Physics and Chemistry of the Earth最新文献

{"title":"Spatial variation in water quality, their ecotoxicological risks, and source apportionment in the Marmara Sea Watershed, Türkiye","authors":"Cem Tokatlı ,&nbsp;Said Muhammad ,&nbsp;Fikret Ustaoğlu ,&nbsp;Bayram Yüksel","doi":"10.1016/j.pce.2026.104280","DOIUrl":"10.1016/j.pce.2026.104280","url":null,"abstract":"<div><div>Streams and rivers are among the most vulnerable water resources to contamination due to their extensive contact with terrestrial environments. Contaminated streams and rivers pose significant threats to downstream ecosystems. This study examined water quality parameters in lotic or potamic habitats (n = 60) that flow into the Marmara Sea. Samples were collected during the spring season (May 12–26, 2024) and analysed for ten limnological parameters, including dissolved oxygen (DO), electrical conductivity (EC), pH, total dissolved solids (TDS), nitrate nitrogen (NO₃–N), nitrite nitrogen (NO₂–N), ammonium nitrogen (NH₄–N), phosphate phosphorus (PO₄–P), turbidity, and salinity. The average values of these parameters were 6.76 mg/L for DO, 4.00 NTU for turbidity, 1.23 mg/L for NO₃–N, 0.20 mg/L for NO₂–N, 0.91 mg/L for NH₄–N, and 0.35 mg/L for PO₄–P. Water quality was further examined using statistical and ecotoxicological indicators to assess potential risks to ecosystems and human health, including concerns related to mucilage formation. This study employed a range of statistical models—including principal component analysis (PCA), cluster analysis (CA), water quality index (WQI), nutrient pollution index (NPI), hazard quotient (HQ), hazard index (HI), and geospatial mapping—to deliver a comprehensive toxicological, statistical, and spatial assessment. The results identified the Gemlik and İzmit Gulfs as the highest-risk zones for the Marmara Sea. Additionally, high population densities, industrial activities, and inadequate infrastructure in the watersheds of the Gemlik, Gebze, Şerefli, and Çanakkale streams were identified as significant stressors for the Marmara Sea ecosystem.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104280"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-temporal InSAR deformation monitoring and pre-impoundment landslide inventory development with kinematic zonation along the Dasu Reservoir, Pakistan","authors":"Hilal Ahmad ,&nbsp;Zhang Yinghua ,&nbsp;Mehtab Alam ,&nbsp;Majid Khan ,&nbsp;Aboubakar Siddique","doi":"10.1016/j.pce.2026.104275","DOIUrl":"10.1016/j.pce.2026.104275","url":null,"abstract":"<div><div>Reservoir-induced landslides pose significant risks during dam operations, yet comprehensive pre-impoundment deformation baselines are rarely established. This study developed a systematic workflow using Small-Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) to map pre-filling slope instability along the Dasu Reservoir in northern Pakistan, a region experiencing significant climate change impacts. Ninety-two Sentinel-1 scenes (January 2022–December 2024) were processed through small-baseline interferogram networks and coherence-based weighting to generate millimetre-scale line-of-sight (LOS) displacement time series and mean velocity maps. Pixels exhibiting persistent motion &gt;2 mm/yr over ≥3 consecutive acquisitions were clustered into landslide polygons. These features were validated against high-resolution optical imagery and SRTM DEM data, then classified into low, moderate, or high hazard levels based on displacement magnitude and spatiotemporal trends. Results identified 27 active landslides, with nine high-hazard features concentrated in three critical zones along the reservoir margins. Zone-based crown-to-toe kinematic analysis revealed seasonal acceleration patterns and cumulative displacements ranging from −152 mm (uplift) to +97 mm (subsidence). High-hazard slopes (e.g., LS3: 2.4 km², LS4: 3.8 km², LS5: 5.5 km²) on the steep right bank exhibited progressive deformation exceeding 40 mm at crown locations, confirmed by UAV surveys. This work establishes a vital pre-impoundment deformation baseline for Geohazard Assessment. It enables detection of future reservoir-induced slope changes, data-driven early-warning thresholds, and prioritisation of critical stabilization zones by advancing Disaster Risk Management.