The HRRR Meteorology, Energy, and Transmission (MET) Toolkit, described in a recent paper in Wind Energy Science, represents a fundamental shift in how atmospheric data can be repackaged for energy applications. Rather than running expensive new simulations from scratch, the researchers at the National Laboratory of the Rockies took the National Oceanic and Atmospheric Administration’s (NOAA) High-Resolution Rapid Refresh (HRRR) model — an operational weather forecasting system that updates hourly across the contiguous United States — and transformed it into a streamlined, energy-ready dataset. The result is a continuous record spanning 2015 through 2025, with plans to extend it annually, all delivered on the same uniform 2-kilometer grid that made the legacy WIND Toolkit so widely adopted.

The challenge the team faced was not a lack of data but rather the format in which it existed. The native HRRR archive stores information in GRIB files on irregular hybrid vertical levels that vary in both space and time, making it nearly impenetrable for energy modelers accustomed to working with standardized datasets. Modern wind turbines, whose rotor hubs sit at heights of roughly 80 to 140 meters above ground, require wind speed measurements at precisely those elevations — something the HRRR’s native vertical structure did not readily provide. The toolkit solves this by linearly interpolating wind speeds from the model’s time-varying native levels to fixed heights above ground, matching the same vertical levels used in the WIND Toolkit: from 10 meters all the way up to 500 meters.

Spatial regridding posed another technical hurdle. The HRRR runs at a native 3-kilometer horizontal resolution, while the legacy WIND Toolkit used a 2-kilometer grid. To maintain continuity with a decade of industry-standard modeling, the researchers employed inverse distance-weighted interpolation with the four nearest neighbors, carefully preserving spatial variability while aligning the projections. The team acknowledges that interpolating from a coarser to a finer grid inherently smooths the data without generating new physical information. Spectral analyses indicate that the effective resolution of the underlying model limits true physical detail to scales of roughly 20 kilometers, regardless of the output grid spacing.

Temporal gaps in the public HRRR archive — approximately 5.1 percent of hours between 2015 and 2018 — required a hierarchical gap-filling strategy. Missing time steps were populated using forecasts from previous initialization hours, then pressure-level data with fewer gaps, and finally linear interpolation for the remaining 25 hours of missing data. The resulting dataset includes wind speed and direction, temperature, pressure, humidity, precipitation, friction velocity, boundary layer height, and heat fluxes, all as hourly instantaneous values.

To evaluate whether this repackaged operational data could truly replace computationally intensive dedicated simulations, the researchers conducted the first comprehensive, multi-model validation of wind resource datasets across the contiguous United States at utility-scale hub heights. They compared the HRRR MET Toolkit against the WIND Toolkit Long-term Ensemble Dataset (WTK-LED), its Climate variant, the 2023 National Offshore Wind (NOW-23) dataset, and an experimental bias-corrected version called BC-HRRR. Validation drew on tall meteorological towers and lidar instruments spanning diverse geographies, from the Gulf of Mexico to the Great Plains to complex mountain terrain in the American West.

The results were striking. Averaged across all stations, the HRRR MET Toolkit achieved a Pearson correlation coefficient of 0.82 with observed hub-height wind speeds, significantly outperforming WTK-LED at 0.67 and WTK-LED Climate at just 0.39. The centered root-mean-square error for the HRRR-based products was approximately 2.1 meters per second, compared to 3.3 meters per second for WTK-LED and 4.4 meters per second for WTK-LED Climate. Perhaps most importantly, the average hub-height wind speed bias for the HRRR MET Toolkit was just 0.10 meters per second, compared to 0.82 meters per second for WTK-LED — an eightfold reduction in systematic error.

For offshore environments, the NOW-23 dataset demonstrated the strongest performance, achieving the lowest mean bias of 0.02 meters per second and the highest correlation of 0.87, validating its regional tuning strategy in which model configurations were selected separately for each offshore region. The HRRR-based products performed robustly offshore as well, closely trailing NOW-23, while the WTK-LED variants showed significant degradation in marine environments.

The researchers also tested whether applying a statistical bias correction — a quantile mapping technique that anchors the HRRR wind speed distribution to the legacy WIND Toolkit baseline — could further improve accuracy. The answer, surprisingly, was largely no. While the bias correction mitigated some negative biases in offshore environments, it provided marginal benefits on land and actually exacerbated positive wind speed biases in complex terrain such as the western United States and the Appalachian Mountains. The team concluded that the native HRRR physics are already sufficiently robust for energy applications without the added complexity of statistical post-processing.

The implications for the rapidly evolving US energy landscape are substantial. With wind energy generation exceeding 10 percent of annual electricity generation in the United States as of 2022, and with grid planners increasingly tasked with integrating variable renewable resources at unprecedented scale, the availability of accurate, continuously updated, high-resolution atmospheric data has never been more critical. The HRRR MET Toolkit is publicly available through the National Laboratory of the Rockies’ Open Energy Data Initiative and the Wind Resource Database, offering both web-based visualization tools and programmatic access for bulk analysis. As the underlying HRRR model continues to improve through ongoing NOAA development — with correlation improving from approximately 0.67 in early versions to 0.8 in the latest iteration — the toolkit is positioned to become the new standard for power system studies across the country, bridging the critical gap between the historical record that ended in 2013 and the operational forecasting capabilities of today.

Subject of Research: Development and validation of the HRRR MET Toolkit for high-resolution wind resource assessment in US power system planning

Article Title: The HRRR Meteorology, Energy, and Transmission (MET) Toolkit: advancing high-resolution atmospheric data for contiguous US energy applications

Article References: Bodini, N., Maric, E., Egerer, U., Buster, G., Lavin, L., Pinchuk, P., Benton, B., & Turner, D. D. (2026). The HRRR Meteorology, Energy, and Transmission (MET) Toolkit: advancing high-resolution atmospheric data for contiguous US energy applications. Wind Energy Science, 11(10), 3919-3948. https://doi.org/10.5194/wes-11-3919-2026

Image Credits: AI Generated

DOI: 10.5194/wes-11-3919-2026

Keywords: HRRR MET Toolkit, wind energy, NOAA, weather modeling, grid integration, wind resource assessment, WIND Toolkit, bias correction, renewable energy, power system planning, atmospheric data, lidar validation

The research team, led by Shaidatul Azdawiyah Abdul Talib of the Malaysian Agricultural Research and Development Institute (MARDI) together with colleagues at Universiti Kebangsaan Malaysia, evaluated projected changes in mean temperature, precipitation, thermal threshold exceedance, and drought behavior using a multi-model ensemble under the RCP 4.5 intermediate-emissions pathway and the RCP 8.5 high-emissions pathway. The baseline period for comparison spans 1984 to 2014, providing three decades of observed climate against which the modeled future could be measured. Before feeding the raw climate model output into their analysis, the team applied statistical bias correction, a crucial step in regional climate studies because global models often misrepresent local temperature and rainfall distributions, particularly in the complex tropical maritime environment of Southeast Asia.

The headline temperature findings are stark. Mean annual temperature across Peninsular Malaysia is projected to rise by approximately 1.36 degrees Celsius under RCP 4.5 and 1.55 degrees Celsius under RCP 8.5 by mid-century relative to the 1984 to 2014 baseline. While a difference of roughly two-tenths of a degree between the two scenarios might sound trivial, the study found that the divergence between emission pathways becomes increasingly pronounced after 2030, meaning that policy choices made this decade will meaningfully shape how hot the peninsula becomes in the decades that follow. The warming signal is statistically significant under both scenarios, indicating that the projected temperature increase is not an artifact of model noise or natural variability.

Perhaps the most striking result concerns heat stress days, defined in the study as days when the daily maximum temperature exceeds 34 degrees Celsius. During the baseline period, Peninsular Malaysia experienced an average of 42 such days per year. By mid-century, that figure is projected to climb to 113 days per year under RCP 4.5 and 132 days per year under RCP 8.5. In practical terms, this means that under the high-emissions pathway, roughly one out of every three days of the year would exceed a temperature threshold associated with significant heat stress for humans, livestock, and crops. The near-tripling of heat stress days, despite what the authors describe as only moderate warming in mean temperature, illustrates a well-established principle in climate science: extremes respond more sensitively to warming than averages do. A modest shift in the mean of a temperature distribution pushes a disproportionately large fraction of days past fixed threshold values, and the study’s results bear this out dramatically for the Malaysian context.

