet neural network, to predict the conditional mean of the high-resolution field. A second network, an Elucidated Diffusion Model, is then trained to correct the regression output. During inference, random noise is fed through the denoising process conditioned on the low-resolution input, and the result is added back to the regression prediction. By sampling multiple random noises, the model produces an ensemble of plausible high-resolution fields, yielding a probabilistic forecast rather than a single deterministic answer.

The Chinese team pushed this framework far beyond its original scope. Where the original CorrDiff work covered Taiwan at 2 kilometer resolution on a 448 by 448 grid, the new model covers nearly all of China on a 1600 by 2400 grid at 3 kilometers, roughly twenty times larger. The researchers extended the target variables beyond surface quantities to include upper-air fields across six pressure levels, added high-resolution orography as an input, and introduced a global residual connection: instead of predicting the full high-resolution field, the networks learn the residual between it and a bilinear interpolation of the coarse input, a strategy that accelerated convergence and improved accuracy. Training relied on 25 kilometer ERA5 reanalysis data from ECMWF as inputs and 3 kilometer reanalysis from the China Meteorological Administration Regional Reanalysis Atmospheric System as ground truth, using 11,688 samples spanning 2019 to 2022.

The experimental design was meticulous. Five regression models and their CorrDiff counterparts were trained with different combinations of input and output variables, probing how dependencies between variables affect downscaling quality. On a validation set drawn from January, April, July, and October 2023, all models consistently beat simple interpolation of the coarse data. Notably, the team found that combining radar reflectivity prediction with general downscaling in a single model compromised accuracy for other variables, while adding orography improved surface pressure predictions. A smaller network failed to match the accuracy of its larger sibling, confirming that model capacity matters at this scale. All models were trained on eight NVIDIA H20 GPUs, with training times per epoch ranging from roughly one to four hours.

When the trained models were applied to real global forecasts, the results were striking. The researchers downscaled 25 kilometer forecasts from two sources: the conventional CMA Global Forecast System and Sphere Fusion Forecast, a deep learning weather model built on Spherical Fourier Neural Operators. Against the operational CMA Mesoscale Model, a 3 kilometer regional numerical prediction system, as a baseline, the downscaled forecasts achieved lower Mean Absolute Error for almost all target variables. Power spectral analysis added nuance: the diffusion model generated richer high-frequency detail than the regression model, and for many near-surface variables its spectra agreed more closely with the reanalysis ground truth than CMA-MESO did, though the authors caution it may also introduce some non-physical high-frequency components.

Perhaps the most compelling evidence came from tropical cyclones. In case studies of Typhoons Khanun and Haikui, the regression-based downscaling produced overly smooth wind fields, a consequence of the over-smoothing tendency of deterministic models, and dramatically underestimated the probability of radar reflectivity exceeding 50 dBZ. CorrDiff, by contrast, reproduced realistic small-scale convective structures and matched the reflectivity distribution of the reanalysis data far more closely, outperforming CMA-MESO in Fractions Skill Scores between 12 and 36 hours of forecast lead time. The team also showed that the ensemble variance of CorrDiff predictions correlates with actual error: grid points with low variance tend to be more accurate, offering a practical tool for identifying where a forecast can be trusted.

The authors are candid about limitations. CorrDiff’s Mean Absolute Error is consistently higher than the regression model’s, an expected consequence of diffusion models optimizing a different objective, and the iterative denoising process makes inference several times slower, a cost that grows further when generating ensembles. Analysis of the correction fields revealed that the diffusion model learns spatially organized adjustments concentrated over meteorologically active regions such as the Qinghai-Tibet Plateau, but also over-corrects in some smooth areas, suggesting that region-aware or variable-dependent correction strategies could improve future versions. The uncertainty estimates, while empirically useful, do not yet correspond to physically grounded measures of atmospheric predictability.

Even so, the implications are profound. A data-driven pipeline that ingests coarse global forecasts, whether from traditional numerical models or from rapidly emerging AI weather systems, and emits 3 kilometer regional forecasts that rival or beat an operational mesoscale model could reshape how national weather services deliver local predictions for agriculture, energy, transportation, and disaster risk. With the model code and sample data released on Zenodo, and with promising directions including fine-tuning on operational data, physically grounded uncertainty quantification, and interpretable deep learning, the study signals that generative AI is no longer just painting pretty pictures of the atmosphere. It is learning to fill in the fine print of the weather itself, one denoising step at a time.

