The study, published in Earth Science Informatics, focused on the Taleghan Dam watershed, a vast catchment of roughly 124,062 hectares in Alborz Province. This semi-arid landscape, tucked into the Alborz mountain range, experiences the classic problems of mountainous drylands: intense seasonal rainfall, erodible slopes, and a delicate balance between water scarcity and destructive floods. The research team, led by Omid Asadi Nalivan of the University of Maragheh together with colleagues from Iran and India, set out to answer two intertwined questions: which locations are most suitable for building watershed dams and check dams, and which parts of the landscape are most prone to flooding in the first place.

What makes the approach remarkable is its grounding in real-world evidence rather than purely theoretical assumptions. The researchers compiled an inventory of 67 stone masonry check dams that had already been built in the watershed, treating these implemented structures as verified examples of successful siting decisions. They also assembled 160 historical flood occurrence points, documenting places where flooding had actually happened. These datasets served as the ground truth for training and testing three distinct machine learning algorithms: Maximum Entropy, better known as MaxEnt; Support Vector Machine, or SVM; and Artificial Neural Network, or ANN. Each algorithm learns differently, and comparing them reveals which approach best captures the complex interplay of factors that make a site suitable.

The models were fed an unusually rich set of environmental information. Twelve conditioning factors, spanning topography, hydrology, geology, and land cover, were compiled at a fine spatial resolution of 10 by 10 meters, meaning every pixel of the final maps represents a patch of ground about the size of a small living room. Before modeling began, the team checked all twelve variables for multicollinearity, the statistical problem that arises when input factors overlap so heavily that models become unstable. Every factor passed the test, with variance inflation factors below 10 and tolerance values above 0.1, the conventional thresholds that signal acceptable independence among predictors.

The results were striking. All three algorithms achieved area under the curve values, the standard measure of predictive skill, in the range of 0.80 to 0.93 for check dam site suitability, a band the researchers classify as very good to excellent. MaxEnt emerged as the clear winner, reaching a validation AUC of 0.93, meaning it correctly distinguished suitable from unsuitable locations with remarkable reliability. For flood susceptibility, MaxEnt again performed strongly, with training and validation AUC values of 0.91 and 0.89 respectively. The datasets were split 70:30 between training and validation, a standard practice that ensures models are judged on data they have never seen, guarding against the trap of simply memorizing the training examples.

Perhaps the most scientifically interesting outcome came from the Jackknife analysis, a technique that measures how much each environmental variable contributes to model performance by systematically removing one factor at a time. Five predictors stood out as the most informative: the Stream Power Index, stream order, slope, drainage density, and rainfall. The Stream Power Index, which combines the accelerating effect of slope with the accumulating effect of upstream flow area, essentially quantifies the erosive energy of flowing water at any point on the landscape. Its dominance makes intuitive sense: check dams work precisely where water has enough energy to cause erosion and flooding, but where the terrain allows a barrier to be built and to function effectively.

Statistical testing reinforced this picture. For flood susceptibility, stream order emerged as the most significant factor, with a coefficient of 0.349 and a p-value of 0.002, while the Stream Power Index showed a strong negative coefficient of minus 1.149 with a p-value below 0.001. In plain terms, the position of a location within the stream network hierarchy, and the erosive power of water moving through it, largely determine whether floods occur there. Higher-order stream reaches, the main channels where tributaries converge, concentrate flow and therefore concentrate both flood risk and the potential benefit of well-placed barriers.

When the trained models were applied across the entire watershed, they identified 147 prioritized sites with substantial potential for watershed dam construction. These locations share a consistent profile: they lie along higher-order stream reaches where flow is concentrated, on moderate slope gradients that allow construction without excessive engineering challenges, and on geotechnically stable rock formations that can anchor a structure securely. Just as revealing are the areas the models rejected. Gypsum-bearing geological units, mapped as the Ekgy formation, were consistently rated as low suitability, and so were zones close to active faults. Gypsum dissolves slowly in water, undermining the foundations of any structure built on it, while fault-proximate zones carry seismic risk that could crack or collapse a dam. The models, in effect, rediscovered sound engineering judgment from the data alone.

The dual mapping of check dam suitability and flood susceptibility gives watershed managers something they have rarely had before: a single, spatially explicit framework that shows both where floods threaten and where interventions will succeed. Instead of allocating conservation budgets across a whole province uniformly, or relying on the intuition of individual surveyors, planners can now direct investment to the 147 identified priority sites, confident that each one combines hydrological need with physical feasibility. The approach is also transferable. Because it relies on freely available geospatial data layers such as digital elevation models, satellite-derived land cover, and geological maps, the same workflow can be applied to mountainous semi-arid catchments anywhere in the world, from Central Asia to the Mediterranean to the American Southwest.