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104275"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146022975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Status analysis of rock landslides in reservoir areas based on displacement monitoring and stability evaluation","authors":"Xiaolin Li ,&nbsp;Xinyi Chen","doi":"10.1016/j.pce.2026.104273","DOIUrl":"10.1016/j.pce.2026.104273","url":null,"abstract":"<div><div>At present, the method of rock mass landslide state analysis in reservoir area has uncertainty in parameter acquisition and model input, and the ability to capture step-type displacement mutation is limited. Therefore, a predictive model for the status of rock landslides in reservoir areas based on displacement monitoring and stability analysis is put forward. Particle Swarm Optimization combined with Support Vector Regression is applied to optimize displacement monitoring, aiming to improve the forecasting and early warning capabilities for landslide disasters. Experimental results show that the proposed model performs well in predicting different types of landslide deformation curves, especially achieving an accuracy of 95.12 % for smooth-type landslide deformations. In contrast, comparison models perform worse. The predictive model based on Cyclic Generative Adversarial Network achieves the lowest monitoring rate of 7.15 frames per second for step-type landslide deformations, while the predictive model based on Genetic Algorithm combined with Support Vector Regression reaches only a 78.54 % recall rate for smooth-type deformations. As the number of iterations increases, the proposed model maintains a prediction accuracy of over 90 %. Meanwhile, the overall prediction accuracy of the other three comparison models remains below 85 %. Overall, the proposed predictive model demonstrates strong mechanical mechanism characterization ability. It accurately predicts the status of rock landslides in reservoir areas and provides a novel monitoring method for concealed rock landslides.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104273"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Forecasting analysis of aftershock sequences in Morocco","authors":"Mohamed Hamdache ,&nbsp;Bogdan Enescu ,&nbsp;Jesús Henares ,&nbsp;Jose A. Peláez","doi":"10.1016/j.pce.2026.104312","DOIUrl":"10.1016/j.pce.2026.104312","url":null,"abstract":"<div><div>This study presents a short-term forecasting analysis for some seismic series in the Ibero-Maghrebian region, with a specific focus on recent occurrences in Morocco, including the Al Hoceima sequences occurring in 1994, 2004, and 2016, as well as the most recent event occurred in Al Haouz on September 2023. The method first determines the parameters of the Omori-Utsu and Gutenberg-Richter relationships for aftershocks, providing updatable probability estimates for earthquakes above a threshold magnitude, which makes it suitable for real-time applications. For the investigated sequences, a number of fairly short learning and forecasting periods (ranging from a few hours to a day) were employed. Gutenberg-Richter <em>b</em>-value estimations for the sequences under consideration were about equal to the standard value of 1.0. The aftershock rate decrease defined by the <em>p</em>-value of the Omori-Utsu relationship also was close to 1.0, which is commonly observed for crustal aftershock sequences. The achieved forecasting analysis indicates that, for all forecasting periods, the aftershock probability of the Al Hoceima 1994 sequence was lower than that of the other sequences. Using a learning period of 3 h after the mainshock, an aftershock of magnitude 5.0 or greater was forecasted for the Al Hoceima 1994 sequence with a probability of less than 10 %, for the Al Hoceima 2004 and 2016 sequences with a probability of approximately 20–30 %, and for the Al Haouz 2023 sequence with a probability of about 10 %, for a 3-h forecasting period, and exceeding 10 % for longer forecasting periods. The findings of this study are promising, and encourage further research in other areas of the Ibero-Maghrebian region for a quasi-real-time forecasting and monitoring system that could improve seismic risk management and evaluation.