Subject of Research: Diffusion-based statistical downscaling of weather forecasts to 3 km resolution over China

Article Title: China regional 3 km downscaling based on residual Corrective Diffusion model

Article References: Sun, H., Jing, H., Dai, Z., Xiao, S., Xue, W., Sun, J., & Lu, Q. (2026). China regional 3 km downscaling based on residual Corrective Diffusion model. Geoscientific Model Development, 19(18), 9235-9262. https://doi.org/10.5194/gmd-19-9235-2026

Image Credits: AI Generated

DOI: 10.5194/gmd-19-9235-2026

Keywords: downscaling, diffusion models, CorrDiff, weather forecasting, deep learning, numerical weather prediction, China, radar reflectivity, typhoons, ERA5, uncertainty quantification, Geoscientific Model Development

The research team, led by Jia Sun of the First Institute of Oceanography under China’s Ministry of Natural Resources, examined tropical cyclone activity across the western North Pacific from 1980 to 2025. This basin is the most active tropical cyclone region on Earth, generating more named storms in an average year than any other ocean basin, and it is the source of nearly every typhoon that threatens eastern China, the Korean Peninsula, and the Japanese archipelago. By tracking where storms formed, the paths they followed, and the points where they made landfall, the researchers assembled one of the most detailed pictures yet of how the region’s coastal hazard landscape is changing.

The central result is a redistribution rather than a simple intensification. South of 20 degrees north latitude, the likelihood that a tropical cyclone would make landfall declined over the study period. North of 30 degrees north, the likelihood increased. That northern band broadly encompasses parts of eastern China, the Korean Peninsula, and Japan, meaning that communities which historically sat at the margins of the typhoon belt are now finding themselves closer to its center of gravity. At the same time, storms have tended to form farther to the northwest, to follow tracks that are both more landward and more poleward, and to move modestly faster along their journeys toward the coast.

One of the most consequential numbers in the study concerns time. The average duration that landfalling cyclones spent over the ocean fell from roughly 5.6 days in the 1980 to 2015 period to about 4.7 days in 2016 to 2025. That difference of nearly a full day may sound abstract, but it has direct operational meaning. A storm that forms closer to shore and races toward the coast more quickly compresses the window available for forecasters to issue warnings, for emergency managers to position resources, and for residents to evacuate or shelter. In densely populated coastal megacities, where evacuation logistics are measured in hours and days, a shortened lead time can translate directly into greater human risk even if the storm itself is no stronger than its predecessors.

Co-corresponding author Guihua Wang of Fudan University emphasized that this changing tempo demands a change in mindset. Preparedness plans, she noted, often lean heavily on past experience, on the assumption that future storms will behave like the ones recorded in historical archives. The new findings show why that assumption is increasingly fragile. Coastal risk assessments and early warning systems, particularly in northern East Asia, must now account for shifting cyclone routes and shrinking travel times, not merely for changes in storm frequency or intensity.

Notably, the study did not detect a significant long-term increase in the maximum intensity that these cyclones achieved. In other words, the strongest storms are not steadily getting stronger in this dataset. That distinction matters for how the risk should be understood and communicated. The evolving threat appears to be driven mainly by where tropical cyclones travel and where they come ashore, rather than by a uniform ramping up of their peak winds. For a coastal planner, this reframes the problem: the question is not only how powerful the next typhoon might be, but where it is increasingly likely to arrive.

What physical mechanisms could be steering storms toward higher latitudes? The researchers point to changes in the large-scale winds over the Pacific, which act as a steering current for tropical cyclones. These environmental flow patterns appear to be guiding more storms on westward and poleward trajectories and toward land. Shifts in the subtropical ridge, the monsoon trough, and the background circulation of the western North Pacific can all alter the preferred corridors along which cyclones move, and the observed trends are consistent with such a steering-driven explanation. The storms are, in effect, being redirected by the atmospheric highways they travel.

The authors are careful, however, not to attribute the entire pattern to long-term climate change. The observed redistribution reflects a combination of forced climate trends and natural climate variability, including multidecadal oscillations in the Pacific system that can redistribute storm activity for years or even decades at a time. Separating the anthropogenic signal from natural swings will require longer observational records and targeted climate model experiments designed to isolate the influence of rising greenhouse gases on steering flows and storm tracks. Until that attribution work matures, the safest interpretation is that the northward shift is real and consequential, even if its ultimate causes remain partly open.