The timing could hardly be better. Around the world, check dams are enjoying renewed attention as climate change intensifies both floods and droughts, and as countries seek nature-based and low-cost solutions to water security. Recent studies from China’s Loess Plateau, where hundreds of thousands of check dams have transformed eroded landscapes, to Jordan, Syria, Nigeria, and Ghana, show growing global interest in these structures. Yet failures happen too, often because dams were sited on unstable ground or in channels where they could not withstand the forces they were meant to control. By demonstrating that machine learning models trained on real construction records can predict suitability with AUC values above 0.9, the Taleghan study offers a template for making every future check dam count. For semi-arid regions facing harsher, less predictable climates, that could mean the difference between conservation budgets that build resilience and money that simply washes downstream.

Subject of Research: Machine learning-based site selection for watershed conservation structures and flood susceptibility mapping in a semi-arid Iranian watershed

Article Title: Feasibility assessment of watershed conservation structure site selection using machine learning models in a semi-arid area, Iran

Article References: Nalivan, O. A., Yousefi, S., Shahbazi, A., & Shahi, N. R. (2026). Feasibility assessment of watershed conservation structure site selection using machine learning models in a semi-arid area, Iran. Earth Science Informatics, 19(11), Article 194. https://doi.org/10.1007/s12145-026-02251-2

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02251-2

Keywords: check dams, machine learning, MaxEnt, flood susceptibility, watershed management, Taleghan watershed, Iran, GIS, soil conservation, semi-arid, Stream Power Index, site suitability

The research, led by Zufan Desta of Adigrat University and Mekelle University together with Emiru Birhane, Mitiku Haile, and Dawit Gebregziabher, focused on the Mechahile subwatershed in the Atsbi-Wenberta district, a highland zone sitting between 2,300 and 3,200 meters above sea level. The region receives roughly 635 millimeters of rain annually, almost all of it concentrated between June and September, under a mean annual temperature of about 14.7 degrees Celsius. Dominant soils include Leptosols, Regosols, Cambisols, and Fluvisols, thin and fragile profiles that erode rapidly under intense seasonal rainfall and long histories of cultivation and grazing. Between 2011 and 2020, teams implemented a suite of physical structures, including deep trenches, trench bunds, half-moons, micro basins, and stone-faced soil bunds, alongside biological interventions such as exclosures and plantations of native and adapted species including Juniperus procera, Olea europaea subsp. cuspidata, and Eucalyptus globulus.

The experimental design is what sets this study apart from much of the existing conservation literature. Rather than simply comparing a treated site with a distant control, the researchers built a factorial randomized complete block design across the hillside. They stratified the landscape into upper, middle, and lower slope positions, sampled two soil depths of 0 to 30 and 30 to 60 centimeters, and placed sampling plots at two distance intervals from conservation structures, 0 to 4 meters and 4 to 8 meters. In total, 96 composite soil samples were collected from 40 by 40 meter plots, supplemented by undisturbed core samples for bulk density determination. Treated plots and adjacent untreated, open-access grazing land were matched for aspect, soil type, topography, land use history, and agroecological conditions, so that observed differences could be attributed primarily to the conservation measures themselves.

The physical results are striking. Bulk density, a measure of soil compaction, dropped by up to 6 percent in conserved plots, with the lowest value of 1.56 grams per cubic centimeter recorded on treated lower slopes, compared with the highest value of 1.66 grams per cubic centimeter in untreated upper-slope soils exposed to erosion and trampling by livestock. Lower bulk density means more pore space, better root penetration, and faster water infiltration, all of which are prerequisites for vegetation recovery on degraded land. Water-stable aggregates, the soil crumbs that resist breakdown by raindrops and running water, increased by 28 percent in treated plots, reaching 64.03 percent in the topsoil of conserved lower slopes against 50.16 percent in untreated areas. Stable aggregates are the architecture of a functioning soil; without them, seasonal rains strip away fine particles and the organic matter bound to them.

Perhaps the most consequential finding concerns water. Soil moisture content was 67 percent higher in treated lower-slope plots at the 30 to 60 centimeter depth, reaching 7.99 percent, than in untreated upper-slope surface soils, which averaged just 4.77 percent. In a semi-arid system with a single annual rainy season, that difference is the difference between seedlings surviving the dry season and dying. The researchers attribute the moisture gains to a combination of reduced runoff behind physical barriers, improved infiltration through less compacted soil, and the water-holding capacity conferred by accumulating organic matter. Notably, moisture peaked in the deeper layer while aggregate stability peaked in the surface layer, a depth-dependent pattern that reflects reduced evaporation and finer subsoil textures below ground and concentrated litter inputs above.