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104312"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146023084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geochemical insights into the depositional environment dynamics of Gondwana sediments in the Tajpur Basin, Bangladesh","authors":"Md Mahmudul Hasan Rakib ,&nbsp;Rahat Khan ,&nbsp;Arabe Khan ,&nbsp;Mehedi Hasan Ovi ,&nbsp;Pradip Kumar Biswas ,&nbsp;Md Saiful Islam ,&nbsp;Md Ali Akbar ,&nbsp;Javed Mallick ,&nbsp;Hoang Thi Hang ,&nbsp;Dhiman Kumer Roy","doi":"10.1016/j.pce.2026.104300","DOIUrl":"10.1016/j.pce.2026.104300","url":null,"abstract":"<div><div>Tajpur Basin Gondwana Sandstone (TBGS) was studied through Petrographic thin section, X-ray Fluorescence (XRF) and Instrumental Neutron Activation Analysis (INAA) to retrieve insight about the major trace elemental concentration and mineralogical assemblages, respectively. The quartz-rich sandstone was found to be enriched in SiO₂ (Avg. 71.95 wt%), Al₂O₃ (Avg. 11.78 wt%), Fe₂O₃ (Avg. 8.29 wt%) and K₂O (4.19 wt%) and depleted in response to other major elements. Among the trace elements Zr, Th, Cr, and Ti is relatively enriched in the samples. Among the trace elements Zr, Th, Cr, and Ti is relatively enriched in the samples. Almost every major and trace element possesses a negative correlation with SiO₂. The total REE contents range from 66.90 to 241.94 mg/kg, whereas the light REEs dominate the samples. Geochemical indices and discrimination diagrams were applied to infer the weathering, provenance, tectonic settings, climatic conditions and depositional environment of the TBGS. The moderate intensity of weathering can be inferred from the average Chemical Index of Alteration (CIA) value of 69.36. Furthermore, the discriminant function (F1–F2) diagram suggests felsic to intermediate provenance, whereas the bivariate plots of trace elements designate the provenance as felsic rock with minor input from recycled source. Tectonically, the samples were found to be derived from passive margin setting. The petrographic, major and trace element data, C-value (0.18–0.74), designate TBGS to be weathered and deposited in semi-humid to semi-arid climatic conditions. The depositional environment of the TBGS is indicative to the freshwater according to the proportional variability of Fe₂O₃ and MgO. Alongside, the sediments seemed to have been deposited in oxidizing environment, supported by V/Cr vs U/Th plot. The basin is marked to be formed due to the complex interaction of isostasy and tectonic stress, and episodic rifting, whereas the cyclic sedimentation events might have controls on sediment variability and accumulation pattern. Correlation with geophysical and geochronological data could have validated the present interpretation. This work represents a first step toward building a more holistic picture of the Gondwana basins of Bangladesh.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104300"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146023086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Distribution, source apportionment, and risk assessment of polycyclic aromatic hydrocarbons in Chinese waters' sediments","authors":"Jiaxin Kuang ,&nbsp;Yan Li ,&nbsp;Qingkun Wang ,&nbsp;Ye Li ,&nbsp;Ke Liu ,&nbsp;Long Chen ,&nbsp;Hang Su","doi":"10.1016/j.pce.2025.104247","DOIUrl":"10.1016/j.pce.2025.104247","url":null,"abstract":"<div><div>To investigate the nationwide pollution status of PAHs in sediments and provide data support and clear direction for water pollution control across China, this study systematically integrated publicly available concentration data of PAHs in surface sediments from Chinese water bodies between 1993 and 2023. It reveals their spatiotemporal distribution patterns, source characteristics, and ecological risks. Results indicate that PAHs pollution exhibits a temporal trend of “initial increase followed by decline,” peaking around 2006. Spatially, it shows a pattern of “higher concentrations in the east (598.06 ng/g) than in the west (580.97 ng/g), and higher in the north (724.39 ng/g) than in the south (485.98 ng/g).” Source apportionment using two receptor models yielded largely consistent results. Comparison indicated APCS-MLR provided more reliable outcomes, attributing contribution rates as follows: biomass combustion (48.26 %), coal combustion (25.49 %), transportation emissions (13.77 %), and petroleum emissions (12.48 %). Ecological risk assessment using both the risk quotient (RQ) method and the toxicity equivalent (TE) method yielded highly consistent results, indicating the highest ecological risk for PAHs in the Northeast region and the lowest in the Central region. The calculated national overall RQ_(NCs) for sediment PAHs was 263.128, and RQ_(MPCs) was 2.631, indicating a medium-to-high risk level.