The implications for agriculture are considerable. Malaysia is a major producer of palm oil, rice, and tropical fruits, and research cited in the study, including work on heat stress during flowering in cereal crops, has shown that even brief episodes of high temperature during critical reproductive stages can sharply reduce yields. Rice, the region’s staple crop, is particularly vulnerable: physiological studies have demonstrated that temperatures above roughly 33 to 35 degrees Celsius during flowering cause spikelet sterility and grain quality deterioration. A tripling of days above 34 degrees Celsius would therefore translate directly into heightened risk for the peninsula’s farming systems, particularly in the rice-growing regions of the northwest such as Kedah and Perlis, as well as for the plantation sector that underpins much of the national economy.

In contrast to the temperature findings, the projections for mean annual precipitation are more nuanced. The study found that changes in total annual rainfall remain modest and spatially heterogeneous, with some areas projected to become slightly wetter and others slightly drier, but without a coherent peninsula-wide trend toward systematic drying. This result aligns with a broader body of Southeast Asian climate research, including studies based on the CORDEX-Southeast Asia regional downscaling framework and CMIP6 multi-model ensembles, which have generally found that mean rainfall changes in the maritime continent are smaller and more spatially variable than temperature changes. The physical explanation lies in the region’s position within the global tropical convergence zone, where rainfall is governed by complex monsoon dynamics, sea surface temperature patterns, and local convection rather than by simple thermodynamic scaling alone.

However, the absence of a large change in mean rainfall does not mean the water cycle is stable. When the team applied the Standardized Precipitation Index, or SPI, a widely used drought metric originally developed by McKee and colleagues in 1993, they found evidence of intensifying hydroclimatic variability. The frequency of moderate drought conditions, defined as periods when the SPI falls to or below minus 1.0, is projected to rise from 2.77 percent of months during the baseline period to 14.72 percent by 2041 to 2050 under RCP 8.5. That represents roughly a fivefold increase in the share of time spent in moderate drought, even though average annual rainfall barely changes. The mechanism is well understood in the climate literature: as the atmosphere warms, it holds more water vapor, roughly 7 percent more per degree Celsius of warming, which intensifies both extremes of the hydrological cycle. Wet events become wetter, dry spells become more pronounced between them, and rainfall arrives in heavier bursts separated by longer dry gaps. The study’s SPI results suggest Peninsular Malaysia is entering precisely this regime.

The authors identify this as the key finding of their work: future climate risk over Peninsular Malaysia arises primarily from rising temperatures and amplified precipitation variability rather than from a systematic decline in mean annual rainfall. This distinction matters enormously for adaptation planning. Water resource managers, farmers, and policymakers who focus solely on average rainfall totals would conclude that the peninsula’s water security is largely intact, and would be dangerously wrong. A climate that delivers the same annual total but concentrates it into fewer, more intense events, punctuated by more frequent drought episodes, poses fundamentally different challenges for reservoir management, flood control, hydropower scheduling, and crop calendars than a climate with stable, evenly distributed rainfall.

The study’s methodology reflects the current state of the art in regional climate projection science. The team employed multivariate quantile mapping bias correction, an approach refined by Cannon and colleagues that adjusts climate model output so that its statistical distribution matches observed data across multiple variables simultaneously, preserving the physical relationships between temperature and precipitation rather than correcting each in isolation. Trend detection was carried out using the Mann-Kendall test with modifications for autocorrelated data, following the methods of Hamed and Rao and of Yue and Wang, which guard against falsely detecting trends in serially correlated hydrological time series. Model performance was evaluated using standard metrics including the Taylor diagram framework introduced by Taylor in 2001, which summarizes how well each model reproduces observed patterns of correlation, variance, and root-mean-square error in a single graphical comparison.

The broader context for these findings is sobering. Recent analyses, including work by Hansen and colleagues published in 2025, argue that global warming has accelerated beyond what many public assessments have conveyed, and regional studies across Asia, from India to Thailand, have documented rising heatwave frequency and intensifying precipitation extremes. The Malaysian results fit squarely within this global pattern while adding regionally specific detail for a nation whose economy, food systems, and population are acutely exposed to tropical heat and hydrological whiplash. The study’s projection that the peninsula will transition toward what the authors describe as a more extreme-prone climate regime, characterized by intensified thermal stress and increasing precipitation variability under continued warming, is a warning that applies with particular force to the decades between now and 2050. How severe that transition becomes depends substantially on which emissions pathway the world follows after 2030, the point at which the study found the scenarios beginning to meaningfully diverge. For Peninsular Malaysia’s 33 million residents, its farmers, and its water managers, the message of the research is clear: the threat is not less rain, but hotter days and a rainfall rhythm that is becoming steadily less reliable.

Subject of Research: Projected changes in temperature, precipitation, and drought variability across Peninsular Malaysia under multi-model ensemble climate scenarios

Article Title: Hydroclimatic variability and intensifying thermal extremes in Peninsular Malaysia under multi-model ensemble climate projections

Article References: Abdul Talib, S. A., Idris, W. M. R., Ju Neng, L., Lihan, T., & Abdul Rasid, M. Z. (2026). Hydroclimatic variability and intensifying thermal extremes in Peninsular Malaysia under multi-model ensemble climate projections. Theoretical and Applied Climatology, 157(11), Article 720. https://doi.org/10.1007/s00704-026-06577-1

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06577-1

Keywords: climate change, Peninsular Malaysia, heat stress days, RCP 4.5, RCP 8.5, Standardized Precipitation Index, drought, hydroclimatic variability, bias correction, multi-model ensemble, Southeast Asia, thermal extremes

The research, carried out by Diksha and Soora Naresh Kumar of the Division of Environmental Sciences at ICAR-Indian Agricultural Research Institute in New Delhi, was funded through India’s National Mission for Sustaining the Himalayan Ecosystem. Its central challenge was one that has long frustrated atmospheric scientists: the Himalaya is a region of extreme topographic complexity, where deep valleys, steep slopes and sharp elevation gradients make ground-based monitoring sparse and satellite retrievals unreliable. Mountains distort the assumptions built into standard satellite products, and the few air quality stations that exist are concentrated in towns and cities, leaving vast stretches of the range effectively unmeasured.

To overcome this, the team fused data from multiple satellite instruments with a state-of-the-art atmospheric model. Aerosol Optical Depth measurements from the Moderate Resolution Imaging Spectroradiometer, processed through the Multi-Angle Implementation of Atmospheric Correction algorithm, provided the aerosol backbone, while the Sentinel-5P TROPOMI instrument supplied column measurements of ozone, carbon monoxide, sulfur dioxide and nitrogen dioxide. These observations were combined with reanalysis data from the Goddard Earth Observing System-Composition Forecast model, which allowed the researchers to estimate near-surface concentrations from satellite-observed columns using modelled vertical scaling ratios and conversions based on other significant atmospheric variables.

The processing pipeline was as important as the data itself. Monthly composites were generated on the Google Earth Engine platform, while annual mosaics were enhanced to a one-kilometre resolution using ArcGIS Pro. Crucially, the raw satellite estimates were then bias-corrected against ground station measurements using three machine learning approaches: Gradient Boosting, Random Forest and Linear Regression. The corrected estimates showed strong agreement with observed data, with correlation coefficients ranging from 0.85 to 0.95, a level of performance that gives considerable confidence in the spatial patterns the study reveals across terrain where no ground instruments exist.

The choice of the 2020 lockdown as a benchmark period gives the analysis an unusual analytical power. The COVID-19 restrictions of 2020 produced an abrupt, near-uniform suppression of traffic, industry and construction across India, creating something close to a natural experiment in emission reduction. Earlier work, including studies of the Indo-Gangetic Plain’s influence on Himalayan air quality, documented how the lockdown visibly cleared the air even at high-altitude sites. By anchoring their timeline to that exceptional year, the researchers could measure precisely how much of the subsequent deterioration represents a rebound of economic activity rather than a continuation of longer-term trends.

The answer, in much of the region, is a great deal. In the Himalayan foothills, levels of carbon monoxide, PM2.5, ozone, sulfur dioxide and nitrogen dioxide all increased in the post-pandemic period. Rate-of-change analyses pinpoint the acceleration after 2021, with the steepest rises concentrated in the valleys of Jammu and Kashmir and in the foothills of Uttarakhand and Himachal Pradesh. These are precisely the areas where post-pandemic recovery of economic activity, expanding transport and growing industrial presence drove pollutant concentrations upward over the study period. The valleys, which trap air under temperature inversions especially in winter, act as natural holding chambers for whatever is emitted into them.