The chemical story mirrors the physical one. Soil organic matter, total nitrogen, available phosphorus, and pH were all significantly higher in treated plots, particularly at lower slope positions and in the topsoil. The pattern follows the logic of erosion and deposition: upper slopes lose fine particles and nutrients, while lower slopes act as depositional zones where sediment, organic residues, and base cations accumulate. Treated plots also showed higher pH, likely because organic matter improves the retention of calcium and magnesium ions and reduces leaching, which buffers acidity and supports nutrient uptake and microbial activity. The interaction effects among management, slope position, and soil depth were statistically significant for moisture, aggregates, and all chemical indicators, confirming that conservation effectiveness is not uniform across a landscape but depends critically on where interventions are placed.

Distance from the structures themselves emerged as a decisive variable. Every measured property, from moisture and aggregate stability to organic matter and phosphorus, was highest within 0 to 4 meters of a bund or trench and declined measurably by 4 to 8 meters. Bulk density was slightly lower near the barriers. This spatial gradient demonstrates that conservation structures work locally, slowing runoff, trapping sediment, and creating fertility hotspots. The practical implication is that bund spacing matters: structures placed too far apart leave wide bands of hillside effectively untreated, a finding consistent with earlier work showing that tighter bund spacing improved soil water content in the Lake Tana basin.

Correlation analysis placed soil organic matter at the center of the entire restoration process. Organic matter correlated positively with available phosphorus, with a Pearson coefficient of 0.648, as well as with total nitrogen, water-stable aggregates, moisture content, and pH, and negatively with bulk density, with a coefficient of minus 0.729. In other words, as organic matter accumulates under exclosures and plantations, it simultaneously feeds nutrient cycling, binds soil particles into stable aggregates, holds water, and loosens compacted profiles. Texture also shifted along the degradation gradient, with untreated plots trending toward coarser, sandier compositions that lose nutrients and moisture more readily, while conserved plots retained more silt and clay, the particle fractions with the greatest surface area and cation exchange capacity.

The study is honest about its limits. Phosphorus remained very low across all plots, even in treated areas, indicating that conservation structures alone cannot correct every fertility deficit and that phosphorus-specific amendments may be needed. The observation window is short, the sampling confined to selected subwatersheds, and biological and hydrological indicators and socio-economic drivers of adoption were not assessed. Long-term monitoring across multiple agroecological zones, the authors argue, is essential to confirm that these gains persist and scale.

Even with those caveats, the message for policymakers is unusually actionable. Ethiopia has invested in soil and water conservation at national scale since the aftermath of the 1973-74 famine, through the integrated watershed programs of the 1980s and 1990s, the national exclosure program, the Productive Safety Net Programme from 2005, and the Sustainable Land Management Programme from 2008. This study provides the quantitative justification for that investment and, more importantly, a design principle for the next phase: targeted placement beats blanket coverage. Conservation structures deliver the greatest returns on lower slopes, in topsoil layers, and close to the structures themselves, which is precisely where erosion and deposition dynamics can be harnessed rather than fought. In a warming climate where rainfall in the Horn of Africa is becoming more erratic, rebuilding the sponge of degraded hillsides is not sentimental conservation. It is infrastructure, and the hills of Tigray are showing exactly how to build it.

Subject of Research: Effects of integrated soil and water conservation measures on physicochemical soil properties in degraded highlands of northern Ethiopia

Article Title: Effect of biophysical soil and water conservation measures on physicochemical soil properties and their implications for climate resilience in degraded hillside landscapes of Tigray, Ethiopia

Article References: Desta, Z., Birhane, E., Haile, M., & Gebregziabher, D. (2025). Effect of biophysical soil and water conservation measures on physicochemical soil properties and their implications for climate resilience in degraded hillside landscapes of Tigray, Ethiopia. BMC Environmental Science, 2(1), Article 24. https://doi.org/10.1186/s44329-025-00037-3

Image Credits: AI Generated

DOI: 10.1186/s44329-025-00037-3

Keywords: soil and water conservation, soil erosion, Tigray, Ethiopia, climate resilience, bulk density, soil moisture, water-stable aggregates, soil organic matter, exclosures, hillside restoration, drylands