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104247"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840040","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Remote sensing and multi-method statistical analysis of snow cover variability and hydrological responses in the Upper Indus Basin","authors":"Nasir Abbas ,&nbsp;Shah Fahd ,&nbsp;Ninglian Wang ,&nbsp;Muhammad Zubair ,&nbsp;Adeel Ahmad Nadeem ,&nbsp;Zeeshan Zafar","doi":"10.1016/j.pce.2025.104263","DOIUrl":"10.1016/j.pce.2025.104263","url":null,"abstract":"<div><div>Climate-driven changes in the cryosphere pose a significant threat to Pakistan's fragile water-energy-food nexus, which depends heavily on the snow-fed waters of the Upper Indus Basin (UIB) for agriculture and power generation. This study analyzed snow cover variability across the UIB from 2009 to 2024 using 8-day snow cover data from the MOD10A2 product. To identify long-term climate trends, 35 years of climate data were examined using the Mann–Kendall trend test. Additionally, available runoff records were evaluated to assess the hydrological behavior of the Gilgit-Baltistan area within the UIB. Statistical methods, including Pearson correlation, Structural Equation Modeling, and wavelet coherence, were used to analyze relationships among snow cover extent, temperature, precipitation, and runoff. Results show seasonal fluctuations in maximum snow cover area (SCA), ranging from 34 % to 83 %. While the overall SCA in Gilgit-Baltistan remained stable (Sen's slope≈0.035 % year⁻¹), rising temperatures, along with increased precipitation and runoff volumes, showed decreasing trends with Sen's slope values of −13.12 mm year⁻¹ and -11.62 mm year⁻¹, respectively. Simultaneously, summer precipitation and temperatures in the region showed significant upward trends. Results indicate that consistent snow cover, combined with decreasing precipitation and runoff, poses a threat to hydropower production, irrigation resources, and downstream water accessibility. Early melting may reduce summer energy production, exacerbate agricultural water stress, and elevate the dangers of both flooding and drought. The effects highlight the vulnerability of the water–energy–food nexus and the urgent need for adaptive management to safeguard power, food security, and climate resilience in UIB.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104263"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840035","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantification of long-term spatio-temporal CO2 dynamics in Indian terrestrial ecosystems","authors":"Ramakant Tiwari ,&nbsp;Dileep Kumar Gupta ,&nbsp;Prashant Kumar Chauhan ,&nbsp;Vivek Singh ,&nbsp;Abhay Kumar Singh","doi":"10.1016/j.pce.2025.104259","DOIUrl":"10.1016/j.pce.2025.104259","url":null,"abstract":"<div><div>Carbon dioxide (CO₂) is a significant contributor to atmospheric climate change, making long-term monitoring crucial for developing effective regional/global mitigation policies. A multidisciplinary approach is adopted in this study, utilizing a combination of AIRS satellite CO₂ data, vegetation (NDVI) data, and fire data to obtain information on atmospheric and ecological dynamics in India. The long-term spatial and time series trends of CO₂ during 2002–2017 provides useful information about regional climate trends. The observations in this study have examined the monthly, annual, and seasonal dynamics of CO₂ across Indian terrestrial ecosystems. The results indicate that there has been a steady increase in CO₂ concentrations at a mean rate of 2.1 ppm annually in India. Seasonal analysis reveals that there are higher concentrations of CO₂ in winter, primarily due to reduced atmospheric mixing, and lower concentrations during the monsoon, resulting from increases carbon uptake by vegetation. Correlation analysis between CO₂ and NDVI on a monthly basis showed strong vegetation effects in April, May, and August and weak correlations in other months due to anthropogenic influences. An effective approach, namely the Standard Deviation Ellipse (SDE), has been used in this study to elucidate the spatial behaviour and directionality stability of CO₂ patterns of concentration that would not have been identified in traditional pixel-based analysis. The results of the SDE indicate that the geographical location of the CO₂ concentration was stationary throughout time, implying that there still existed emission hot spots in central India. This study primarily presents a dynamic and holistic view of the Indian carbon landscape by integrating a solid SDE model with the use of various data source during 2002–2017. The outcomes of the present study provide useful guidance on mitigation measures and highlight the need to resolve long-term sources of emissions.