Not every pollutant followed the same trajectory, and the exceptions are scientifically revealing. Oxides of nitrogen, reported as NOx, remained comparatively stable in high-altitude regions throughout the study period. This stability suggests that the most remote parts of the range, far from roads and industry, have so far been buffered from the rebound affecting the lowlands, and that the pollution problem in the high Himalaya is driven less by local emissions than by transport of pollutants from the plains. The Indo-Gangetic Plain to the south is one of the most densely populated and polluted regions on Earth, and its plumes routinely ride upslope winds into the mountains, a dynamic that earlier studies of black carbon on Himalayan glaciers have documented in detail.

The distinction between foothill and high-altitude behaviour matters enormously for what happens next. The Himalaya is often called the water tower of Asia, and its glaciers feed rivers that sustain hundreds of millions of people. Deposited dark particles, whether black carbon or mineral dust mixed with combustion products, darken snow and ice surfaces and accelerate melting, linking air quality directly to water security. Aerosol pollution also alters regional radiation budgets and monsoon dynamics. A gradual increase of aerosol pollution over the so-called Third Pole over the past four decades has already been flagged as a driver of regional climate change, and the new study’s finding that air quality in the Indian Himalayan Region is worsening over time adds urgency to that warning.

There are also direct public health implications. Fine particulate matter and the gaseous pollutants tracked in the study are associated with respiratory and cardiovascular disease, adverse birth outcomes and aggravated asthma, with effects documented across age groups from children to the elderly. Mountain communities, often assumed to enjoy clean air, may face exposure patterns that mirror those of the plains, particularly in valley towns where tourism traffic, diesel generators, brick kilns and heating emissions accumulate under inversions. The foothill cities of Uttarakhand and Himachal Pradesh have expanded rapidly, and studies of Dehradun have already linked urbanisation to deteriorating air quality and changing temperature dynamics. The new regional maps put those local findings into a range-wide context.

Methodologically, the study demonstrates how far satellite-based air quality assessment has come. The combination of TROPOMI’s trace gas columns, MAIAC-corrected MODIS aerosol data, GEOS-CF reanalysis for vertical structure, and machine learning bias correction against sparse ground stations now yields surface concentration estimates at one-kilometre resolution even over terrain that defeats conventional monitoring. Comparable approaches have been applied successfully in Delhi, across the Arabian Peninsula, and in Thailand, suggesting the Himalayan workflow could be replicated for other mountain ranges, from the Andes to the Tibetan Plateau, where ground networks are thin but health and climate stakes are high.

What the study cannot yet say, and what the authors’ own framing makes clear, is where the region goes from here. The data end at 2024, and the post-2021 acceleration may continue as infrastructure projects, hydropower construction and tourism expand deeper into the mountains, or it may moderate as cleaner vehicle standards and cooking fuel transitions take hold. What the analysis does establish is a baseline and a method: a reproducible, validated way to watch pollution evolve across one of the planet’s youngest and most fragile mountain systems, season by season and valley by valley. For a region whose air was once its proudest asset, that watching has now become a necessity, and the satellites are already keeping score.

Subject of Research: Spatiotemporal variation of air pollutants across the Indian Himalayan Region from 2020 to 2024 using satellite data and machine learning bias correction

Article Title: High altitude alerts: temporal and spatial variations in air quality over the Indian Himalaya Region

Article References: Diksha, & Kumar, S. N. (2026). High altitude alerts: temporal and spatial variations in air quality over the Indian Himalaya Region. Air Quality, Atmosphere & Health, 19(10), Article 237. https://doi.org/10.1007/s11869-026-02122-1

Image Credits: AI Generated

DOI: 10.1007/s11869-026-02122-1

Keywords: air pollution, Indian Himalayan Region, Sentinel-5P TROPOMI, MODIS, machine learning, bias correction, Google Earth Engine, GEOS-CF, PM2.5, nitrogen dioxide, COVID-19 lockdown, Himalayan valleys

The problem begins with scale. To study how climate change has reshaped Alpine catchments since the mid-twentieth century, researchers often use the Weather Research and Forecasting model, or WRF, to dynamically downscale coarse global reanalysis data to a much finer grid. In this study, the team took the Twentieth Century Reanalysis Version 3, a global dataset with a spatial resolution of roughly 100 kilometers, and downscaled it to 2 kilometers with hourly output for the period 1969 to 2015. That is a remarkable achievement, but even 2 kilometers is too coarse to capture the fine-grained reality of a glacierized catchment, where snowfields, rock debris, and ice interact with the atmosphere in ways that vary from one slope to the next.

When the researchers evaluated the WRF output against observations from four meteorological stations spanning nearly 900 meters of elevation in the upper Kauner Valley, they found a pronounced cold bias in the diagnosed 2-meter air temperature at elevations above roughly 2600 meters. The culprit was the model’s coarse representation of snow- and ice-covered surfaces. In the WRF land surface scheme, grid cells above about 2900 meters were classified as permanent Snow or Ice, and the cooling influence of those bright, frozen surfaces bled into the simulated near-surface air temperature. The result was air that was too cold in exactly the places where snowmelt and glacier melt are most sensitive to temperature, a distortion that would ripple through every downstream calculation.

Conventional bias correction, such as empirical quantile mapping, adjusts model output by matching its statistical distribution to that of observations. But that approach requires long, high-quality observational records, and such records are notoriously scarce at high elevations in the Alps. Stations above 2700 meters are few, gappy, and prone to sensor failures, and their statistical properties may not be stationary over decades of rapid warming. Rather than lean on sparse data, Hofmeister and his colleagues turned to the model’s own atmosphere for the answer.

Their method exploits a fundamental property of the atmosphere: temperature generally decreases with height according to a lapse rate. In the WRF simulations, the researchers found that at the lowest atmospheric model levels, temperatures were strongly influenced by the underlying land cover, with snow- and ice-covered grid cells showing markedly colder values and considerable scatter. But higher up, between the first and twentieth vertical levels, the temperature-elevation relationship became almost perfectly linear, with very little scatter. At these heights, the model atmosphere was essentially free of surface contamination, governed by a uniform domain-wide lapse-rate structure. The lapse rate at the first model level averaged minus 0.78 degrees Celsius per 100 meters in a test period, close to the free-atmospheric value of minus 0.65, and it decreased steadily with height until it approached zero at the highest levels.

The correction workflow is a three-step procedure applied to hourly time series. First, the team computed spatial averages of hourly temperature and geopotential height across the 49 grid cells covering roughly 200 square kilometers of the Kauner Valley domain, at the first and twentieth vertical levels. Averaging in space reduces the dependence of the lapse rate on land cover type. Second, they calculated an hourly temperature lapse rate from the difference between the two levels, capturing the sub-daily variability of the atmospheric temperature gradient rather than assuming a fixed value. Third, they used that hourly lapse rate to extrapolate the temperature at the first atmospheric model level, roughly 17 meters above the ground on average, down to any station or catchment elevation. Because the reconstruction starts from the free atmosphere rather than the surface, it sidesteps the surface-related biases entirely. The mean lapse rate derived between the two levels over 1973 to 2015 was minus 0.57 degrees Celsius per 100 meters, slightly steeper in summer than in winter.

Temperature alone was not enough. Because snowmelt and glacier melt models rely on energy balance calculations in which latent heat fluxes play a substantial role, the team also applied a sequential, thermodynamically consistent correction of relative humidity, adjusting moisture to match the corrected temperatures so that the physical relationship between the two variables was preserved. Notably, they left shortwave radiation and wind speed untouched, since physics-based hydrological models such as WaSiM, the Water Flow and Balance Simulation Model used in this study, already account for topographic effects on radiation and can represent wind-driven snow redistribution on their own.

The validation was rigorous and multi-layered. Against station observations, the corrected temperatures showed substantially better agreement than the uncorrected output, with the improvement most dramatic at the highest station, Vernagt Pegel at 2640 meters, where the maximum distance between observed and modeled temperature distributions dropped from 0.13 to 0.03. The consequences for hydrological modeling were striking. When WaSiM was forced with uncorrected WRF temperatures, it could not reproduce the observed dynamics of snow accumulation and melt, glacier mass balance, snow cover, or streamflow: the cold bias suppressed melt so severely that simulated annual streamflow was nearly 50 percent too low, and the simulated glacier mass balance stayed unrealistically positive during a period when Alpine glaciers were actually shrinking. With the corrected temperatures, the model matched the observed mean annual streamflow almost exactly, and the Kling-Gupta efficiency of daily streamflow improved from 0.42 to 0.71. Adding the humidity correction pushed performance further, raising the efficiency to 0.78 and improving the representation of late-summer flows during peak glacier melt, when the Fagge carries its maximum meltwater load.

