The study, led by researchers analyzing nearly four decades of weather station records, satellite imagery and ocean indices, set out to answer a deceptively simple question: when a local climate shifts, how much of that shift belongs to the planet’s natural cycles, and how much belongs to us? To do this, the team contrasted two twenty-year windows, using a single weather station in central Tepic with continuous data from 1980 to 2018 and only 2.6 percent missing readings. They standardized monthly temperature and precipitation anomalies and correlated them with the multivariate El Niño index, a composite measure built from sea-level pressure, winds, sea surface temperature and cloudiness over the tropical Pacific. They then layered on a Bayesian network analysis, a probabilistic technique that estimates the chance of a climatic outcome given an ENSO phase, even when no strict cause-and-effect relationship exists.

The temperature signal was unambiguous. Average temperature climbed from 21.9 degrees Celsius in the historical period to 22.9 degrees in the recent one. Maximum temperatures rose by about 0.8 degrees, but the most dramatic change came at night: minimum temperatures jumped by roughly 1.3 degrees, and during the autumn and winter months of October through January, the increase exceeded 2 degrees. Rising nighttime minimums are a classic fingerprint of urbanization, because asphalt and concrete absorb heat all day and release it slowly after dark. They are also among the most dangerous manifestations of warming for human health, depriving bodies of the overnight recovery window during heat waves and straining the cardiovascular systems of the elderly and children.

Precipitation told a more turbulent story. The annual total barely moved, hovering near 1,130 to 1,160 millimeters across both periods, but the seasonal rhythm shifted sharply. Summer rainfall declined by between 2 and 6 percent, while autumn and winter rains surged by 13 to 20 percent. That combination, drier summers and wetter autumns, is a recipe for both water stress and flooding, and Tepic is poorly equipped for the latter. The city’s drainage infrastructure is more than fifty years old, its natural watercourses have been obstructed by construction, and buildings have spread into flood-prone zones. The researchers warn that increasingly intense autumn storms will overwhelm sewers and urban rivers, a problem they trace directly to decades of unplanned growth.

The ENSO analysis revealed something unexpected: the Pacific’s grip on Tepic’s climate is not static, and it is moving in opposite directions for temperature and rain. El Niño’s influence on temperature variation doubled, from 10 percent in the historical period to 20 percent in the recent one, with warm Niño events now reliably associated with temperature increases and cool Niña events with decreases. The correlation between the ocean index and minimum temperature flipped entirely, from an inverse relationship of minus 10 percent to a positive one of 19 percent. In winter and spring months, the probability that a Niño event would raise minimum temperatures leapt from 8 percent historically to 34 percent in the current period. Meanwhile, ENSO’s influence on precipitation weakened, falling from 17 percent to 8 percent, yet the anomalies it produces grew more extreme, with the amplitude of wet and dry swings widening in the recent record.

Those shifting probabilities matter for forecasting. During summer El Niño events, the probability of reduced rainfall doubled from 10 percent to 20 percent, and the phenomenon’s seasonal footprint contracted from July through September to June through August, heightening the risk of drought precisely when crops and reservoirs need water most. In winter, Niña events now carry a 33 percent probability of precipitation decrease, up from 24 percent, threatening the severe droughts that historically struck between March and May. On the wet side, autumn El Niño events still bring torrential rains, though with diminished probability, 14 percent versus 31 percent historically. The Bayesian analysis also showed anomaly ranges stretching from severely dry to extremely wet conditions, a sign that when ENSO does act on Tepic’s rain, it acts harder.

To measure the human hand directly, the team turned to satellites. They processed thermal infrared imagery from Landsat 5, 7 and 8 for March of 1986, 2000 and 2018, converting raw radiance into surface temperature through standard brightness-temperature transformations calibrated across the different sensors. The picture that emerged was of a city cooking itself. In 1986, surface temperatures across Tepic ranged from 20.1 to 36.9 degrees Celsius. By 2000 the maximum had climbed to 43.9 degrees, and by 2018 it reached 45.3 degrees, roughly 8 degrees hotter than the historical ceiling. The heat island itself, defined as the zone of very high surface temperatures, covered 9.2 percent of the city in 1986, shrank in relative terms to 6.8 percent in 2000 and 6.1 percent in 2018, but in absolute area it exploded: the 2018 footprint amounted to 345.5 hectares against 217.1 hectares in 2000, a 62.8 percent expansion.

The geography of that expansion tells its own story. In 2000 the heat island was concentrated in the historic center. By 2018, scorching zones had appeared in the southeast, precisely where the city’s fastest growth occurred. The urban footprint doubled between the mid-1980s and 2015, and grew another 43.6 percent between 2000 and 2018 alone, much of it at the expense of the forests surrounding the city, including the Sierra de San Juan, a state-level biosphere reserve that supplies critical ecosystem services. The region lost about 30 percent of its adjacent forest cover over the study period. Vegetation absorbs thermal radiation and releases moisture, creating a microclimate buffer against planetary-scale phenomena; strip it away, the researchers argue, and the city loses its natural shield, leaving local temperatures more exposed to swings driven by the Pacific.

That loss of buffering may explain the study’s most provocative finding: the apparent link between heat island growth and ENSO’s doubled influence on temperature. The relationship remains descriptive rather than statistically proven, and the authors are careful to say so. They relied on a single weather station, made only three satellite observations across four decades, and could not perform a formal statistical correlation of heat island area against climate variability. They call for annual monitoring, in-situ temperature measurements of different urban surfaces, an inventory of green spaces, and the inclusion of future climate scenarios. They also suggest examining other ocean-atmosphere drivers, from the Pacific Decadal Oscillation to the North American monsoon and hurricanes, that could be shaping Tepic’s climate alongside ENSO.

Even with those caveats, the implications reach far beyond one mid-sized Mexican city. Tepic’s population of nearly half a million in its metropolitan area, hosting over 40 percent of the state of Nayarit’s people, makes it a template for hundreds of cities across the developing world where urban sprawl is outpacing planning. The consequences documented here, higher electricity demand and greenhouse emissions, water stress, intensified heat waves, flooding from aging drainage, and risks to public health ranging from viral disease to psychological strain, are the standard costs of unplanned growth. The researchers’ prescription is blunt: expand meteorological monitoring networks, invest in climate action research, and treat urban planning as an essential instrument for building climate-resilient cities. As global temperatures rise and the Pacific keeps oscillating, the cities that protect their forests and green space, the study suggests, will be the ones best armored against both.

Subject of Research: The influence of ENSO and the urban heat island on temperature and precipitation variation in Tepic, western Mexico

Article Title: Influence of ENSO and the urban heat island on climate variation in a growing city of the western Mexico

Article References: Nájera González, O., Nájera González, A., Marceleño Flores, S., & Flores Vilchez, F. (2024). Influence of ENSO and the urban heat island on climate variation in a growing city of the western Mexico. One Ecosystem, 9, Article e125302. https://doi.org/10.3897/oneeco.9.e125302

Image Credits: AI Generated

DOI: 10.3897/oneeco.9.e125302

Keywords: ENSO, El Niño, urban heat island, climate change, Tepic, Mexico, urbanization, deforestation, Landsat, precipitation, Bayesian networks, urban planning

The study’s methodological core was a head-to-head comparison of Random Forest, Support Vector Machine, Gradient Tree Boosting, and K-Nearest Neighbors classifiers, all trained on identical input variables to ensure a fair contest. The team used six spectrally equivalent Landsat bands—Blue, Green, Red, NIR, SWIR1, and SWIR2—together with five spectral indices: NDVI for vegetation density, SAVI to minimize soil brightness effects in sparse vegetation, EVI to reduce atmospheric influence, NDBI to highlight built-up surfaces, and NDWI to detect surface water. All imagery was pre-processed using Level-2 Surface Reflectance products, with the LEDAPS atmospheric correction applied to Landsat 5 Thematic Mapper data and the LaSRC correction applied to Landsat 8 and Landsat 9 datasets, ensuring radiometric consistency across the twenty-year time series.

Random Forest emerged as the clear winner. It achieved the highest Overall Accuracy across all three study years—92 percent in 2005, 89 percent in 2015, and 89 percent in 2025—with Kappa coefficients of 0.87 throughout. Gradient Tree Boosting followed closely with accuracies between 86 and 88 percent, while K-Nearest Neighbors produced satisfactory results in the 84 to 86 percent range. Support Vector Machine lagged behind at roughly 72 to 73 percent, with Kappa values as low as 0.55 in 2005, indicating substantial classification confusion for spectrally similar vegetation classes. The researchers attribute Random Forest’s dominance to its ensemble architecture, which builds many decision trees on bootstrap samples and random feature subsets, then determines the final class by majority voting—a design that handles high-dimensional data, nonlinear relationships, and noisy inputs while resisting overfitting.