The practical implications for coastal communities are immediate. Populations and infrastructure in southern regions that have long been the traditional focus of typhoon preparedness may face somewhat reduced exposure, while cities and provinces farther north, some of which have comparatively less hardened defenses and less deeply ingrained cyclone cultures, may need to plan for a hazard that was once rare on their doorstep. Building codes, drainage systems, evacuation routes, and insurance models calibrated to historical landfall statistics may all understate the risk now emerging in places like the Korean Peninsula and northern Japan. Yesterday’s tropical cyclone map, as the study’s authors put it, may simply not describe tomorrow’s risk.

As the planet continues to warm, understanding how the geography of tropical cyclone hazard evolves will be essential for protecting hundreds of millions of people along the world’s most storm-exposed coastlines. This study adds a crucial nuance to that picture: the danger is not uniformly growing, but it is moving. For the communities of northern East Asia, the message is to watch not only how strong the next storm may become, but where it is increasingly likely to go, and how much warning time it may leave behind as the risk shifts steadily northward.

Subject of Research: Poleward redistribution of tropical cyclone landfall hazards along East Asian coasts

Article Title: East Asia's tropical cyclone risk is moving steadily north

Article References: East Asia's tropical cyclone risk is moving steadily north. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: tropical cyclones, typhoons, East Asia, western North Pacific, landfall, poleward shift, climate change, storm tracks, early warning systems, coastal risk, steering winds, natural variability

The spatial structure of the 2025 anomalies was as noteworthy as their magnitude. Rather than a uniform wetting, the year displayed a striking north-south contrast across central and eastern China, with conditions skewed toward a wetter north and a drier south. This dipole pattern matters enormously for water resource management, agriculture and flood defense, because it concentrates hydrological stress in different regions at different times. In the north, above-normal rainfall swollen by record-breaking seasonal rain episodes filled reservoirs and saturated soils, while parts of the south faced periods of deficit that required careful allocation of irrigation supplies. Climate scientists note that such reversed or amplified precipitation gradients are consistent with the kind of circulation disruptions expected as the East Asian summer monsoon responds to a warming background climate, though attribution of any single year’s pattern requires careful analysis beyond the scope of the annual summary.

Temperature statistics for the year were dominated by the persistence of heat rather than a single spectacular spike. The annual count of high-temperature days, a metric that tracks how often stations exceeded defined heat thresholds, climbed to a new historical maximum, underscoring that 2025 was not merely warm on average but repeatedly hot in ways that stress human health, power grids and crop development. Central and eastern China endured the fourth-strongest large-scale heatwave since systematic records began in 1961, an event distinguished by its spatial footprint and intensity. Compounding the summer burden, autumn brought so-called “autumn tiger” episodes, a colloquial Chinese term for unseasonably hot spells that arrive after the nominal end of summer and extend the season of heat stress well into what should be a cooling period. The combination of an exceptional summer heatwave and lingering autumn warmth pushed cumulative heat exposure metrics to levels without precedent in the observational archive.

The precipitation side of the ledger was equally remarkable, with multiple seasonal benchmarks falling in the same year. The rainfall amount and duration of the North China rainy season both set new records, a doubly significant outcome because the length of the rainy season governs how long soils remain saturated and how sustained flood risk persists across the densely populated North China Plain. Separately, the West China autumn rain phenomenon, a characteristic autumnal rainfall regime over the country’s western interior, also reached record intensity, and national autumn precipitation as a whole established a new high. For three distinct seasonal precipitation measures to break records simultaneously is unusual and points to a persistently anomalous circulation configuration during the second half of the year, with moisture transport pathways repeatedly directed toward regions where they produced prolonged, heavy and impactful rainfall.

The consequences of this wet, energetic year were felt most acutely during the summer, when exceptionally severe rainstorm processes struck North China, Northeast China and Inner Mongolia. These episodes broke daily and cumulative rainfall records at numerous individual stations, overwhelming urban drainage systems and rural flood defenses alike, and ultimately triggered regional major flooding across the Haihe River Basin, one of China’s principal river systems that drains the Beijing-Tianjin-Hebei heartland. The Haihe Basin has a long and well-documented history of catastrophic floods, and modern water infrastructure was engineered with historical maxima in mind; rainfall events that rewrite those maxima test the assumptions embedded in decades of hydraulic planning. Long-lasting and intense West China autumn rain added a second wave of agricultural disruption, interfering with autumn harvesting operations and delaying winter-wheat sowing, the planting window on which the following year’s wheat harvest depends. Farmers in affected areas faced the difficult choice between harvesting waterlogged fields and risking grain quality losses.