Crucially for climate research, the correction preserved the long-term warming signal. Comparing the corrected product with homogenized observations at Obergurgl, a station in a neighboring valley with records stretching back to 1851, the team found that both the observations and the corrected WRF temperatures showed a significant warming trend of about 0.05 to 0.06 degrees Celsius per year over 1973 to 2015, and both detected the well-documented abrupt Alpine warming shift in May 1987. A 42-year simulation of the catchment reproduced the observed transition from near-equilibrium glacier conditions in the 1980s to continuous mass loss thereafter, along with a significant shortening of snow cover duration after 1985. Because the method relies on lapse rates derived from the climate model itself rather than on local observations, it is transferable to other glacierized mountain regions where ground data are lacking. As the authors note, future work should test the approach in catchments with different climatic and topographic settings, but the message is already clear: physically consistent corrections of model output can rescue hydrological simulations in the world’s most water-critical landscapes, without erasing the climate signal that makes those simulations worth running in the first place.

Subject of Research: Physically based bias correction of regional climate model temperatures for hydrological simulation in a glacierized Alpine catchment

Article Title: Technical note: A physically based correction of WRF temperature using atmospheric lapse rates for high-resolution hydrological simulations in a glacierized catchment

Article References: Hofmeister, F., Fan, X., Pfeiffer, M., Marzeion, B., Schaefli, B., & Chiogna, G. (2026). Technical note: A physically based correction of WRF temperature using atmospheric lapse rates for high-resolution hydrological simulations in a glacierized catchment. Hydrology and Earth System Sciences, 30(19), 6265-6288. https://doi.org/10.5194/hess-30-6265-2026

Image Credits: AI Generated

DOI: 10.5194/hess-30-6265-2026

Keywords: WRF model, temperature lapse rate, bias correction, glacierized catchment, hydrological modeling, Alps, snowmelt, glacier mass balance, streamflow, regional climate model, relative humidity, climate change

The research, led by Mahmoud Hesham of Port Said University together with colleagues from Cairo University, evaluated daily rainfall records from 23 ground stations spanning the years 2003 to 2024. Against this observational baseline, the team compared five products: the Climate Hazards Group InfraRed Precipitation with Station Data version 2.0, known as CHIRPS; the Climate Prediction Center Morphing Technique, or CMORPH; the PERSIANN-CDR dataset built on artificial neural networks; the Global Precipitation Measurement mission’s Integrated Multi-satellite Retrievals Final Run version 07, abbreviated GPM IMERG-F V07; and ERA5, the fifth-generation atmospheric reanalysis produced by the European Centre for Medium-Range Weather Forecasts. Each of these products estimates rainfall in a fundamentally different way, blending satellite infrared and microwave sensors, gauge data, or numerical weather model output, and each carries its own characteristic strengths and blind spots.

The stakes are considerable. In arid regions, rainfall is rare, localized, and often violent, arriving as short-lived cloudbursts that can trigger flash floods in wadis and urban areas that were designed with little drainage capacity. Hydrological models and flood forecasting systems depend on accurate precipitation inputs, and where gauges are sparse, satellite and reanalysis products are often the only practical source of spatially continuous rainfall information. But these products were largely developed and calibrated in wetter climates, and their performance in hyper-arid environments, where the signal-to-noise ratio of rainfall detection is poor, has long been an open question.

The evaluation produced a nuanced picture rather than a single winner. At the daily and monthly timescales, CHIRPS V2.0 generally delivered the lowest root mean square error, suggesting that its blend of infrared satellite imagery and station data captures the day-to-day rhythm of Egyptian rainfall better than its competitors. When the analysis shifted to yearly totals and maximum yearly values, however, ERA5 and GPM IMERG-F V07 took the lead, indicating that the reanalysis model and the latest generation of the GPM mission are more reliable for long-term water resources assessment and extreme-rainfall analysis. This scale-dependence is a critical finding for practitioners: the best data source for a flash-flood early warning system may not be the best source for planning a decades-long water management strategy.

Event detection proved equally revealing. GPM IMERG-F V07 showed reliable rainfall-event detection across Egypt, with probability of detection values exceeding 0.45 throughout the country. The probability of detection, a standard verification metric that measures the fraction of observed rain events correctly identified, is especially important in arid lands where a missed event is often a missed flood. ERA5 also performed well in detecting rainfall events at several northern stations, where Mediterranean weather systems bring the bulk of Egypt’s modest precipitation. These results suggest that modern satellite retrievals have matured to the point where they can meaningfully supplement, and in some contexts partially substitute for, ground observations in gauge-poor regions.

Detecting that rain fell is only half the battle, however. The next question is how much fell, and here the raw satellite and reanalysis products carry systematic biases that can propagate into hydrological models with serious consequences. To address this, the researchers tested three bias-correction methods of increasing sophistication: linear scaling, a simple multiplicative adjustment of the mean; quantile mapping, a statistical technique that reshapes the entire distribution of estimated rainfall to match the observed distribution; and a dense neural network, a deep learning model capable of learning complex, nonlinear relationships between the estimated rainfall and the true gauge-measured amounts.

The neural network approach was engineered with considerable care. A feedforward dense neural network was optimized separately for each of the 23 stations, with the architecture and training regime tuned individually: the models used between 2 and 6 hidden layers, 16 to 256 neurons per layer, 150 to 300 training epochs, batch sizes ranging from 16 to 128, learning rates between 0.0005 and 0.0050, and dropout rates from 0.0 to 0.2. This station-specific optimization reflects a key insight of modern machine learning practice in the geosciences: rainfall error structures are local phenomena, shaped by topography, distance from the coast, and prevailing weather patterns, and a single global correction model would likely average away the very details that matter most.

The payoff was substantial. Deep-learning-based bias correction reduced error by 33 percent at the daily timescale, 22 percent at the monthly timescale, and 34 percent at the yearly timescale. These are not marginal refinements; a one-third reduction in daily rainfall error can materially change the output of flood forecasting models, which are acutely sensitive to the timing and magnitude of individual rain events. The fact that the largest relative improvement appeared at the yearly scale suggests the neural networks were particularly effective at correcting the cumulative biases that accumulate over seasons and years, precisely the errors that matter for water resources planning in a country where every drop counts.

Perhaps the most important conclusion of the study is a cautionary one: there is no universal best product, and bias correction is not a magic bullet. The best-performing rainfall product varied by location, hydrological zone, and temporal scale, meaning that practitioners in Egypt should select rainfall datasets based on local rainfall conditions and the intended hydrological use. Bias correction, meanwhile, should be applied carefully, because its effectiveness depends on rainfall behavior and timescale. A correction that works brilliantly for Mediterranean coastal stations may fail in the southern desert, and a method tuned for daily extremes may distort long-term averages. This echoes a growing body of literature warning that statistical correction techniques, including machine learning approaches, must be validated for the specific application at hand rather than assumed to be universally beneficial.

For Egypt, the practical implications are immediate. The evaluated and corrected products can serve as complementary rainfall inputs for hydrological modeling, flash-flood forecasting, extreme-rainfall analysis, and water-resources management in arid and gauge-sparse regions, not only in Egypt but across the broader belt of deserts and semi-arid lands where similar data scarcity prevails. As climate change alters rainfall patterns in the Mediterranean and North Africa, and as urban expansion places more people and infrastructure in the path of rare but devastating flash floods, the ability to extract trustworthy rainfall information from satellites, refined by machine learning, becomes an essential tool of resilience. This study provides both a practical guide for choosing among the available datasets and a demonstration that deep learning can meaningfully close the gap between what satellites see and what actually falls from the sky.