Class-level accuracy assessment reinforced this picture. Under Random Forest, water bodies and plantation forests recorded User’s Accuracy values of 0.95 and 0.96 in 2005, with Producer’s Accuracy reaching 0.92 and 0.97 respectively. By 2015 and 2025, water classification had become near-perfect across the ensemble methods, with accuracy values approaching 1.00. In contrast, Support Vector Machine struggled badly with grassland in 2005, recording a Producer’s Accuracy of just 0.19, and confused woodland/shrubland and bare land classes throughout the study period. The confusion matrices confirmed that Random Forest produced the most balanced omission and commission errors across nearly all land cover classes, a critical property when the resulting maps feed directly into change detection analysis.

The spatial story that emerged from the highest-accuracy maps is striking. Woodland and shrubland remained the dominant land cover class—Random Forest estimated it at 31,129 square kilometers in 2005—but it experienced progressive fragmentation and conversion to grassland, agricultural land, and bare land. Built-up areas expanded substantially, particularly between 2005 and 2015, when Random Forest estimated an increase of 1,371.5 square kilometers, or 184 percent. Agricultural land also grew, reflecting intensifying farming activity and land conversion on communal lands surrounding the park. Notably, relatively stable vegetation patterns were observed within the protected core of Kruger National Park itself, suggesting that formal conservation management is effectively buffering the park’s interior even as surrounding landscapes transform.

Perhaps the most alarming trend is hydrological. Water bodies declined consistently across all classifiers and all time intervals. Random Forest estimated a loss of 1,052.9 square kilometers—76 percent—between 2005 and 2015, and the long-term 2005 to 2025 analysis showed a 76 percent decline overall. K-Nearest Neighbors estimated an even steeper 83 percent reduction. Because water availability regulates vegetation productivity, wildlife distribution, and ecosystem resilience in semi-arid savannas, this persistent contraction of rivers, dams, and seasonal water points signals increasing hydrological and environmental stress, likely driven by the combined influence of rainfall variability, rising temperatures, and growing human demand for water resources.

The transition matrices built from the Random Forest maps revealed a savanna in constant flux. Between 2005 and 2015, approximately 17,829 square kilometers of woodland and shrubland persisted unchanged, but 5,442 square kilometers converted to grassland, 2,346 square kilometers to natural forest, and 1,942 square kilometers to bare land. Grassland, in turn, showed 4,815 square kilometers converting back into woodland and shrubland. These substantial bidirectional exchanges between woody vegetation and grassland suggest that the ecosystem remains highly dynamic, with vegetation succession, regeneration, fire disturbance, and herbivory by elephants and other large mammals operating simultaneously alongside anthropogenic land conversion. Some transitions likely represent natural ecological processes rather than permanent degradation alone.

The researchers interpret these patterns as the product of interacting drivers rather than any single force. Climate variability—with annual rainfall ranging from roughly 450 to 750 millimeters and summer temperatures frequently exceeding 30 degrees Celsius—strongly regulates vegetation productivity and surface water availability. Fire regimes and herbivory continuously restructure vegetation communities within the park. Beyond the park boundaries, population growth, agricultural expansion, settlement development, and infrastructure construction drive landscape transformation, particularly along park edges and on communal lands. The transboundary nature of the study area adds further complexity, as differences in land-use policies, governance systems, and socio-economic conditions across South Africa, Mozambique, and Zimbabwe shape the intensity and spatial distribution of change.

Methodologically, the study demonstrates that integrating machine learning with multi-temporal Landsat imagery processed in Google Earth Engine provides a robust framework for monitoring complex, heterogeneous savanna landscapes. The team collected 1,163 reference samples distributed across eight land cover classes, splitting them 70 percent for training and 30 percent for independent validation, with identical datasets used across all four algorithms. Training samples were digitized as polygons on high-resolution imagery and converted to pixel-based points using stratified sampling, minimizing labeling errors and spatial bias. The authors acknowledge that the 30-meter resolution of Landsat inevitably produces mixed pixels along woodland-grassland transitions, riparian corridors, and fragmented peri-urban environments, contributing to some classification uncertainty regardless of algorithm.

The implications extend well beyond Kruger National Park. Accurate, continuous land cover monitoring can support early detection of vegetation degradation, agricultural encroachment, urban expansion, and hydrological change, enabling timely conservation interventions. The transition matrices offer conservation agencies a practical tool for identifying priority areas for ecological restoration and evaluating management effectiveness. In transboundary conservation landscapes like the Great Limpopo Transfrontier Conservation Area, such evidence can strengthen collaborative planning among the three nations sharing the ecosystem. The researchers suggest that future work should integrate higher-resolution optical, radar, LiDAR, and UAV datasets with advanced deep learning approaches, including Convolutional Neural Networks and Vision Transformers, to resolve fine-scale landscape features that medium-resolution satellite imagery cannot capture. For now, the study stands as compelling scientific evidence that ensemble-based machine learning, particularly Random Forest, offers the most reliable approach for mapping and monitoring one of Africa’s most ecologically significant protected landscapes as human pressures continue to reshape it.

Subject of Research: Machine Learning Reveals Two Decades of Landscape Change Around Kruger National Park

Article Title: Multi-temporal land use/land cover dynamics (2005–2025) and machine learning algorithm comparison in and around kruger national park

Article References: Gwala, O. Z., Altoom, M. B., Rameolo, A., & Gebreslasie, M. (2026). Multi-temporal land use/land cover dynamics (2005–2025) and machine learning algorithm comparison in and around kruger national park. Frontiers in Remote Sensing, 7, Article 1915380. https://doi.org/10.3389/frsen.2026.1915380

Image Credits: AI Generated

DOI: 10.3389/frsen.2026.1915380

Keywords: remote sensing, machine learning, Random Forest, Kruger National Park, land use land cover, change detection, savanna ecosystems, Landsat, conservation, Google Earth Engine, Multi-temporal, land

The research team, led by Mohamed L. Traore with Opeyemi S. Aniramu and Oluwagbenga O.I. Orimoogunje, focused on an area of roughly 33 square kilometers within the Upper Guinea savannah zone, lying between latitudes 11°20′ and 11°50′ North and longitudes 9°10′ and 10°10′ West. Guinea ranks among Africa’s leading gold producers, and Siguiri hosts some of the country’s most extensive mining operations, including the industrial ventures Société Anglogold Ashanti de Guinée (SAG) and Société Minière de Dinguiraye (SMD). Since industrial mining began in 1996, the landscape has undergone substantial environmental change, making Kintinian an ideal natural laboratory for studying mining-induced degradation.

What sets this study apart methodologically is its integrated analytical framework. The researchers adopted the Geographical Pathways to Environmental Sustainability (GPES) framework, which links human-induced pressures to environmental outcomes across spatial and temporal scales. Rather than relying on a single satellite snapshot, they combined multi-temporal Landsat imagery from 1990, 2000, 2010, and 2016 with Sentinel-1 radar data, high-resolution Google Earth Pro imagery, and extensive field validation. The team deliberately chose the Soil-Adjusted Vegetation Index (SAVI) over the more commonly used Normalized Difference Vegetation Index (NDVI) because SAVI incorporates a soil adjustment factor that corrects for the bright, exposed ground surfaces characteristic of mining landscapes—conditions under which NDVI’s readings become unreliable.

The land-cover classification itself was performed using a supervised Random Forest algorithm, a machine-learning approach that builds hundreds of decision trees from bootstrap samples and assigns classes by majority vote. The researchers trained the model on 2,000 sample pixels distributed equally across five land-cover categories—water body, forest, agricultural land, built-up and bare soil, and rock—and validated it against 500 independent samples. Accuracy was impressive: overall classification accuracy rose from 86.4 percent in 1990 to 92.7 percent in 2016, with kappa coefficients climbing from 0.82 to 0.89, indicating strong agreement between satellite-derived maps and ground truth.

The temporal analysis revealed dramatic shifts. In 1990, six years before industrial mining commenced, forest and agricultural land together covered more than 60 percent of the study area—539.24 and 537 square kilometers respectively. By 2000, just four years after mining began, built-up and bare soil had nearly doubled from 271.48 to 569.46 square kilometers, a gain of roughly 298 square kilometers in a single decade. Agricultural land, meanwhile, lost about 80 square kilometers during that same period as people abandoned farming for mining. The trend intensified between 2000 and 2010, when farmland plummeted by 346 square kilometers—a striking illustration of what the authors describe as the socio-economic

The choice of SAVI deserves closer attention, because it reflects a broader methodological debate in remote sensing science. NDVI, which exploits the contrast between red-light absorption by chlorophyll and near-infrared reflectance from leaf structure, has been the workhorse vegetation index since the earliest satellite missions. Yet in landscapes where vegetation is sparse and soil brightness varies widely, NDVI tends to conflate genuine plant vigor with the reflectance of the ground beneath. SAVI addresses this by introducing a soil-brightness correction factor into its formulation, damping the influence of exposed mineral surfaces. In a mining landscape scarred by excavated pits, tailings, and freshly turned laterite, that correction is not a technical nicety but a prerequisite for meaningful measurement. The authors’ decision to apply SAVI in a West African gold-mining context also fills a gap: the index has been validated extensively in precision agriculture and viticulture, but its use in assessing the environmental consequences of artisanal and industrial mining in Guinea had remained largely unexplored.