Tropical cyclone activity also ran above normal on both sides of the life-cycle ledger, with the number of storms forming in the western North Pacific and the number making landfall on the Chinese coast both exceeding their climatological averages. Autumn typhoons successively impacted South China, and several systems struck during national holiday periods, when travel surges, crowded tourist destinations and family gatherings dramatically increase the population exposed to coastal and inland hazards. The report highlights that these holiday-timed landfalls carried elevated disaster risk, because evacuation logistics, transportation capacity and emergency communications are all strained when hazards coincide with peak seasonal migration. The concurrence of an active typhoon season with the record autumn rainfall regime illustrates how multiple hazard streams can overlap within a single season, compounding rather than merely adding to regional impacts.

Not every region and season told the same story, and the report is careful to document the counterpoints to the dominant warm-wet narrative. Meteorological drought, defined by sustained precipitation deficits and elevated evaporative demand, displayed prominent regional and sub-seasonal variations, appearing and receding across different areas on timescales short enough to complicate drought monitoring and response. Cold-wave outbreaks occurred more frequently than the long-term norm, a reminder that a warming climate does not eliminate severe winter intrusions and that mid-latitude circulation variability can still deliver sharp, damaging cold snaps. Severe convective weather, including the short-lived but violent storms that produce damaging winds, hail and tornadoes, was active nationwide, and the national total of gale days reached its highest level since 1991. Spring sand-dust events over northern China were also more frequent than the long-term average, linking the year’s weather to the condition of arid and semi-arid source regions upwind.

One of the most consequential findings of the assessment concerns the bottom line of disaster economics. Despite the sheer frequency of extreme weather and climate events through 2025, the overall losses attributed to meteorological disasters came in below the 2015-2024 decadal average. The authors of the report do not interpret this as evidence that the hazards were mild; rather, the outcome reflects the cumulative effect of China’s sustained investment in early warning systems, forecast skill, disaster preparedness, infrastructure hardening and coordinated emergency response, which together reduce the vulnerability that converts a hazardous event into a catastrophic loss. The distinction between hazard frequency and disaster impact is a central theme of modern climate risk science, and 2025 offered a vivid national-scale demonstration that improved forecasting and preparedness can bend the loss curve even as the hazard curve rises.

Taken together, the 2025 assessment contributes to a growing body of evidence that China’s climate is shifting toward conditions in which record warmth, intense precipitation and compound seasonal extremes are no longer rare curiosities but recurring features of the national climate. The annual report series itself, now spanning eight consecutive years in a peer-reviewed international journal, provides a consistent methodological framework that allows year-to-year comparison and long-term trend detection, a resource that climate researchers, water managers, agricultural planners and policy makers increasingly rely upon. As the global mean temperature continues its upward trajectory, the report’s documentation of tied temperature records, simultaneous seasonal precipitation records and a rising count of heat days offers a granular, regionally detailed picture of what a warming world looks like in practice for one of the most populous and economically significant nations on Earth. The full findings are available in Atmospheric and Oceanic Science Letters, published by the Institute of Atmospheric Physics at the Chinese Academy of Sciences.

Subject of Research: The state of China's climate in 2025, including record warmth, precipitation extremes and high-impact weather events.

Article Title: China's 2025 climate: Distinct warm‑wet conditions and multiple new climate records

Article References: China's 2025 climate: Distinct warm‑wet conditions and multiple new climate records. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: China climate 2025, record temperature, precipitation records, Haihe River Basin floods, heatwave, typhoons, autumn rain, drought, cold waves, National Climate Center, Atmospheric and Oceanic Science Letters, climate monitoring

A new study led by The University of Hong Kong together with Imperial College London claims to have found the missing piece. Writing in Nature Geoscience, Professor Dazhi Xi and Dr Jianan Chen of the HKU Department of Earth and Planetary Sciences, alongside Professor Ralf Toumi of Imperial College, report that a warming atmosphere does not simply become moister in a way that guarantees more rain. Instead, it also becomes drier in a specific, measurable sense that suppresses rainfall, effectively acting as a brake on tropical cyclone precipitation. This constraining influence, the researchers argue, is strong enough in many climate models to offset much of the rainfall intensification driven by stronger storms.