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104259"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Differentiated grouting reinforcement for mines based on material rheology and coal–rock bonding characteristics","authors":"Bowen Fan ,&nbsp;Ping Chang ,&nbsp;Zhijun Wan ,&nbsp;Yuan Zhang ,&nbsp;Ruiqiang Liu ,&nbsp;Bin Du","doi":"10.1016/j.pce.2025.104249","DOIUrl":"10.1016/j.pce.2025.104249","url":null,"abstract":"<div><div>Roof breakage and weak coal seams frequently trigger rib spalling and roof collapse at longwall faces, posing major risks to mine safety. Grouting reinforcement is an effective measure to stabilise fractured coal–rock masses, yet its performance strongly depends on selecting materials that balance structural requirements, construction safety and economic feasibility. This study presents a combined laboratory–field evaluation of four commonly used grouting materials—ordinary Portland cement (PISC), single-component inorganic material (JIRM), two-component inorganic material (TIRM) and two-component organic material (SPM). Their rheological behaviour, reinforcement effects on coal–rock composites and failure characteristics were examined, followed by full-scale verification at an operating longwall face. The results show clear differentiation in material performance: TIRM develops the highest compressive strength and is suited to high-stress reinforcement; SPM exhibits superior macroscopic ductile failure characteristics, enabling rapid stabilisation in emergency conditions; JIRM provides excellent flowability and cost advantages for large-scale pre-grouting; whereas PISC shows comparatively limited applicability. On this basis, a differentiated material selection strategy was formulated and successfully applied in the field, resulting in reduced material consumption, lower operational cost and markedly improved face stability. The study offers a practical framework for matching grouting materials to geological and operational demands, providing guidance for mines with similar working conditions.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104249"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning approach for enhanced estimation of phosphorus adsorption isotherm parameters","authors":"Xiaogang Li ,&nbsp;Zelin Hu ,&nbsp;Yilian Li ,&nbsp;Danqing Liu ,&nbsp;Xiaoying Yang","doi":"10.1016/j.pce.2025.104242","DOIUrl":"10.1016/j.pce.2025.104242","url":null,"abstract":"<div><div>Adsorption is one critical mechanism affecting phosphorus migration across different environmental compartments. However, traditional laboratory-based parameter estimation for phosphorus adsorption isotherm models is laborious and results are limited to specific experimental conditions, restricting their broader applicability for assessing phosphorus transport dynamics in the real-world heterogeneous conditions. This study explored using machine learning to decipher the complex, nonlinear dynamics between the parameters of phosphorus adsorption isotherm models and common soil properties utilizing published experiment results. Our study results have confirmed the feasibility of such approach, with the best-performing models achieving R² values of 0.93 and 0.94 for the Q_max and LogK_L parameters, respectively, of the Langmuir model, and R² values of 0.99 for both the K_F and b parameters of the Freundlich model. For Q_max (Langmuir) and K_F (Freundlich), both indicating soil adsorption capacity, clay content was identified as the most significant influencing soil property. For K_L (Langmuir), indicating the strength of phosphorus adsorption, clay content (positive effect) and pH (negative effect) were identified as the two primary influencing soil properties. For the Freundlich b parameter, indicating favorability of phosphorus adsorption, pH and organic matter content were determined to be critical factors, both with positive effects. While focused on phosphorus, the newly developed framework for estimating adsorption model parameters possesses significant transferability to other pollutants. The developed models and adsorption parameter estimates are practically valuable for mapping regional soil adsorption capacity, calibrating and refining pollutant transport model parameters, and thereby supporting the formulation of effective pollution control and remediation strategies.</div></div>","PeriodicalId":54616,"journal":{"name":"Physics and Chemistry of the Earth","volume":"142 ","pages":"Article 104242"},"PeriodicalIF":4.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}