The study period itself was chosen with care. The authors restricted their multi-temporal comparison to 1990, 2000, 2010, and 2016 because persistent cloud cover after 2016 limited the availability of comparable, cloud-free satellite imagery over this part of Guinea. This constraint is a familiar one for remote sensing in the humid tropics, where seasonal cloud and haze can render optical sensors blind for months at a time. By anchoring the analysis to dates with reliable imagery, the researchers ensured that observed changes in land cover and vegetation condition reflect real landscape transformation rather than atmospheric artifacts. The 1990 to 2016 window also captures the full arc of industrial expansion, beginning six years before the establishment of industrial mining in 1996 and extending through two decades of intensive operation.

Field validation formed the third pillar of the methodology. Ground-truthing excursions allowed the team to confirm that pixels classified as bare soil, forest, or farmland corresponded to conditions on the ground, a step that becomes especially important in heterogeneous mining landscapes where spectral signatures can be ambiguous. This combination of satellite observation, radar data, and physical verification reflects a growing consensus in environmental monitoring science that remote sensing is most powerful when paired with local knowledge and direct observation.

The analytical framework also incorporated multi-criteria decision analysis weighted through the Analytic Hierarchy Process, or MCDA-AHP, a structured technique for translating expert judgment into explicit, comparable weights across criteria. Applied here within the GPES lens, the approach allowed the researchers to integrate spectral, thermal, land-cover, and soil-related indicators into a single spatial characterization of environmental vulnerability. Rather than treating vegetation loss, soil exposure, and land transformation as isolated symptoms, the framework maps where these pressures converge, identifying zones of differing vulnerability across the study area. Such spatially explicit vulnerability mapping is a practical tool for land managers, because it directs attention to the places where remediation and restoration efforts are likely to yield the greatest benefit.

The broader context of the findings extends well beyond Kintinian. Gold mining across West Africa has expanded rapidly in recent decades, driven by rising global demand, favorable prices, and the accessibility of alluvial deposits to artisanal miners. In Guinea, gold is extracted in nearly every village of the Kintinian sub-prefecture, and the local population has woven mining into a livelihood portfolio that also includes agriculture, livestock rearing, fishing, trade, and handicrafts. The study notes that these alternative activities are themselves affected by mining, creating a feedback loop in which environmental degradation undermines the very diversification that communities rely upon for resilience. When farmland is converted to mining pits, households lose agricultural capacity; when vegetation cover declines, soil stability and microclimate regulation weaken, further diminishing the productivity of remaining agricultural land.

This dynamic illustrates a central tension in sustainability science: the same extractive activity that generates employment, revenue, and foreign exchange can simultaneously erode the ecological foundations of local livelihoods. The authors frame their work within the three pillars of sustainability—social, economic, and environmental—arguing that understanding their interaction is essential for designing interventions that do not simply restrict mining but redirect it toward more sustainable practices. Their findings offer evidence-based guidance for land restoration strategies and remediation planning in mining-affected landscapes of the Global South, where regulatory capacity may be limited but the ecological stakes are high.

The study also contributes to a methodological frontier. Previous assessments of mining impacts have often been localized, based on single-date observations, or focused on individual environmental components in isolation. By combining land-cover change detection, SAVI-based vegetation assessment, soil-exposure indicators, and spatial vulnerability modelling within one coherent framework, the researchers demonstrate how multiple lines of geospatial evidence can be synthesized into an integrated picture of environmental change. This integration matters because mining-induced degradation operates through interacting pathways: deforestation exposes soil, exposed soil accelerates erosion, erosion degrades remaining vegetation, and the resulting landscape fragmentation reduces ecosystem resilience. Capturing these interactions requires the kind of multi-indicator, multi-temporal approach the study exemplifies.

For the communities of Kintinian, the practical implications are tangible. The quantified baseline of degradation between 1990 and 2016 provides a benchmark against which future change can be measured, enabling evidence-based environmental management rather than reactive crisis response. The identification of spatially distinct vulnerability zones offers a starting point for prioritizing restoration, whether through revegetation of abandoned mining sites, stabilization of exposed soils, or the protection of remaining forest fragments. And the demonstration that SAVI can reliably track vegetation condition in a soil-dominated mining landscape opens the door for routine, low-cost monitoring across other gold-producing regions of West Africa, where similar dynamics of expansion, degradation, and partial recovery are unfolding. As the authors acknowledge, their work establishes a foundation upon which future investigations of post-mining sustainability can build, turning a quarter-century of satellite observation into a roadmap for ecological recovery.

Subject of Research: Remote sensing assessment of gold mining impacts on vegetation and soil degradation in Kintinian, Guinea

Article Title: Assessing the impacts of gold mining on vegetation and soil degradation using multi-temporal remote sensing in Kintinian, Siguiri prefecture, Guinea

Article References: Traore, M. L., Aniramu, O. S., & Orimoogunje, O. O. I. (2026). Assessing the impacts of gold mining on vegetation and soil degradation using multi-temporal remote sensing in Kintinian, Siguiri prefecture, Guinea. Environmental and Sustainability Indicators, 32, Article 101570. https://doi.org/10.1016/j.indic.2026.101570

Image Credits: AI Generated

DOI: 10.1016/j.indic.2026.101570

Keywords: gold mining, remote sensing, SAVI, land degradation, Guinea, Siguiri, soil vulnerability, Landsat, vegetation index, GPES framework, MCDA-AHP, environmental sustainability

The research team, led by botanists working with the National Herbarium of Uzbekistan, spent three consecutive years documenting the flora of the Yozyovon sands between 2023 and 2025. Their surveys, combined with historical herbarium specimens from collections in Tashkent, Moscow, and St. Petersburg, produced a comprehensive inventory of 296 vascular plant taxa belonging to 175 genera and 44 families. What makes this number remarkable is not just its size but its composition: nearly 45 percent of the recorded species are psammophytes, plants so specialized for life on sand that they cannot survive anywhere else.

Among these sand specialists are species found nowhere else on Earth. The team identified 14 trigger taxa that qualify the site for protection under the Important Plant Area, or IPA, framework, an international standard for identifying sites of exceptional botanical significance. One of them, Calligonum elegans, is listed as Endangered on the global IUCN Red List and in Uzbekistan’s national Red Data Book. Its rediscovery in the Yozyovon sands extends its confirmed range beyond what was documented in a 2007 assessment, which had restricted the species to the Sokh River basin. Three voucher-backed localities, including one inside the proposed IPA boundary, now confirm that this endangered shrub persists in the lowland sands of Central Fergana.

Other trigger species include Astragalus rubellus, another nationally Endangered endemic known from just five historical specimens collected between 1928 and 1952, and three Fergana Valley endemics: Asparagus ferganensis, Haplophyllum ferganicum, and Limonium ferganense. Nine additional taxa qualify under a regional subcriterion that captures species whose conservation importance is tied to habitat specialization rather than formal global threat listings. Many of these plants have broad geographic ranges but tiny, fragmented areas of occupancy, a pattern the researchers describe as especially pronounced in arid landscapes where suitable sandy habitats exist only as isolated remnants.

To quantify how fast the sands are vanishing, the team turned to remote sensing. They applied Random Forest classification, a machine-learning algorithm, to Landsat satellite imagery from four reference years: 1991, 2000, 2015, and 2025. The results are stark. Within a fixed study area of 416.24 square kilometers, mapped sand declined from 125.95 square kilometers in 1991 to just 54.79 square kilometers in 2025, a reduction of 56.50 percent. The largest single loss occurred between 2000 and 2015, when sand extent dropped by nearly 38 square kilometers at an annualized rate of 2.53 square kilometers per year. Classification accuracy was high across all epochs, with overall accuracy ranging from 0.925 to 0.983.

The drivers of this collapse are not mysterious. An adapted threat assessment aligned with IUCN standards identified 18 direct pressures affecting the site, with irrigated agriculture, hydrological modification, and substrate disturbance ranking as the dominant forces. Shifting agriculture, agro-industrial farming, and other ecosystem modifications all scored as High impact, reflecting their broad spatial scope, rapid degradation rates, and ongoing activity. Sand extraction, road construction, and land leveling further physically alter the substrate, while water abstraction for irrigation destabilizes dune systems and accelerates salinization.