Subject of Research: Evaluation and bias correction of satellite and reanalysis rainfall products across Egypt's hydrological zones

Article Title: Evaluation and bias correction of high-resolution satellite and reanalysis rainfall products across Egypt’s hydrological zones

Article References: Hesham, M., Helmi, A. M., Elkiki, M., Hamed, Y., & Selim, T. (2026). Evaluation and bias correction of high-resolution satellite and reanalysis rainfall products across Egypt’s hydrological zones. Theoretical and Applied Climatology, 157(9), Article 601. https://doi.org/10.1007/s00704-026-06519-x

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06519-x

Keywords: satellite precipitation, bias correction, Egypt, CHIRPS, GPM IMERG, ERA5, deep neural networks, quantile mapping, flash flood forecasting, arid climate, hydrological modeling, rain gauges

The research, led by Binyam Workeye Asfaw of Virginia Tech’s Department of Biological Systems Engineering with colleagues at Virginia Tech, Morehead State University, and Virginia Tech’s School of Animal Sciences, tested whether satellite soil moisture observations could serve as a calibration target in their own right. The team worked in the upper Stroubles Creek watershed, a 14.5-square-kilometer headwater catchment in Montgomery County, Virginia, monitored by the Virginia Tech StREAM Lab. The watershed sits in the Valley and Ridge physiographic province, underlain by dolomite and limestone geology riddled with springs and sinkholes. It receives roughly 1,200 millimeters of precipitation annually, about 760 millimeters of which returns to the atmosphere through evapotranspiration. Crucially, the catchment is dominated by saturation excess runoff, meaning that water only runs off the land where the soil has already filled to capacity, a process controlled largely by topography and shallow soil storage.

That hydrologic regime made the site an ideal proving ground for the Soil and Water Assessment Tool with a Variable Source Area modification, known as SWAT-VSA. Unlike conventional model setups that lump land areas into uniform hydrologic response units, SWAT-VSA redistributes soil and hydrologic properties according to classes of the topographic index, a measure of how strongly each landscape position accumulates moisture from upslope. In the Stroubles Creek application, the watershed was divided into three topographic index classes covering 56, 38, and 6 percent of the area, from driest ridge tops to wettest convergent hollows. This structure matters because in saturation excess landscapes, runoff is wildly disproportionate: the wettest class, just 6 percent of the watershed, generated roughly 30 percent of surface runoff on a per-area basis, while the largest, driest class contributed only about 9 percent. Capturing that spatial organization is essential if a model is to represent where water actually moves.

The satellite data at the heart of the study came from NASA’s Soil Moisture Active Passive mission, or SMAP, which measures surface soil moisture globally at a native resolution of 36 kilometers, far too coarse for a watershed of 14.5 square kilometers. The researchers turned to an enhanced 10-kilometer SMAP product and then downscaled it further to 500 meters using a pretrained quantile random forest model from the mlhrsm package in R. The downscaling algorithm fuses an impressive array of auxiliary data: Sentinel-1 radar backscatter, MODIS land surface temperature, Landsat-derived vegetation indices, a 10-meter digital elevation model, POLARIS soil properties including clay, sand, and bulk density, and National Land Cover Database classes. The result was a daily volumetric soil moisture estimate for every 500-meter cell in the watershed from April 2015 through August 2022, spatially averaged into a single watershed-mean time series for calibration.

There was a catch. When compared against ground measurements, the downscaled product showed a compressed dynamic range: it peaked at roughly 32 percent volumetric water content while field observations reached about 45 percent, and it systematically muted the wetting pulses that follow heavy rain. Rather than discarding the product, the team applied a deliberately parsimonious, event-based bias correction. Whenever daily effective precipitation exceeded a threshold of 5 millimeters, the depth needed to bring the top 50 millimeters of soil from field capacity to saturation, the satellite estimate was nudged upward by a multiple of the uncertainty bound reported by the downscaling model itself. Importantly, none of the correction parameters were derived from the in-situ field data, preserving the independence of the evaluation dataset. The correction amplified event-scale wetting responses while leaving dry-period behavior and long-term means untouched.

For ground truth, the researchers relied on an unusually dense field campaign: 25 measurement locations spread across a 4.2-hectare mixed-grass pasture within the watershed, sampled on 20 dates between March 2023 and January 2024, one to three days after precipitation events. The observations spanned nearly the full moisture range, from near wilting point to near saturation, between 0.15 and 0.45 cubic meters of water per cubic meter of soil. Point measurements were aggregated to a field-scale mean using topographic index weighting consistent with the model’s internal structure, a scaling choice justified by prior work showing that in this landscape, spatial moisture patterns are governed more by terrain than by local soil variation.

The team then ran three calibration experiments: one targeting streamflow only, one targeting satellite soil moisture only, and a multi-objective version optimizing both simultaneously. Single-objective runs used a differential evolution algorithm evaluating 650 parameter vectors, while the multi-objective case employed the non-dominated sorting genetic algorithm NSGA-II to map out a Pareto front of optimal trade-off solutions. The results were striking. All three models performed comparably on streamflow, with Nash-Sutcliffe efficiency values between 0.54 and 0.58 during calibration and 0.55 to 0.56 during the independent evaluation period. But the soil moisture picture diverged dramatically. The streamflow-calibrated model explained only 50 percent of observed field-scale soil moisture variability, while both the soil-moisture and multi-objective calibrations explained 88 percent. The multi-objective model cut the root mean squared error from 0.05 to 0.03 cubic meters per cubic meter and reduced bias to a negligible minus 0.8 percent.

Yet the study’s most important contribution may be its cautionary lesson about statistics versus physical realism. The soil-moisture-only calibration achieved excellent agreement with moisture observations, but it did so partly through compensatory parameter adjustments: it converged on lower available water capacity and evapotranspiration compensation parameters that dampened vertical moisture redistribution, producing systematically overestimated low flows and an altered water balance partitioning. Calibrated soil parameters deviated from database values by as much as 155 percent for available water capacity and 136 percent for saturated hydraulic conductivity, a reminder that regional soil datasets often misrepresent local conditions. The multi-objective approach, by contrast, narrowed the range of acceptable parameter values for the most sensitive soil and evapotranspiration parameters, reducing equifinality, the persistent problem in hydrology where many different parameter combinations produce equally good fits. Fewer equally acceptable solutions mean more confidence that the model is right for the right reasons.

The authors are careful to delineate the limits of their findings. The study watershed is small, humid, and strongly terrain-controlled, and the SWAT-VSA framework applies specifically to saturation excess environments; infiltration-excess or arid catchments may not benefit from the topographic index structure. The bias correction relies on watershed-average precipitation and may fail where rainfall varies sharply across space, and the relative contributions of the native SMAP retrieval, the machine learning downscaling, and the bias correction were not disentangled. Still, the implications are considerable. Satellite soil moisture products offer global coverage at essentially no marginal cost, making them a promising calibration resource for the many watersheds that lack gauges altogether. Combined with terrain-informed model structures and multi-objective optimization, they appear capable of constraining the internal states of hydrological models, not just the water leaving them, bringing field-scale predictions of soil wetness, and the agricultural and water quality decisions that depend on them, within reach.

Subject of Research: Using downscaled and bias-corrected satellite soil moisture data to calibrate watershed hydrological models

Article Title: Calibration using downscaled and bias-corrected satellite soil-moisture data can improve watershed model representation of soil-moisture variability

Article References: Asfaw, B. W., Maksud, S., Fuka, D. R., Collick, A. S., White, R. R., & Easton, Z. M. (2026). Calibration using downscaled and bias-corrected satellite soil-moisture data can improve watershed model representation of soil-moisture variability. Hydrology and Earth System Sciences, 30(18), 5999-6018. https://doi.org/10.5194/hess-30-5999-2026

Image Credits: AI Generated

DOI: 10.5194/hess-30-5999-2026

Keywords: soil moisture, SMAP, satellite remote sensing, watershed modeling, SWAT-VSA, model calibration, hydrology, downscaling, bias correction, saturation excess runoff, multi-objective optimization, streamflow

The work, published as a preprint under review at the journal Wind Energy Science by Hadi Hoghooghi and Carlo L. Bottasso of the university’s Wind Energy Institute, builds on a previously developed digital-shadow framework. A digital shadow, in this context, is a computational model of a physical turbine that runs alongside the real machine, continuously estimating quantities that cannot be directly measured, such as the bending moments acting along the blade structure. The Munich team’s innovation lies in how they correct the inevitable biases that creep into such models when reality refuses to behave exactly as the equations predict.

At the heart of the method is a linearized aeroelastic model, a computationally efficient mathematical description of how the turbine’s blades, drivetrain, and tower respond to aerodynamic forces. Linearized models are fast enough to run in real time, which is essential for a digital shadow that must keep pace with the living turbine. Their weakness, however, is accuracy. Simplifications made to keep the model tractable introduce systematic errors, or biases, that vary with the operating condition. A model that performs well in steady, uniform winds may drift badly when the rotor faces turbulent, wake-disturbed, or yawed inflow.