The key to the finding lies in a quantity known as the column saturation deficit, which measures the gap between the actual amount of water vapour present in an atmospheric column and the amount that would be present at complete saturation, the threshold at which condensation and precipitation proceed freely. Rain normally forms when water vapour condenses into cloud droplets, coalesces into raindrops and falls to the surface. Under global warming, however, the moisture-holding capacity of the atmosphere rises exponentially with temperature, following the Clausius-Clapeyron relationship. Even if relative humidity were to remain perfectly constant, the absolute distance to saturation widens substantially as temperatures climb. In this precise technical sense, the air becomes drier: it can hold far more vapour than it actually contains.

That widening unsaturation has two distinct consequences for a typhoon’s rain production. The first is evaporation on the way down. Raindrops that condense at high altitudes inside a tropical cyclone must descend through the lower and middle troposphere before reaching the ground, and in a warmer climate the surrounding air in those layers is further from saturation. Dry environmental air efficiently strips moisture from falling drops, evaporating them before they can reach the surface and robbing the storm of rainfall that its clouds have already produced.

The second consequence operates at the source. Tropical cyclones are not closed systems; they continuously draw in environmental air through their circulation. When this drier air is entrained into the storm’s updrafts, it dilutes the moisture supply feeding condensation, suppressing cloud and rain formation from the outset. Together, these two effects, the evaporation of falling rain and the inhibition of new condensation, reduce what atmospheric scientists call precipitation efficiency: the fraction of the water vapour flowing through a storm that actually ends up as rainfall at the ground.

To establish the result, the team analysed climate model simulations, satellite observations and reanalysis data, and found a consistent signal: as the climate warms, tropical cyclones become measurably less efficient at converting available moisture into rain. Across the datasets examined, greater atmospheric dryness correlates with lower rainfall efficiency, a robust negative relationship that persisted despite differences in models and observational products. The researchers emphasise that this correlation is not a modelling artefact but a thermodynamic constraint that any credible projection of future cyclone rainfall must incorporate.

Beyond diagnosing the mechanism, the study proposes a unified framework for thinking about tropical cyclone rainfall. In this view, total rainfall depends on three factors rather than two: the intensity of the storm, the amount of water vapour available in the atmosphere, and the precipitation efficiency with which that vapour is converted into rain. The third factor is itself shaped by two opposing influences in a warming climate. Stronger storms tend to raise precipitation efficiency, while increasing atmospheric dryness tends to depress it. The framework does not guarantee that efficiency will fall everywhere; if future storm intensification proves strong enough, it could outweigh the suppressive effect of dryness in some models and regions. But the analysis shows that in a number of existing climate projections, the dryness effect dominates, which explains why their rainfall increases fall below thermodynamic expectations.

The practical stakes are considerable. Coastal communities, disaster management agencies and infrastructure planners rely on projections of hurricane and typhoon rainfall to design flood defences, plan evacuations and build long-term climate resilience. If conventional expectations systematically overstate future rainfall, flood protection may be designed around numbers that misrepresent the true risk; if they understate it in regions where storm intensification wins out, communities could be caught unprepared. By explicitly accounting for the dryness mechanism, the new framework offers a way to narrow this uncertainty and produce more defensible rainfall and flood-risk assessments for adaptation planning.

The authors are careful to note the limits of the current work. Global climate models do not fully resolve some of the fine-scale physical processes involved in cloud formation, rain evaporation and the entrainment of dry air into storm circulations, so future high-resolution simulations will be needed to refine the quantitative projections. Nevertheless, the consistent negative relationship between atmospheric dryness and rainfall efficiency across multiple independent datasets gives the finding unusual robustness for a study of projected future behaviour. For a field that has long assumed that a moister atmosphere automatically means a rainnier future for the world’s most destructive storms, the message is a sobering one: the same warming that loads the atmosphere with water also opens a gap that keeps much of that water from ever falling as rain.

Subject of Research: Thermodynamic constraints of atmospheric dryness on future tropical cyclone rainfall under climate warming

Article Title: HKU study reveals how atmospheric dryness constrains typhoon rainfall leading to lower-than-expected increases

Article References: HKU study reveals how atmospheric dryness constrains typhoon rainfall leading to lower-than-expected increases. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: tropical cyclones, typhoons, atmospheric dryness, climate change, precipitation efficiency, column saturation deficit, rainfall projection, flood risk, Nature Geoscience, thermodynamics, climate models, University of Hong Kong