What makes the Yozyovon sands particularly significant is their biogeographical position. The Fergana Valley, wedged between the Tian Shan and Pamir-Alay mountain systems, is one of Central Asia’s most distinctive and densely populated landscapes. Three IPAs have already been identified there, but all are associated with foothill or mountain environments. The Yozyovon sands represent something fundamentally different: the last extensive remnant of lowland psammophytic ecosystem in Central Fergana, shaped by aeolian processes, groundwater dynamics, and substrate specialization rather than altitudinal gradients.

The floristic composition reflects this uniqueness. The life-form spectrum is dominated by therophytes, annual plants that complete their life cycles during brief periods of favorable moisture, comprising 58 percent of the flora. This pattern reflects adaptation to seasonal aridity and unstable substrates. Meanwhile, the woody component, including species of Populus, Salix, Tamarix, and Haloxylon, contributes to dune stabilization and habitat structure, though field observations confirm increasing fragmentation from grazing and fuelwood collection. The site also functions as a relict lowland analog of desert-riverine landscapes that once occupied larger portions of the Fergana Depression following the retreat of the Syr Darya.

The proposed IPA covers 54.79 square kilometers of remaining mapped sandy habitat and would become the fourth IPA in the Fergana Valley if formally recognized. The Yazyavan Natural Monument, a small existing protected area, sits within the broader proposed boundary, but the site currently lacks formal protected status at the landscape scale. The researchers emphasize that IPA designation does not impose immediate legal restrictions, making it a pragmatic tool for reconciling conservation with the socioeconomic realities of one of Central Asia’s most intensively cultivated regions.

Beyond its regional significance, the study offers a methodological template for conservation planning in data-poor arid environments worldwide. By integrating floristic surveys, herbarium verification, standardized threat assessment, and validated satellite-based land-cover analysis, the researchers demonstrate how evidence-based prioritization can work even where botanical knowledge is uneven and ecosystems are transforming rapidly. As global commitments under the Kunming-Montreal Biodiversity Framework push nations toward conserving 30 percent of terrestrial areas by 2030, the fate of overlooked lowland sand ecosystems like the Yozyovon may well determine whether that target means anything for the world’s most specialized and vulnerable flora.

Subject of Research: Floristic assessment and habitat loss analysis of the Yozyovon Sand Massif as a proposed Important Plant Area in the Fergana Valley, Uzbekistan

Article Title: The Yozyovon Sand Massif as a proposed Important Plant Area in the Fergana Valley: floristic value, habitat loss, and threats

Article References: Tojibaev, K. S., Hoshimov, H. R., Sharipov, I. K., Akbarov, F. I., Abdullaev, S. S., Batoshov, A. R., Karimov, F. I., Dekhkonov, D. B., Kor, L., Baasanmunkh, S., Gil, H.-Y., Oh, A., & Choi, H. J. (2026). The Yozyovon Sand Massif as a proposed Important Plant Area in the Fergana Valley: floristic value, habitat loss, and threats. Nature Conservation, 64, 229-281. https://doi.org/10.3897/natureconservation.64.200242

Image Credits: AI Generated

DOI: 10.3897/natureconservation.64.200242

Keywords: Important Plant Area, Fergana Valley, Uzbekistan, psammophytic flora, sand desert ecosystems, habitat loss, Random Forest classification, Landsat, IUCN Red List, Calligonum elegans, endemic plants, Central Asia

The research, led by Ibrahim Inuwa Yahaya of Central South University in China and colleagues from institutions across China and Nigeria, focused on Wudil, a local government area whose economy and food security depend heavily on rain-fed agriculture. The team’s core analysis spanned 1986 to 2020, a period long enough to capture both natural climate variability and sustained anthropogenic change. Rather than relying on a single vegetation metric, the researchers deployed a battery of remote-sensing and drought indices, each probing a different dimension of ecosystem health.

At the heart of the analysis sits the Normalized Difference Vegetation Index, or NDVI, the workhorse of satellite vegetation monitoring. NDVI exploits the fact that healthy chlorophyll-rich leaves absorb red light strongly while reflecting near-infrared radiation, so the ratio between these two bands serves as a proxy for green biomass. But greenness alone can be misleading: a landscape can appear green while its vegetation is physiologically stressed. That is why the team paired NDVI with the Normalized Difference Moisture Index and the Normalized Difference Water Index, which trace vegetation and surface water content through complementary spectral signatures, and with the Soil Moisture Index, which estimates water availability in the ground itself.

To capture drought dynamics, the researchers turned to the Vegetation Condition Index, which compares current NDVI against its historical range to flag anomalous stress, and the Standardized Precipitation Evapotranspiration Index, or SPEI. Unlike simpler rainfall-based drought measures, SPEI accounts for atmospheric demand for water by incorporating the balance between precipitation and potential evapotranspiration. In a warming climate, this distinction matters enormously: hotter air pulls more moisture from soils and plants, so even unchanged rainfall can translate into intensifying drought. The indices were analyzed alongside ground-based climatic variables, including temperature, humidity, and precipitation records.

The long-term picture that emerges is unambiguous. Across the 1986 to 2020 window, Wudil experienced a sustained decline in vegetation health and moisture status, alongside increasing drought severity and reduced ecological resilience. The study attributes these trends to a 1.8 degrees Celsius rise in temperature over the study period and an approximately 15 percent decline in annual rainfall. In a semi-arid environment where crops and natural vegetation already operate close to their water limits, a shift of this magnitude fundamentally alters the water balance, pushing ecosystems and farming systems toward thresholds from which recovery becomes progressively harder.

Crucially, the researchers did not treat climate as the sole culprit. The historical record is overlaid with continuing human pressures: urban expansion consuming open land, deforestation stripping woody cover, and agricultural practices that fail to conserve soil and water. The interaction between these forces is what makes the Wudil case scientifically interesting and practically alarming. Climate stress weakens vegetation, making it more vulnerable to clearing and overuse; land-use change, in turn, removes the canopy and root systems that buffer the landscape against drought, creating a feedback loop of degradation.

To examine the most recent phase of this dynamic, the team applied Decision Tree Classification to Landsat 8 Operational Land Imager data for three snapshot years: 2020, 2022, and 2024. Decision trees are supervised machine-learning classifiers that partition the spectral feature space through a series of binary decisions, making them transparent and interpretable compared with more opaque algorithms. This approach allowed the researchers to map five land classes with particular relevance to local livelihoods and policy: street vegetation, forest, built-up land, irrigation land, and fallow land.

The post-2020 assessment revealed a landscape in flux. Street vegetation and irrigated land fluctuated between the snapshot years, suggesting that these categories respond rapidly to short-term economic and climatic pressures, while built-up land remained relatively stable, indicating that the pace of urban consolidation had not dramatically accelerated in the four-year window. The most striking result, however, concerned forest cover: between 2020 and 2024, Wudil lost 14.7 percent of its forest. For a semi-arid region where remaining woody vegetation provides shade, soil stabilization, fuelwood, and habitat, a loss of this magnitude in just four years represents a serious acceleration of the longer-term decline.

The study’s multi-index design carries broader methodological significance. Single-indicator assessments of vegetation change can conflate climatic and anthropogenic signals, because both drivers leave fingerprints on greenness and moisture. By combining vegetation, moisture, water, soil, and drought indices with an independent land-use classification, the researchers built a more robust attribution framework: climate indices track the environmental stress, while the land-use maps track the human restructuring of the surface. The convergence of evidence from both streams strengthens the conclusion that Wudil’s degradation is a compound phenomenon rather than the product of either driver alone.

The authors argue that their findings point to an urgent need for integrated land management, reforestation, climate adaptation, and sustainable urban and agricultural planning in Wudil and comparable semi-arid environments. In practical terms, that means pairing tree-planting and forest protection with water-efficient irrigation, land-use zoning that constrains unplanned expansion, and farming practices designed to conserve scarce moisture. As climate change intensifies across the Sahelian belt and West Africa’s drylands, the Wudil study offers both a warning and a template: satellite-based monitoring that separates climatic stress from human pressure can identify where intervention will matter most, before the window for ecological recovery closes.