Hoghooghi and Bottasso’s solution is a learning-based bias correction layered on top of the physics model. A neural network is trained to learn the operating-condition-dependent discrepancies between the model’s predictions and actual measurements from the turbine. Crucially, the network does not replace the physics-based model; it augments it. The learned corrections adjust the model’s outputs while preserving its underlying structure, which means the digital shadow retains the interpretability and extrapolation behavior of an engineering model while gaining the data-fitting power of machine learning. This hybrid philosophy reflects a growing consensus in wind energy research: pure physics models are too rigid for messy atmospheric reality, while pure machine learning models are too opaque and data-hungry to be trusted with the structural safety of a multi-million-euro machine.

To validate the framework, the researchers turned to field data from a 3.5 MW eno126 turbine, gathered during two measurement campaigns in October 2020. The datasets were divided according to the complexity of the inflow conditions the turbine experienced: simple inflow, meaning relatively clean and uniform wind; complex inflow, involving turbulent or disturbed conditions such as those created by wakes from neighboring turbines; and mixed or transitional conditions that combine elements of both. After filtering, roughly 49 hours of simple-inflow data and 23 hours of complex-inflow data remained, with portions reserved for testing, along with an additional 7.5 hours of mixed and transitional data used to probe how well the method generalizes beyond its training conditions.

The headline result concerns blade bending moments evaluated at the 25 percent span location, a station partway along the blade where fatigue loads are critical to structural design. Without bias correction, the digital shadow’s damage equivalent load estimates, a standard fatigue metric that condenses a turbine’s entire loading history into a single number, carried errors of 14 to 24 percent. With the neural-network correction in place, those errors fell below 5 percent. The largest improvements appeared precisely where they were most needed: under complex inflow conditions, where the linearized model’s assumptions break down most severely and where uncorrected predictions were least trustworthy.

The team also developed a hybrid strategy that combines a baseline bias correction with the neural-network-based correction. The motivation is robustness. A neural network, however well trained, can behave unpredictably when fed conditions far outside its training distribution, and wind is nothing if not unpredictable. By blending the learned corrections with a simpler, more conservative baseline approach, the hybrid method maintains reliable performance across a wider range of operating and inflow conditions than either component could achieve alone. This kind of belt-and-braces design is likely to appeal to industrial operators, who must balance the promise of advanced analytics against the consequences of a rare but catastrophic misestimate.

The practical implications extend well beyond academic benchmarking. Accurate fatigue load estimation is the foundation of modern turbine monitoring and predictive maintenance. If an operator knows how much cyclic stress each blade has actually endured, maintenance can be scheduled based on real accumulated damage rather than conservative generic assumptions, extending component life and reducing downtime. For offshore turbines, where inspection visits are enormously expensive, the value of trustworthy digital shadows is even greater. Load estimates also feed into lifetime extension assessments, a rapidly growing concern as the first large generation of European wind turbines approaches the end of their design lives.

Yet the study has drawn pointed criticism during the open peer discussion that Copernicus journals conduct publicly. Reviewer J. Gordon Leishman acknowledged that the revised manuscript is shorter, more focused, and clearer in its comparisons between the uncorrected shadow, the baseline correction, the neural-network approaches, and the hybrid method. But he argued that the central scientific concern remains unresolved: the entire evaluation rests on a single turbine at a single site, instrumented in one way, during one autumn month. The so-called unseen test data, he noted, still come from the same machine and campaign, constituting holdout validation within one dataset rather than genuinely independent external validation. Because a learned correction may absorb turbine-specific structural properties, controller behavior, sensor characteristics, and site peculiarities, there is no evidence yet that the method transfers to another turbine, another season, or another site. Leishman also questioned the independence of the validation quantity itself, observing that root and inboard blade bending moments are structurally related, so agreement at a second spanwise station checks structural-load estimation without proving the underlying aerodynamic loading was correctly reconstructed. In a separate comment, he further raised a potential conflict-of-interest concern, noting that corresponding author Bottasso serves as Editor-in-Chief of Wind Energy Science and arguing that an independent editor should handle the manuscript.

These criticisms do not erase the significance of the result; they define its boundary. What Hoghooghi and Bottasso have demonstrated is that physics-plus-learning digital shadows can achieve remarkable accuracy on the turbine and conditions where they were trained and tested, with the gains concentrated exactly in the difficult flow regimes that matter most. The authors themselves flag wake flows, yawed conditions, offshore environments, additional turbine platforms, and strongly transient operation as future work, which is precisely the list of challenges any claim of broad applicability must eventually confront. If the method survives that wider testing, the payoff could be substantial: a generation of turbines whose structural health is known not from conservative design envelopes but from continuously corrected, physically grounded, learning-enhanced estimates of the real loads they carry through every gust. For an industry betting trillions on machines that must endure twenty-five years of punishment in the sky, that kind of visibility may prove one of the most consequential technologies of the energy transition.

Subject of Research: Hybrid physics and machine learning digital shadows for wind turbine fatigue load estimation

Article Title: Brief communication: Enhanced wind turbine fatigue load estimation using digital shadows with data-driven bias correction

Article References: Hoghooghi, H., & Bottasso, C. L. (2026). Brief communication: Enhanced wind turbine fatigue load estimation using digital shadows with data-driven bias correction. https://doi.org/10.5194/wes-2026-156

Image Credits: AI Generated

DOI: 10.5194/wes-2026-156

Keywords: wind energy, digital shadow, fatigue loads, bias correction, neural network, aeroelasticity, damage equivalent load, predictive maintenance, turbine monitoring, machine learning, blade bending moments, Wind Energy Science

The core idea behind downscaling is deceptively simple. Global climate models divide the planet into grid cells that are typically one hundred kilometres or more across, a resolution at which coastlines, mountain ranges, and individual thunderstorms are essentially invisible. Regional climate models take the large-scale output from these global models and re-simulate it over a smaller domain at much finer resolution, adding back the topographic and coastal complexity that shapes local weather. In Australasia, where the Great Dividing Range, the complex coastline of Bass Strait, and the steep rainfall gradients of the tropical north all exert powerful influences on climate, this added detail is not a luxury but a necessity for credible projections.

The CORDEX-Australasia programme has now completed successive ensembles based on the fifth and sixth phases of the Coupled Model Intercomparison Project, known respectively as CORDEX-CMIP5 and CORDEX-CMIP6. These coordinated collections of regional simulations, produced by modelling groups across Australia and beyond, allow scientists to separate the robust signals of climate change from the noise of model-to-model differences. The synthesis of this fifteen-year effort, published in PLOS Climate by Jason Evans of the University of New South Wales and a large team of collaborators from institutions across the region, documents both the scientific achievements and the formidable challenges that remain.

One of the clearest findings to emerge from the programme is that regional models demonstrate genuine added value over the global models that drive them. This is not merely a matter of sharper pictures. The high-resolution simulations represent temperature extremes, heavy precipitation events, coastal processes, and complex terrain with substantially greater fidelity, which matters because these are precisely the phenomena that cause the most damage to communities and infrastructure. When a regional model can resolve the sea breeze that moderates a coastal heatwave, or the orographic uplift that triggers a flash flood on a mountain slope, its projections of future extremes become far more useful to planners and emergency managers.

The CMIP6-based projections now available for Australasia show robust agreement on one central fact: substantial warming through the twenty-first century is essentially locked in across the region. The magnitude of that warming, however, depends overwhelmingly on which emissions pathway the world follows, making human choices the dominant source of uncertainty in temperature projections. Precipitation is a different story altogether. Rainfall responses vary considerably from one subregion to another, with some areas projected to become drier and others facing changes that are harder to pin down, reflecting the intricate interplay of atmospheric circulation, ocean temperatures, and local geography that governs water availability across the continent and its island neighbours.

Beyond the headline projections, the past decade and a half has seen significant methodological advances that have quietly revolutionised the field. Convection-permitting modelling, which runs regional models at kilometre-scale resolutions fine enough to simulate thunderstorms directly rather than approximating them with statistical formulas, has moved from research curiosity to operational tool. Benchmarking frameworks now allow systematic evaluation of model performance against observations, while improved bias-correction methods help translate raw model output into information that hydrologists, engineers, and farmers can actually use. Together these advances have sharpened both the physical realism of the simulations and the credibility of their evaluation.

The practical payoff of this scientific investment is now visible across nearly every sector of Australian and Pacific society. Regional climate projections underpin decision-making in water resource management, where reservoir operators need to know how inflows will shift in the decades ahead. Agricultural planners use the projections to assess how growing seasons and crop water requirements will change. Disaster risk reduction agencies, energy system planners, public health authorities, and national climate services all draw on the same coordinated datasets, meaning that a single scientific infrastructure quietly supports adaptation decisions worth billions of dollars across the region.