Subject of Research: Climatic and anthropogenic drivers of vegetation dynamics in semi-arid Wudil, Nigeria, assessed through multi-index remote sensing and decision tree land-use classification

Article Title: Unraveling the complex nexus of climatic and anthropogenic drivers on vegetation dynamics in Wudil: a multi-index analysis (1986–2020) with decision tree-based land-use assessment (2020–2024)

Article References: Yahaya, I. I., Wang, C., Yahaya, M. S., Ogbue, C. P., Al-Masnay, Y. A., Basiru, H. A., Auwal, A., & Muhammad, A. M. (2026). Unraveling the complex nexus of climatic and anthropogenic drivers on vegetation dynamics in Wudil: a multi-index analysis (1986–2020) with decision tree-based land-use assessment (2020–2024). Theoretical and Applied Climatology, 157(11), Article 723. https://doi.org/10.1007/s00704-026-06654-5

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06654-5

Keywords: vegetation dynamics, remote sensing, NDVI, SPEI, drought, land-use change, deforestation, semi-arid, Nigeria, Landsat, decision tree classification, climate change

The scale of the shift is striking when measured against the historical record. Between 1909 and 2019, just nine detailed maps of the delta were produced, roughly one every thirteen years, and no cartographic product has been published since 2019. The gap between the 1983 map of the delta and the 2019 topographic sheets spans thirty-six years, a period during which countless natural changes and human interventions reshaped the landscape. Field observations, however meticulous, offered only punctual and fragmented glimpses of a system in constant motion.

Satellite observation began to change that calculus with the launch of the Landsat program in 1972, but early imagery was sporadic and coarse, with spatial resolutions of thirty to eighty meters. Between 1972 and 2015, researchers identified 917 available Landsat scenes covering the delta, and 675 of those were acquired before 2008, when the full archive finally opened to the public. After that year, the parallel operation of Landsat 5, 7, and 8 pushed acquisition rates to fifty-one scenes per year by 2014 and 2015. The real breakthrough, however, came with the European Union’s Copernicus program and the Sentinel-2 missions, launched in 2015 and augmented by satellites launched in 2017 and 2024.

Sentinel-2 delivers free optical imagery at spatial resolutions down to ten meters with a revisit time of five days, and the current three-satellite constellation provides coverage of both land and water surfaces that the study’s authors describe as unparalleled. The accessibility, consistency, and quality of this data now allow near-continuous monitoring of seasonal changes in water extent, vegetation health, vegetation fires, water quality parameters, and shifts in the waterline and submerged sandbars along the delta’s maritime margin. Platforms such as the Copernicus Data Space Ecosystem have democratized access, enabling specialists, local institutions, NGOs, and even the general public to visualize and interpret satellite information quickly.

To demonstrate what this new observational capacity can achieve, the researchers turned to a dramatic natural experiment: the vegetation fires that swept through the delta between January and March 2025. From mid-February into the first days of March, the area between Letea and the Sfântu Gheorghe branch showed intense burning activity. Two Sentinel-2 acquisitions, on February 20 and February 25, 2025, captured the fire’s evolution in remarkable detail. In natural color composites built from the satellite’s red, green, and blue bands, smoke plumes are clearly visible on February 25, while the February 20 image shows only incipient signs of burning.

The shortwave infrared composites, combining bands 12, 8A, and 4, tell an even sharper story. Pixels corresponding to active burning glow in intense red and orange tones, burned scars appear brown, and unaffected vegetation remains dark green. The compact shape of the fire, its south-eastern orientation, and its visible expansion in just five days point to rapid burning driven by wind. A Landsat 9 image from February 20, 2025, captured an active event on the same day, and thermal detections from NASA’s FIRMS service, which integrates MODIS and VIIRS sensors, confirmed the fire’s position and orientation with orange detection points that correlate with the Sentinel-2 observations.

The multi-sensor picture extends well beyond single snapshots. Sentinel-3 data from the World Fire Atlas, available since the winter of 2016–2017, indicated several burning events in the targeted area with a peak of activity in the second half of February. Although spatially coarser than Sentinel-2, these daily detections support rapid regional assessment. Seasonal analysis of the full record shows that most fires occur in winter and spring, with an annual minimum in summer. The number of detected events shows a positive trend overall, though 2021, 2023, and 2024 saw minimal activity. In 2025, the atlas recorded the highest number of observations since 2016, with more than forty events.

Products from the European Forest Fire Information System, which automatically generates burnt area delimitations from MODIS and Sentinel-2 observations with weekly updates and a minimum mapped area of thirty hectares, revealed an expansion of affected areas exceeding 10,000 hectares across the delta in the first three months of 2025. The researchers caution that the cumulative detection method can make events appear extensive in both space and time, but most individual occurrences were limited, lasting no more than five days and covering areas under 1,000 hectares. Temporal analysis of FIRMS detections showed a significant surge between February 21 and 28, 2025, with more than 800 detection points recorded within a single ten-day interval, supported by high values of the Fire Radiative Power indicator, which measures the rate of radiative energy emitted by a fire at the moment of observation.

Perhaps the most consequential finding concerns data availability itself. An analysis of Sentinel-2 metadata for the four tiles that fully cover the delta, spanning all three active platforms from 2015 to 2025, revealed an average of more than 1,400 scenes per tile, with the tile 35TPK recording the highest count at 1,964 images. After filtering for scenes with less than ten percent cloud cover, between 400 and 700 clear images remain per tile. That frequency allows, on average, very precise observations approximately every ten days for the least favorable tile, a density the authors describe as remarkable for monitoring dynamic processes in the delta. Cloud cover remains one of the most limiting factors despite the theoretical revisit time, which underscores the added value of high-temporal-resolution sensors such as Sentinel-3, MODIS, and VIIRS, and of derived products like EFFIS, in bypassing those constraints.

The study’s authors processed all data in open-source environments such as QGIS and SAGA GIS, and the methodology is freely replicable in other wetland regions and protected areas in Romania and worldwide. Historical maps, available through open data initiatives such as geo-spatial.org, still hold value for understanding the delta from a period when human interventions were limited and targeted, and comparing them with current satellite imagery offers perspectives on channel dynamics and morphological change over the last century. But the conclusion is unambiguous: without Earth Observation data, monitoring the changes and dynamics of the Danube Delta in real time with modern research means is simply not possible. What was once a landscape mapped once every thirteen years can now be observed nearly every week, and the fires of early 2025 stand as a vivid demonstration of what that new observational power makes visible.

Subject of Research: Earth Observation satellite monitoring of vegetation fires and environmental change in the Danube Delta, Romania

Article Title: A paradigm shift in monitoring the Danube Delta (Romania): Insights from Earth Observation Data

Article References: Budileanu, M., Serban, I. D., & Constantinescu, S. (2025). A paradigm shift in monitoring the Danube Delta (Romania): Insights from Earth Observation Data. Scientific Annals of the Danube Delta Institute, 30, 55-66. https://doi.org/10.3897/saddi.30.160580

Image Credits: AI Generated

DOI: 10.3897/saddi.30.160580

Keywords: Danube Delta, Earth Observation, Sentinel-2, Copernicus, vegetation fires, remote sensing, wetland monitoring, FIRMS, EFFIS, Landsat, Romania, satellite imagery

The research, led by Sheila Baber of the University of Maryland College Park together with colleagues at Monash University and the University of Strasbourg, tackles a problem that has long frustrated satellite scientists: how do you measure something as subtle as irrigation in a territory you cannot enter? Ground surveys are impossible in occupied, front-line regions, and national statistics are too coarse and too slow to capture a rapidly evolving crisis. The team’s answer was a semi-supervised machine learning pipeline that blends multiple streams of satellite data into a single, field-scale picture of who still has water and who does not.

The technical core of the approach rests on a simple physical insight: irrigated fields look different from rainfed ones in measurable ways. They are cooler, because evaporating water carries away heat, a signal captured by the thermal infrared sensors aboard Landsat 8 and Landsat 9. They are greener and moister, which shows up in vegetation indices such as NDVI, NDMI, and the Green Chlorophyll Vegetation Index derived from Sentinel-2 imagery. And they consume more water than rainfall alone can supply, a discrepancy revealed by comparing actual evapotranspiration from the FAO WaPOR dataset against cumulative precipitation from the CHIRPS rainfall record. Each of these inputs was resampled to a uniform 30-meter resolution, matching the footprint of a Landsat pixel and the scale of the region’s large center-pivot fields.

Because no ground truth exists inside the war zone, the team had to bootstrap their own labels. For the years 2019 through 2022, they ran an unsupervised K-Means clustering algorithm across the multi-sensor feature stack, initially splitting the landscape into nine clusters to account for both crop type and water status. Analysts then assigned irrigation labels to each cluster by cross-checking against visible pivot irrigation systems, proximity to canals, and the behavior of vegetation, thermal, and evapotranspiration time series. These cluster-derived labels then trained a random forest model of 300 trees, which was applied to the uncertain years of 2023 through 2025, when the dam’s destruction meant that irrigation could no longer be assumed to follow historical patterns.

The transition year of 2023 posed particular challenges. Some fields showed early-season irrigation signatures before the dam failed and then dried out mid-season, creating an intermediate class the researchers called interrupted irrigation. Overgrown abandoned fields near the front lines also mimicked the wet, cool signatures of genuine irrigation. To handle these ambiguities, the team supplemented their automated pipeline with AlphaEarth embedding features, latent representations learned from global satellite data, and a small hand-labeled corpus of fewer than 500 reference points. For 2024 and 2025, they reverted to the original feature set, reasoning that fields still equipped for irrigation would be able to water their crops without interruption.

Validation relied on a stratified random sampling design, a statistical framework that corrects for the biases inherent in any classification map. The researchers drew 1,577 response points, stratified by how frequently each location had been mapped as irrigated between 2020 and 2023, and then photointerpreted each point using high-resolution Planet imagery and the same physical criteria used in classification. Because the strata were defined by irrigation frequency rather than by the map classes themselves, a post-stratification estimator was applied to produce bias-adjusted area figures with honest uncertainty bounds.

The numbers tell a stark story. Before the full-scale invasion, between 2019 and 2021, irrigated area in the two oblasts averaged 187.94 thousand hectares, with an uncertainty of plus or minus 14.43 thousand hectares, representing about 4 percent of baseline cropland and 8 percent of summer-season cropland. In 2022, as Russian occupation expanded, that figure fell to 140.04 thousand hectares. The dam’s destruction in 2023 cut it further to 74.97 thousand hectares, with fields showing a patchwork of outcomes: some lost irrigation entirely when the reservoir drained, while others briefly retained water in canal networks. By 2024, irrigated area had collapsed to 15.95 thousand hectares, roughly a 91 percent decline from the pre-war average, and by 2025 it stood at just 5.76 thousand hectares.

Equally revealing is what happened to the land itself. Fields that had been irrigated for water-hungry summer crops such as maize, soy, and sunflower were systematically abandoned or converted to winter crops like wheat, barley, and rapeseed, which require far less supplemental water. By 2024, more than half of the remaining irrigated extent was planted with winter crops, a clear signal of farmers adapting to a world without the Kakhovka reservoir. The accuracy assessment reflects the difficulty of the later years: user’s accuracy for the irrigation class dropped from above 93 percent in pre-war years to as low as 21 percent in 2024, when irrigation had become an extreme minority class covering less than 0.3 percent of cropland. The authors stress that their stratified area estimates remain statistically sound despite these map-level confusions, because the sampling design mathematically corrects for classification bias.

The findings carry weight beyond Ukraine. Irrigation accounts for roughly 70 percent of human freshwater withdrawals globally, and the methods developed here offer a template for monitoring water infrastructure failure in any conflict zone or inaccessible region. The World Bank has estimated that rebuilding the Kakhovka dam and its pumping stations would cost around 1.5 billion US dollars, work that cannot begin while fighting continues. Climate projections add urgency: CMIP models suggest southern Ukraine will grow hotter and drier, with the growing season lengthening by about a month by mid-century even as precipitation declines by roughly 26 millimeters. The study’s authors argue that whatever form reconstruction takes, decentralizing irrigation infrastructure away from single points of failure like the Kakhovka network will be essential, and that satellite-based monitoring capacity will be crucial for managing water in a changing climate.

Limitations remain. The evapotranspiration product is available only at 300-meter resolution and precipitation data at roughly 5.5 kilometers, which limits applicability to smaller fields, and a classifier trained in this semi-arid southern region may over-detect irrigation if transferred to cooler climates. The study also does not attempt to quantify the volume of water applied, only the extent of irrigated land. Still, the demonstration is powerful: a transferable, semi-supervised workflow that can be rapidly deployed to map crop water use where no ground crew can safely walk, providing an evidentiary foundation for reparations discussions, reconstruction planning, and the broader science of agriculture under conflict.

Subject of Research: Satellite mapping of irrigation loss in southern Ukraine following the Kakhovka Dam destruction

Article Title: A semi-supervised multi-sensor remote sensing pipeline for mapping irrigation dynamics in ground-truth deficient conflict zones: a case study of the Kakhovka Dam failure

Article References: Baber, S., Skakun, S., Sadeh, Y., Shumilo, L., Oliinyk, O., Hosseini, M., Wagner, J., Khan, A. Q., Nair, S., Kotcharlakota, A., & Becker-Reshef, I. (2026). A semi-supervised multi-sensor remote sensing pipeline for mapping irrigation dynamics in ground-truth deficient conflict zones: a case study of the Kakhovka Dam failure. Frontiers in Remote Sensing, 7, Article 1920951. https://doi.org/10.3389/frsen.2026.1920951

Image Credits: AI Generated

DOI: 10.3389/frsen.2026.1920951

Keywords: Ukraine, Kakhovka Dam, irrigation, remote sensing, satellite imagery, machine learning, evapotranspiration, food security, conflict zones, crop mapping, Sentinel-2, Landsat

The research team, led by Tolulope Ayodeji Olatoye with Glynnis Elizabeth Daries and Raymond Nkwenti Fru, built their analysis on the Normalized Difference Built-up Index, or NDBI, a remote-sensing technique that exploits the way impervious surfaces reflect shortwave infrared light differently from vegetation and water. By computing NDBI values from Landsat 5 Thematic Mapper imagery captured in 2004 and Landsat 8 Operational Land Imager scenes from 2014 and 2024, the researchers could track built-up density across a province spanning roughly 372,831 square kilometres, an area dominated by the arid Karoo Basin where water security and land stability determine whether settlement is sustainable at all. Cloud-free dry-season scenes were radiometrically calibrated, atmospherically corrected using dark object subtraction, and geometrically aligned so that change over time could be measured on a consistent footing.

What the satellites recorded is a landscape undergoing dramatic transformation. Grassland, which covered 74.37 percent of the province in 2004 at 277,267 square kilometres, collapsed to just 32.96 percent by 2024, a loss of more than 154,000 square kilometres. Barren land expanded nearly threefold over the same interval, from 73,417 to 222,867 square kilometres, a signature of accelerating desertification. Dense vegetation fared worst of all, plummeting 86 percent from 9,022 square kilometres to a mere 1,262. Meanwhile, built-up area grew by approximately 292 percent, from 91 to 357 square kilometres, a seemingly modest absolute figure that carries enormous significance in a province where towns are separated by hundreds of kilometres of semi-desert.

The crucial insight came from disaggregating that growth. Using Random Forest supervised classification with 500 decision trees, validated against 600 reference points per epoch with overall accuracies between 83 and 88 percent and Kappa coefficients above 0.77, the team mapped six land-cover classes and then overlaid the resulting built-up clusters with land tenure records and infrastructure networks. The result was unambiguous: 78 percent of new built-up growth occurred within five kilometres of the mining-transport corridors that thread through the province, most notably the Sishen–Saldanha iron ore export line. Along that corridor, built-up density fell from 4.87 square kilometres per 100 square kilometres within the five-kilometre buffer to just 0.12 beyond fifty kilometres, a 97.5 percent drop that the authors describe as infrastructure functioning simultaneously as an artery of development and a vector of exclusion.

Spatial statistics confirmed the pattern was no accident of visual interpretation. Global Moran’s I, a measure of spatial autocorrelation ranging from minus one for perfect dispersion to plus one for perfect clustering, returned a value of 0.347 with a z-score of 8.94 and a p-value below 0.001, indicating that built-up density and degradation patterns cluster non-randomly across the province’s 233 municipal wards. Local Indicators of Spatial Association, or LISA, pinpointed statistically significant degradation hotspots in the communal rangelands of Namaqualand, with a local Moran’s I of 0.412, and the Kalahari communal areas at 0.387, both highly significant, while cold spots of low degradation coincided with the corridor buffer zones and the urban nodes of Kimberley and Upington.

The tenure contrast is the study’s most striking finding. Degradation density on communal tenure land, the spatial legacy of apartheid’s Bantustan policies, reached 66.9 square kilometres per 100 square kilometres, compared with 32.9 on freehold land, 47.6 in Black majority wards, and just 14.0 in White majority wards. Relative risk ratios placed communal areas at 4.8 times the degradation risk of White majority wards and 2.1 times that of freehold zones. The authors benchmark these figures against comparable arid-region studies and find the disparity exceeds the roughly two-to-threefold accessibility gaps reported for hot-arid Isfahan in Iran and for green-space equity in the oasis city of Urumqi, China, suggesting that South Africa’s tenure-based spatial legacy compounds rather than merely parallels infrastructure-driven inequality elsewhere in the Global South.

To convert these observations into a planning instrument, the team constructed a Composite Spatial Justice Index, or CSJI, aggregating three purpose-built indicators: a Proximity to Service Deprivation Index capturing cumulative deprivation scores that climb from 0.12 for settlements within two kilometres of services to 1.00 beyond twenty kilometres; the Degradation-Vulnerability Overlay Index tying degradation burden to tenure category; and an Infrastructure Access Gradient measuring distance-decay from the named corridor. Each sub-indicator was min-max normalized to a common zero-to-one range and combined with equal weights, a choice the authors stress was made before data collection as a deliberate design decision rooted in Plaatje’s critique of dispossession rather than retrofitted to the results.