None of this would have been possible without sustained investment in shared computing infrastructure. The National Computational Infrastructure in Australia has provided the enormous processing power and data storage that regional climate modelling demands, and coordination has been strengthened further through the National Partnership for Climate Projections, which links state and federal programmes into a coherent whole. This model of collaboration, in which different levels of government and multiple research institutions contribute to a common international framework, has become a quiet success story in how complex scientific infrastructure can be built and maintained over decades.

Significant challenges nonetheless remain on the horizon. Regional climate modelling is computationally voracious, and the demand for ever-finer resolution continually pushes against available supercomputing capacity. Persistent model biases, particularly in simulating precipitation, continue to limit confidence in some projections. Designing ensembles that efficiently sample the range of possible futures without duplicating effort is an ongoing scientific puzzle, and perhaps the most difficult challenge of all is communicative rather than computational: translating sprawling multi-model datasets into decision-ready guidance that a council engineer or a hospital administrator can act upon without needing a doctorate in climate science.

The strategic priorities identified for the coming years point toward an ambitious future. Kilometre-scale convection-permitting modelling is expected to become the new standard, capturing storm-scale processes that current models still approximate. Enhanced coupling between regional models and Earth system components such as oceans, vegetation, and carbon cycles promises more complete simulations. Coordinated super-ensembles combining many models and scenarios, the responsible integration of artificial intelligence techniques to accelerate both simulation and analysis, stronger observational constraints to narrow uncertainty, and expanded support for vulnerable remote Australian islands and Pacific Island nations all feature prominently in the roadmap. For the low-lying atoll nations of the Pacific, whose very existence is threatened by rising seas and intensifying storms, the stakes of this work could hardly be higher.

The fifteen-year story of CORDEX-Australasia is ultimately a story about turning planetary-scale physics into locally meaningful knowledge. As the synthesis makes clear, delivering robust, high-resolution climate information for adaptation and resilience across Australasia will require sustained investment, continued collaboration, and relentless innovation. But the foundation has been laid, the datasets exist, the methods have matured, and the region now possesses a scientific capability that its founders could scarcely have imagined when the first coordinated downscaling experiments began. The task ahead is to ensure that this capability keeps pace with a changing climate and the growing urgency of the decisions it must inform.

Subject of Research: Regional climate downscaling and high-resolution climate projections for Australasia under the CORDEX programme

Article Title: Fifteen years of regional climate downscaling in CORDEX-Australasia

Article References: Evans, J. P., Thatcher, M., Argüeso, D., Di Luca, A., Trancoso, R., Andrys, J., Chapman, S., Di Virgilio, G., Dowdy, A., Fita, L., Gibson, P. B., Grose, M., Howard, E., Ji, F., Kala, J., Katzfey, J., Khadke, L., Mughal, O., Ng, B., … Ye, H. (2026). Fifteen years of regional climate downscaling in CORDEX-Australasia. PLOS Climate, 5(9), e0001066. https://doi.org/10.1371/journal.pclm.0001066

Image Credits: AI Generated

DOI: 10.1371/journal.pclm.0001066

Keywords: CORDEX-Australasia, regional climate modelling, climate downscaling, CMIP6, convection-permitting models, climate extremes, precipitation projections, Australasia, Pacific Islands, climate adaptation, bias correction, climate services

The study, published in npj Sustainable Agriculture, tackles a long-standing weakness in drought science. Most drought assessments rely on indices built from precipitation or temperature, which describe the climatic drivers of drought but say little about the actual water available to plant roots. Agricultural drought, by contrast, is fundamentally a soil moisture phenomenon: it occurs when water held in the root zone falls below the levels crops need to sustain growth and yield. By working directly with bias-corrected soil moisture outputs from seventeen climate models in the Coupled Model Intercomparison Project Phase 6, or CMIP6, the team shifted the analytical focus from the atmosphere to the ground itself, capturing the variable that matters most to farmers.

Methodologically, the framework is built on careful statistical ground. The researchers used the Global Land Data Assimilation System, specifically the GLDAS-Noah v2.1 dataset produced by NASA, as their reference. This product integrates observational forcing, reanalysis data and satellite assimilation to deliver a globally consistent, gap-free simulation of soil moisture at a resolution of 0.25 degrees, making it a reliable benchmark against which climate model outputs can be judged. Because satellite-only soil moisture products suffer from cloud interference, retrieval uncertainties and sensor limitations, the physically based GLDAS record offers the temporal continuity that a global drought assessment demands. The team aggregated soil layers in both the reference data and the model outputs to represent total root-zone storage, ensuring that the two datasets described comparable conditions relevant to crops.

Raw climate model output is notoriously prone to systematic bias, arising from simplified assumptions, imperfect parameterisations and structural errors in how physical processes are represented. To address this, the researchers applied a linear scaling bias correction, adjusting each model’s monthly soil moisture values by the ratio of the GLDAS reference mean to the model mean over the 2000 to 2014 baseline period. Crucially, this approach corrects the systematic mean bias while preserving the temporal structure of each simulation, so the historical trends and relative anomalies simulated by the models remain intact. When the seventeen models were evaluated across a suite of performance metrics, including bias, correlation, the index of agreement, the Kling-Gupta efficiency and root mean square error, the bias-corrected multi-model ensemble mean outperformed every individual model, and it was therefore selected for the drought projections.

With corrected soil moisture in hand, the team computed the Standardised Soil Moisture Index, or SSMI, using a non-parametric formulation based on the empirical Gringorten plotting position rather than a fitted parametric distribution. This choice matters in a warming world. Parametric distributions assume that the statistical properties of soil moisture remain stationary, yet under climate change those properties evolve continuously, and a fixed distribution fitted to historical data can distort future drought estimates. The non-parametric approach makes no such assumption, ranking each month’s soil moisture against the full empirical record and converting those probabilities into a standardised index. Negative SSMI values flag dry conditions, classified from abnormally dry down to exceptionally dry, while positive values indicate wet conditions. A Kolmogorov-Smirnov test confirmed that the empirical distribution adequately described both the reference and the corrected ensemble data, lending statistical confidence to the classification.

The projections, spanning 2015 to 2050 under two Shared Socioeconomic Pathways, reveal a clear intensification of drought characteristics toward mid-century. SSP2-4.5 represents a middle-of-the-road future in which emissions stabilise and radiative forcing reaches 4.5 watts per square metre by 2100, while SSP5-8.5 describes a high-emissions world of rapid, carbon-intensive growth pushing forcing to 8.5 watts per square metre. Under both scenarios, the number of land pixels classified in the severe to exceptional drought categories increases over time, and mean SSMI values decline steadily, with the deterioration most pronounced under SSP5-8.5. Continental analyses show strong and consistent drying signals in South America and Europe, gradual downward trajectories in Asia and North America, and pronounced interannual swings in Australia, reflecting that continent’s acute climate sensitivity and its potential for both extreme wet spells and prolonged dry ones.

One of the study’s more counterintuitive findings concerns drought frequency. Globally, the average number of drought events actually declines slightly under the high-emissions scenario, from roughly 9.1 events to 8.6 over the projection period, yet total drought duration remains essentially unchanged at around 68 months. The explanation is that individual droughts last longer under SSP5-8.5, so fewer distinct events fit into the same window. Rather than many short dry spells, the future favours fewer but more protracted episodes of soil moisture deficit, a shift that is arguably more dangerous for agriculture, since persistent multi-season deficits deplete reservoirs, exhaust soil water reserves and push crops past critical physiological thresholds in ways that brief droughts do not.

To translate these hazard projections into agricultural consequences, the team developed a Drought Exposure Index that combines normalised drought frequency, total duration and mean intensity through a geometric mean, then overlays the result on global cropland data from the Land Cover Climate Research Data Package. The index ranges from zero to one, with higher values signalling greater exposure. The results expose stark regional disparities: the highest exposure values cluster over major croplands in South Asia, southern Africa, South America, southern Europe and parts of the United States. Under SSP2-4.5 the global mean index reaches 0.71, but under SSP5-8.5 it climbs to 0.78 by 2050, a substantial rise in exposure attributable to the emissions pathway alone. Asia, which holds 34 percent of the world’s agricultural area, followed by South America with 22 percent, face the largest absolute areas of drought-affected cropland, with the affected share trending upward in every continent through 2050.