The index survived rigorous validation. A sensitivity analysis using Principal Component Analysis to derive an alternative, data-driven weighting scheme produced ward rankings correlated with the equal-weighting scheme at a Spearman rank correlation of 0.88, and identified the same top-quartile Spatial Justice Priority Areas in 91 percent of cases. Bootstrap resampling with 1,000 iterations yielded 95 percent confidence intervals on every ward score, and comparison against an independent benchmark produced a Pearson correlation of 0.92. The authors are equally candid about limits: the design is cross-sectional and correlational, so it identifies where disparities co-occur and cluster, not why they arose. Confounders such as rainfall variability, grazing intensity, and soil type were not independently modeled, and the tenure-degradation link is framed as a motivating association warranting causal investigation rather than proof of mechanism.

The philosophical framing is what sets the work apart. Sol Plaatje, the journalist, novelist, and founding Secretary-General of the South African Native National Congress, arrived in Kimberley in 1894 as a postal worker and spent his career documenting how the 1913 Natives Land Act confined Black South Africans to ecologically marginal territory and restricted their freedom of movement, a grievance he famously illustrated with the observation that birds cross borders freely regardless of color. The study maps each of his principles onto an analytical component: legal equality underpins the CSJI, resistance to dispossession motivates the tenure-differentiated degradation analysis, the mobility critique becomes the infrastructure access gradient, and his practice of education through documentation informs the use of NDBI clustering as a pedagogical tool that makes hidden geographies of inequality visible to communities and policymakers alike.

The policy implications are concrete. The findings align with South Africa’s Spatial Planning and Land Use Management Act of 2013, which mandates inclusive and developmental spatial planning, as well as the National Development Plan 2030 and the Integrated Urban Development Framework. The authors recommend designating strategic densification zones along existing corridors at Kimberley, De Aar, and the Orange River, paired with mixed-use zoning and affordable housing, while redirecting infrastructure investment toward peripheral areas to flatten the access gradient and prioritizing community-based rangeland management in the Namaqualand and Kalahari hotspots. They also propose a provincial Spatial Justice Observatory to maintain standardized monitoring using the CSJI. The researchers caution that the index is a transferable diagnostic approach rather than a fixed template, requiring adaptation to local data conditions, but its core message stands: in the arid geographies Plaatje once walked, satellite pixels can now measure the injustice he could only describe, and the evidence demands that densification be planned as an instrument of equity rather than left to reproduce the exclusionary logics of the past.

Subject of Research: Geospatial measurement of settlement morphology, land tenure inequality, and inclusive densification in South Africa's arid Northern Cape Province using NDBI clustering and a composite spatial justice index

Article Title: Settlement morphology, inclusive densification, and Sol Plaatje's spatial justice: NDBI clustering pedagogy for equitable urban growth in South African arid geographies

Article References: Olatoye, T. A., Daries, G. E., & Fru, R. N. (2026). Settlement morphology, inclusive densification, and Sol Plaatje's spatial justice: NDBI clustering pedagogy for equitable urban growth in South African arid geographies. Frontiers in Sustainable Resource Management, 5, Article 1921810. https://doi.org/10.3389/fsrma.2026.1921810

Image Credits: AI Generated

DOI: 10.3389/fsrma.2026.1921810

Keywords: NDBI, spatial justice, Northern Cape, remote sensing, land tenure, urban densification, Landsat, Moran's I, Sol Plaatje, arid environments, GIS, South Africa

The study focused on seven coastal subdistricts—Panceng, Ujungpangkah, Bungah, Sidayu, Manyar, Gresik, and Kebomas—chosen because they harbor the region’s principal mangrove ecosystems while experiencing sharply different levels of urban pressure. Gresik Regency is a strategic and rapidly developing coastal zone where industrialization, urbanization, and transport continue to intensify, placing simultaneous pressure on the atmosphere and on the coastal wetlands that might otherwise buffer it. Indonesia holds the largest mangrove area on Earth, but between 1985 and 2019 the country lost roughly 430,000 hectares of mangroves to aquaculture, agriculture, plantations, and coastal development, a loss that directly erodes the national carbon sink.

The methodology combined ground truth with satellite eyes. In the field, survey teams laid 100-meter transects perpendicular to the shoreline and placed 10 by 10 meter observation plots along each line at 0, 50, and 100 meters. Within these plots they identified mangrove species and measured tree diameters at breast height, feeding those measurements into species-specific allometric equations for genera such as Rhizophora, Avicennia, Sonneratia, Ceriops, Lumnitzera, Xylocarpus, and Bruguiera. Biomass was converted to carbon content using the standard assumption that carbon constitutes half of dry biomass, and carbon stocks were then translated into absorbed carbon dioxide using the atomic mass ratio of CO2 to carbon, a conversion factor of 3.67 derived from the atomic weights of carbon and oxygen.

Space-based analysis relied on 2023 Landsat 8 imagery processed on the Google Earth Engine platform. Cloud-masking algorithms produced a clean annual composite, from which the Normalized Difference Vegetation Index was calculated as an input to machine-learning classification using Random Forest and CART algorithms. Training data came from the field plots, and classification accuracy was verified with a confusion matrix. Thirteen mangrove species were confirmed across the coast, with Avicennia and Rhizophora dominating. The mapping revealed dramatic contrasts: Ujungpangkah Subdistrict held 1,615.54 hectares of mangrove, while Kebomas, one of the most developed areas, retained just 0.75 hectares, with much of its former mangrove land repurposed for development.

On the emissions side, the researchers inventoried carbon dioxide equivalents from transport fuel, livestock and poultry, liquefied petroleum gas consumption, electricity use, and population-related activities, applying standard emission factors such as 2.31 kilograms of CO2 per liter of gasoline and 3 kilograms of CO2 per kilogram of LPG. The totals exposed a stark urban-rural gradient. Gresik Subdistrict, anchored by urban mobility and logistics, emitted the most at 6,795.91 tonnes of CO2-equivalent per year, with transportation alone contributing 5,712.52 tonnes—roughly 84 percent of its total. Registered vehicles in the regency number 801,969, a figure that underscores how rapid motorization drives coastal emissions. At the other extreme, Sidayu, with its traditional fisheries and agriculture, produced the least, at 2,375.30 tonnes per year.

Mangrove carbon performance proved equally uneven, and in an instructive way. Bungah recorded the highest carbon stock density at 189.35 tonnes of carbon per hectare and the highest sequestration rate at 694.91 tonnes of CO2 per hectare per year, yet Ujungpangkah’s sheer extent gave it the largest totals: 264,803.16 tonnes of stored carbon and 971,827.60 tonnes of annual CO2 sequestration. Ujungpangkah’s forests flourish at the estuary of the Bengawan Solo River, whose heavy sediment load—over 1,700 tonnes of suspended material annually—continuously builds new land through accretion. A telling comparison emerged between Manyar and Sidayu: despite having more mangrove area, Manyar stored less carbon overall, showing that stand structure, species composition, density, and productivity matter as much as raw acreage.

Balancing the ledger against local emissions, four subdistricts—Ujungpangkah, Bungah, Manyar, and Sidayu—emerged as net carbon sinks with absorption capacity exceeding their anthropogenic output. Ujungpangkah’s mangroves sequester nearly a million tonnes of CO2 annually against local emissions of just 2,750.34 tonnes, a net balance of minus 969,077.26 tonnes. But Kebomas, Panceng, and Gresik all ran deficits. Gresik Subdistrict’s mangroves, covering only 10.67 hectares, store a mere 346.03 tonnes of carbon and sequester 1,269.92 tonnes of CO2 per year, leaving more than 5,500 tonnes of CO2-equivalent unabsorbed each year. Kebomas, with three-quarters of a hectare of mangrove against 5,220.32 tonnes of annual emissions, is the most ecologically insolvent of all.

The spatial arithmetic of restoration follows directly. To neutralize current emissions, the analysis concluded that Kebomas needs an additional 52.07 hectares of mangrove for annual CO2 sequestration—or 191.08 hectares total to also restore ideal carbon storage. Panceng requires 50.21 additional hectares for sequestration, and Gresik 57.10 hectares, with Gresik needing the largest storage-based expansion at 198.89 hectares beyond its existing 10.67. The authors argue that in these urban-industrial zones, policy must move beyond passive conservation toward active landscape reconstruction, ideally prioritizing fast-growing, high-sequestration species, and that rehabilitation targets should be anchored to minimum area requirements rather than aspiration.

The broader significance of the study lies in its framework as much as its numbers. Few previous works have integrated field-based biomass inventories, satellite-derived mangrove mapping, and comprehensive emission accounting at the subdistrict scale, and fewer still have translated the resulting deficits into concrete hectare targets. Because mangroves sequester carbon in both woody biomass and organic coastal sediments—storing more carbon per unit area than most tropical terrestrial forests and holding it for decades to centuries—they represent one of the most efficient natural climate solutions available to coastal nations. They also deliver co-benefits that engineered offsets cannot: wave damping, shoreline protection against abrasion and sea-level rise, habitat for coastal biota, and filtration of pollutants through their root systems.

For Indonesia, the stakes extend to international commitments under the Paris Agreement and to the integrity of coastal communities facing tidal flooding and erosion. The Gresik results show that even where a regency enjoys an overall carbon surplus, thanks largely to one conserved estuarine forest, urban zones can accumulate unabsorbed emissions as habitat fragmentation and land conversion strip away sink capacity. The researchers frame their hectare-level estimates as a practical planning tool for blue carbon management: site-specific restoration quotas that can guide where to plant, how much to plant, and how to monitor whether restored mangroves are actually closing the emissions gap. In a warming century, the message from this stretch of the Java Sea is that every hectare of mangrove now carries a quantifiable climate price—and that the deficit, in some places, is only about fifty hectares wide.

Subject of Research: Estimating the mangrove area required to offset anthropogenic carbon emissions in coastal Gresik Regency, Indonesia

Article Title: Estimation of mangrove requirements for the absorption and storage of anthropogenic carbon emissions in the coastal area of Gresik Regency, East Java Province, Indonesia

Article References: Sodikin, Salsabiela, M., & Candra Maulana, Y. (2026). Estimation of mangrove requirements for the absorption and storage of anthropogenic carbon emissions in the coastal area of Gresik Regency, East Java Province, Indonesia. BioRisk, 24, 169-184. https://doi.org/10.3897/biorisk.24.189492

Image Credits: AI Generated

DOI: 10.3897/biorisk.24.189492

Keywords: mangroves, blue carbon, carbon sequestration, carbon emissions, Gresik Regency, Indonesia, remote sensing, Landsat, GIS, mangrove restoration, climate change mitigation, coastal management

The authors frame this chain as a progression through distinct levels, beginning with multi-source sensing and ending with outcomes such as low-carbon urban development. At the foundation of that chain sits a deceptively simple question: how well do different Earth-observing systems actually represent the wildly heterogeneous surfaces of a city? A cross-sensor comparison featured in the collection examined imagery from Landsat, Sentinel-2, and the Italian hyperspectral mission PRISMA over the same urban setting. The results dismantled any hope of finding a single best satellite for urban work. Hyperspectral data improved the separation of certain transitional urban-density classes, where fine spectral detail distinguishes surfaces that broadband sensors blur together. Yet those gains proved class-specific and heavily conditioned by spatial resolution, acquisition timing, preprocessing choices, and the behavior of the classifier applied. The takeaway is that sensor selection is not a search for a universally superior platform but a choice among complementary observational strengths, each suited to different urban questions.

Still, spectral, spatial, and temporal information, however carefully combined, describe what a city looks like far more readily than what it does. A second contribution in the collection tackled exactly this gap by integrating high-resolution imagery with nighttime lights, building information, and points and areas of interest to identify urban functional zones in three of China’s largest cities: Beijing, Chengdu, and Shanghai. This fusion of physical and socioeconomic representations pushed overall mapping accuracies above 90 percent, a remarkable figure for a task as conceptually slippery as deciding where a district functions as commercial, residential, industrial, or mixed use. The work draws on a broader current in the literature linking remote sensing with social sensing and semantic information to capture the mixed and probabilistic nature of urban functions, where a single block may simultaneously host housing, retail, and logistics.

But the study also exposed one of the field’s most persistent weaknesses: transferability. When the models trained on one Chinese megacity were applied to another, performance declined markedly, indicating that functional signatures remain tightly bound to local urban structure. A commercial zone in Chengdu does not radiate the same multi-source fingerprint as one in Shanghai. For planners hoping to export mapping tools across borders and climates, the message is sobering. The editorial’s authors treat this not as a failure but as a structural insight: urban functional intelligence is context-laden, and any framework claiming global applicability must explicitly account for how place shapes the relationship between what a city looks like and what it does.

Representation becomes genuinely useful for resilience only when it helps explain environmental processes tied to urban form and function. A seasonal analysis of land surface temperature across four functional zones on Xiamen Island, included in the collection, demonstrated that urban heat is not governed by a single city-wide relationship. The relative influence of geographic position, building density, population density, floor-area ratio, and landscape configuration shifted depending on land use and season. This finding echoes wider evidence that the cooling effects of green infrastructure, and the drivers of urban heat more generally, vary with morphology, season, and location. Recent work modeling wind-driven vegetation cooling through a source-medium-receiver framework reinforces the point that thermal resilience emerges from interactions among heat sources, the urban medium, and receiving populations, not from any one variable.

The practical implication is significant: resilience planning demands spatially differentiated and temporally adaptive analysis rather than universal prescriptions. A recommendation to plant trees along boulevards may cool a coastal district in summer while doing little for a dense inland commercial core in a different season. The Xiamen study’s context dependence suggests that heat mitigation investments should be guided by zone-specific and season-specific models, and that planners should be wary of benchmarks imported from other cities. In an era when climate adaptation budgets are finite and heat waves are intensifying, that nuance could translate directly into saved lives and avoided infrastructure costs.

Analysis, however, only matters if it changes what gets built or governed, and this is where the Research Topic turns from observation to intervention. A quasi-natural-experiment evaluation of China’s AI innovation-pilot-zone policy found that AI-oriented policy can contribute to lower urban carbon emissions through pathways including green technological innovation and environmental regulation. Crucially, the effects were conditioned by enabling factors such as digital infrastructure and financial development; AI policy did not operate in a vacuum. The editorial notes that this finding broadens the role of AI beyond an analytical technique, showing that technological capability can contribute to resilience-related outcomes when embedded in institutions, infrastructure, and policy. It is a rare empirical bridge between the governance layer of urban systems and measurable low-carbon outcomes.

Reading the collection as a whole, the authors extract three lessons that could serve as a manifesto for the next decade of urban GeoAI. First, urban resilience requires multiple representations: no single sensor, dataset, or model can capture a city’s physical, functional, environmental, and socioeconomic complexity at once. Second, context and transferability matter at every stage; sensor performance, functional-zone mapping, thermal drivers, and even policy effectiveness vary with place, season, urban form, and institutional capacity, and models developed in one setting may not transfer cleanly to another. Third, resilience depends on connecting analytical levels: observation alone is insufficient, and value is created only when representation supports understanding, understanding enables diagnosis, and diagnosis informs intervention. Each lesson underscores that the field’s hardest problems are less about raw computing power than about integration and interpretation.

The editorial also looks forward, and here it is candid about what the collection did not achieve. None of the four contributions implemented a geospatial foundation model or an urban digital twin, and the authors explicitly include these elements in their conceptual framework as forward-looking opportunities drawn from the wider literature rather than as accomplishments of the collection itself. Geospatial foundation models, which learn reusable representations from vast and increasingly multimodal geospatial archives, are shifting attention away from bespoke, single-city classifiers toward general-purpose urban intelligence. In parallel, urban digital twins offer environments in which heterogeneous observations, models, simulation, visualization, and feedback can be organized around concrete planning and operational decisions, potentially bridging the stages from sensing to action that the framework describes.

Yet greater technical capability, the authors caution, does not eliminate the need for interpretability, uncertainty communication, and institutional accountability. As model outputs move closer to decisions with material consequences, such as where to site cooling infrastructure or how to regulate emissions, explainable GeoAI becomes essential rather than optional. The editorial proposes a demanding yardstick for the next generation of urban GeoAI: systems should be judged not only by classification accuracy or predictive performance but by whether they generalize across cities, represent uncertainty honestly, integrate physical and social dimensions, and deliver information at the scales and times at which urban decisions are actually made. The transition from observing cities to building usable urban intelligence, the authors conclude, is where remote sensing and AI can make their most consequential contribution to resilient urban futures, turning streams of pixels and data points into knowledge that planners, regulators, and communities can act upon before the next heat wave, flood, or emissions target arrives.

Subject of Research: Remote sensing and artificial intelligence for building resilient cities

Article Title: Editorial: Building resilient cities with remote sensing and AI

Article References: Abbas, S., Qamer, F. M., & Zia, A. (2026). Editorial: Building resilient cities with remote sensing and AI. Frontiers in Remote Sensing, 7, Article 1991820. https://doi.org/10.3389/frsen.2026.1991820

Image Credits: AI Generated

DOI: 10.3389/frsen.2026.1991820

Keywords: remote sensing, artificial intelligence, urban resilience, GeoAI, urban heat, functional zones, Landsat, Sentinel-2, PRISMA, digital twins, foundation models, carbon emissions