The implications reach well beyond hydrology. Soil moisture droughts of the kind projected here directly threaten food security, and the authors frame their findings against the Sustainable Development Goals on Zero Hunger, Climate Action and Life on Land. Recent history illustrates the stakes: the Millennium Drought that gripped Australia from 1997 to 2010, the simultaneous droughts of 2010 in the Amazon, southwestern China and Russia, the Texas-Mexico and East African droughts of 2011, and the United States summer droughts of 2012 and 2016 all devastated harvests, economies and livelihoods. The new projections suggest that events of this character will become longer and more widespread, particularly in the very regions that feed growing populations.

The study also charts a path forward. Climate-resilient farming practices such as precision irrigation, conservation tillage and drought-resistant crop varieties can improve soil moisture retention, while agroforestry and regenerative agriculture strengthen soil structure and water-holding capacity. On the water side, managed aquifer recharge, rainwater harvesting and smart irrigation systems can stretch supplies through dry periods, and transboundary cooperation becomes essential where rivers cross borders. The authors argue that drought risk must be embedded in national adaptation plans, supported by improved monitoring networks, early warning systems and climate-informed decision tools, and that advances in remote sensing, data assimilation and machine learning will sharpen future assessments. They are candid about the limitations that remain, including the simplicity of linear bias correction, uncertainties in land cover change and the difficulty of separating human influences from natural variability. Even so, by pinpointing where soil moisture droughts and croplands will collide hardest, the research gives policymakers a map of where every adaptation dollar will matter most.

Subject of Research: Global projections of soil moisture droughts and agricultural drought exposure under climate change scenarios

Article Title: Projecting global soil moisture droughts under climate change: characteristics, agricultural exposure, and adaptation insights

Article References: Moradian, S., Gharbia, S., Sonny, F., Torabi Haghighi, A., & Olbert, A. I. (2026). Projecting global soil moisture droughts under climate change: characteristics, agricultural exposure, and adaptation insights. npj Sustainable Agriculture, 4(1), Article 80. https://doi.org/10.1038/s44264-026-00144-x

Image Credits: AI Generated

DOI: 10.1038/s44264-026-00144-x

Keywords: soil moisture, agricultural drought, CMIP6, climate change, SSP scenarios, Standardised Soil Moisture Index, GLDAS, bias correction, drought exposure, food security, climate adaptation, water management

The research, led by Felicia Yeboah of the University of Ghana with colleagues at the International Water Management Institute and Kenya’s State Department for Irrigation, focused on the Central Highlands, a 52,064-square-kilometer region spanning eleven counties around Mount Kenya. The team evaluated five widely used satellite precipitation products, CHIRPS, PERSIANN-CCS, PERSIANN-PDIR, TAMSAT, and MSWEP, against ground observations from eight rain gauge stations between 2012 and 2021. They also tested a model-derived river discharge dataset, VegDischarge, which simulates streamflow across more than 650,000 African river segments by routing runoff from the VegET agro-hydrologic model through the mizuRoute routing framework.

The validation relied on a battery of standard hydrological statistics. Root mean square error, or RMSE, quantifies the magnitude of deviations; the Nash-Sutcliffe Efficiency, or NSE, measures predictive skill relative to simply using the observed mean; the correlation coefficient captures how well timing and variability align; and relative mean bias reveals systematic over- or underestimation. Categorical metrics, the probability of detection and the false alarm ratio, assessed how well each product identified rainy days. The verdict was sobering: no product was uniformly reliable, and every one of them overestimated rainfall at most stations, a pattern the authors attribute to orographic processes that satellite algorithms struggle to resolve in complex terrain.

Still, two products stood out. CHIRPS achieved monthly RMSE values between 34.85 and 134.26 millimeters, NSE values from minus 0.94 to 0.44, and correlations of 0.52 to 0.87, while MSWEP recorded RMSE of 24.17 to 125.66 millimeters, NSE from minus 0.91 to 0.65, and correlations of 0.45 to 0.84. The two excelled in complementary ways: MSWEP generally reduced magnitude errors, whereas CHIRPS captured temporal variability more faithfully, posting stronger correlations at five of the eight stations. TAMSAT and the PERSIANN variants fared considerably worse, with TAMSAT producing NSE values as low as minus 4.56 at Munyaka during the long rains and RMSE figures approaching 1,000 millimeters at the annual scale.

Performance also shifted with season and elevation, a finding with direct operational consequences. During the March-to-May long rains, MSWEP was the most dependable product; during the October-to-December short rains, PERSIANN-PDIR took the lead; and across full years, CHIRPS proved most consistent. At wet, high-elevation sites such as Naro Moru and Nyahururu, all products degraded sharply, while in drier, semi-arid locations like Archer’s Post and Wamba they aligned far better with gauges. The lesson, the authors stress, is that product selection must be matched to the timescale, season, and location of the intended application, because no single dataset reproduces rainfall magnitudes reliably everywhere.

The discharge side of the analysis told a similar story of promise undermined by bias. VegDischarge successfully reproduced the seasonal rhythm of river flow, with peaks in April and May and a secondary rise in November tracking the bimodal rainfall regime. But the quantitative agreement was weak, with NSE values ranging from minus 2.28 to just 0.23 and correlations between 0.46 and 0.71. The model underestimated peak flows at the high-yielding Naro Moru stations, overestimated discharge elsewhere, and systematically underrepresented low flows across nearly all sites, a shortcoming traced to the underlying VegET model’s tendency to simulate baseflow at or near zero.

To rescue these datasets, the team applied four bias correction techniques: linear scaling, quantile mapping, the delta method, and a Random Forest machine learning model. The three statistical approaches improved matters to varying degrees, but the Random Forest method was in a class of its own. Trained on the relationship between satellite estimates and gauge observations, the ensemble of decision trees captured the complex, nonlinear error structures that simple scaling and distribution-based adjustments could not. After correction, monthly precipitation NSE rose to a consistently positive 0.85 to 0.95 across all eight stations, correlations climbed to 0.93 to 0.98, and RMSE was roughly halved at most sites, falling to between 9.05 and 46.23 millimeters.

The same technique worked equally dramatic improvements on streamflow. Corrected discharge reached NSE values of 0.80 to 0.88 and correlations of 0.92 to 0.95 at five of the six gauging stations, while preserving the variance structure of the hydrograph rather than smoothing it away. One station, A5 Naro Moru, exposed a structural limitation: Random Forest predictions are bounded by the range of their training data, so the method could not extrapolate to that catchment’s extreme flood peaks, and linear scaling actually performed better there. The authors note this weakness matters most precisely for flood applications, and they point toward hybrid, physics-informed correction approaches as the next frontier.

The implications extend well beyond Kenya. Sparse observational networks plague much of Africa, where global hydrological products are often adopted without local validation, and errors in rainfall inputs propagate through every downstream model that depends on them. An annual water-balance check lent physical credibility to the corrected datasets, yielding plausible positive storage changes of 2.8 to 8.1 cubic kilometers per year between 2018 and 2021 for this recharge-dominated water tower. For a region where agriculture, biodiversity, and growing urban populations all compete for water, the bias-corrected CHIRPS and VegDischarge datasets offer something previously scarce: defensible, high-accuracy estimates on which allocation decisions and flood planning can credibly rest.

Challenges remain. With only eight rainfall stations and six discharge gauges, the study’s corrections may not generalize to underrepresented peripheral areas, and the transferability of machine learning models across regions, time periods, and hydroclimatic regimes is still unproven. The authors also call for formal uncertainty quantification and for testing the newly released CHIRPS version 3 in similar environments. Even so, the study delivers a clear and consequential message: in data-scarce mountain regions, systematic bias, not random noise, is the dominant error in global hydrological products, and machine learning has matured into an essential tool for turning satellite eyes and model simulations into water management instruments that communities can actually trust.

Subject of Research: Bias correction of satellite precipitation products and model-derived river discharge in Kenya's Central Highlands

Article Title: Toward reliable water resource assessment: Correcting bias in satellite-based rainfall and model-derived discharge in Kenya’s Central Highlands

Article References: Yeboah, F., Owusu, A., Leh, M., Akpoti, K., Mekonnen, K., Adamseged, E. M., Odera, E., & Velpuri, N. (2026). Toward reliable water resource assessment: Correcting bias in satellite-based rainfall and model-derived discharge in Kenya’s Central Highlands. PLOS Water, 5(9), e0000623. https://doi.org/10.1371/journal.pwat.0000623

Image Credits: AI Generated

DOI: 10.1371/journal.pwat.0000623

Keywords: satellite precipitation, bias correction, machine learning, Random Forest, CHIRPS, MSWEP, VegDischarge, Kenya, Central Highlands, hydrology, streamflow, water resources