Remote Sensing Applications-Society and Environment最新文献

{"title":"Assessment of spectral indices and water color combinations for detecting algal blooms in coastal subtropical shallow lakes","authors":"Mariê Mello Cabezudo ,&nbsp;Matheus Henrique Tavares ,&nbsp;Ng Haig They ,&nbsp;David da Motta Marques","doi":"10.1016/j.rsase.2025.101678","DOIUrl":"10.1016/j.rsase.2025.101678","url":null,"abstract":"<div><div>Algae and cyanobacteria blooms are a growing concern for the quality of aquatic ecosystems, but logistics and cost constraints often limit their monitoring. The use of spectral indices through remote sensing can help detect blooms in places that are difficult to access or have limited available data. However, differences in water optical properties and sensor configuration may affect the accuracy of these indices in inland waters. Here, we assessed whether multiple spectral indices and one colour algorithm based on the International Commission of Illumination colour space (CIE) could increase the accuracy of bloom detection in a shallow coastal lakes system using different satellites. We first calibrated thresholds for the indices against visually detectable blooms and tested the agreement of various algorithm combinations. We found the threshold adjustment did not improve bloom detection for Landsat 8/9 and Sentinel-2, but it is essential for Landsat 5. Bloom areas obtained with CIE combined with the Adjusted Floating Algae Index (AFAI), for the Landsat series, and the Normalized Difference Chlorophyll Index (NDCI), for Sentinel-2, resulted in the best overall accuracy. The CIE algorithm helped reduce false positives in non-blooming lakes. Our results show that using single algorithms with CIE can be applied to retrieve accurate bloom occurrence and areas with multiple sensors; however, these must be tailored according to local characteristics. The methods validated here can be applied to understand the long-term variability of bloom events in lake systems located in regions that are inaccessible or that suffer from a lack of data.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101678"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144780143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cloud removal with compact diffusion models: A residual block-based approach","authors":"Leandro Henrique Furtado Pinto Silva ,&nbsp;João Fernando Mari ,&nbsp;Mauricio Cunha Escarpinati ,&nbsp;André Ricardo Backes","doi":"10.1016/j.rsase.2025.101680","DOIUrl":"10.1016/j.rsase.2025.101680","url":null,"abstract":"<div><div>Satellites are powerful tools for remote sensing, as they enable the imaging of large areas with high quality. However, satellites can be prone to artifacts such as clouds, which can negatively influence the analysis of these images. Thus, researchers have widely investigated cloud removal techniques to mitigate these artifacts, leveraging the rise of generative artificial intelligence methods. These techniques, although powerful, require a high computational cost, which limits their use in real-time applications, embedded devices, and environmental monitoring systems, where computational resources are often limited. Therefore, this work presents an approach based on compact latent diffusion, where the denoising model uses attention channels and residual block operations. In addition, we evaluated different training loss functions, which help the model perform cloud removal across various land cover types. Considering a resource-constrained approach, we investigated different experimental configurations using Pareto Front to optimize the most promising experiments. Our results demonstrate a balance between reconstruction quality and computational cost compared to baseline. Our approaches have between 48% and 82% fewer parameters while presenting competitive results for similarity, noise, and perceptual metrics.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101680"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Applying the science impact framework to understand real-world applications and impacts of ICESat and ICESat-2 data on decision-making","authors":"Siobhan L. Light ,&nbsp;Molly E. Brown ,&nbsp;Aimee R. Neeley ,&nbsp;Thomas A. Neumann","doi":"10.1016/j.rsase.2025.101669","DOIUrl":"10.1016/j.rsase.2025.101669","url":null,"abstract":"<div><div>Assessing the societal impact of satellite remote sensing datasets is essential to understanding how these data influence decision-making and to identifying opportunities for further engagement. However, measuring such impacts remains challenging for missions serving diverse stakeholder communities. In this study, we evaluate the broader impact of NASA's Ice, Cloud, and land Elevation Satellite (ICESat) and its successor mission, ICESat-2 by adapting the scientific impact framework (SIF), originally developed to assess public health research, into an Earth science-specific version (e-SIF). This framework captures data dissemination, community awareness, data-driven actions, measurable changes, and future influence, moving beyond traditional academic metrics to assess the missions' reach and effectiveness. Our findings reveal extensive global usage of ICESat and ICESat-2 data, with applications including, but not limited to, shallow water bathymetry, climate mitigation strategies, and forest management. By comparing the prevalence of topical areas in academic literature to real-world applications, we found that althoughthe cryosphere is the most frequently studied domain, differences between research focus and practical use highlight potential areas where further research could better support stakeholders. We found that ICESat and ICESat-2 data are widely employed by national and international governmental and non-governmental organizations but found only limited use by private sector and local governments. We recommend that the ICESat-2 Applications Team expand outreach efforts to these sectors to enhance dissemination of mission data. Furthermore, numerous ICESat-2 applications benefit from long-term data continuity, reinforcing the need for a successor mission. This study demonstrates the feasibility to use e-SIF to evaluate the impact of Earth science missions.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101669"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144738670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigation of evolutionary fuzzy MPCM machine learning and probabilistic SVM models for Butea monosperma species mapping","authors":"Payel Mani ,&nbsp;Dipanwita Dutta ,&nbsp;Anil Kumar","doi":"10.1016/j.rsase.2025.101667","DOIUrl":"10.1016/j.rsase.2025.101667","url":null,"abstract":"<div><div>Accurate identification of plant species at an optimal level of precision remains a major challenge in ecological observations when using conventional classification methods. This study explores the potentiality of multi-temporal datasets with machine learning classifiers for the identification and distribution of Butea monosperma tree species, a native floral species grown in many countries of South and Southeast Asia. For identifying optimum combinations of temporal images the Euclidian distance-based separability analysis was employed on the multi-temporal GCI, MSAVI2 indices database (24 toatal temporal dates). This study uses the fuzzy Modified Possibilistic <em>c-</em>Mean (MPCM) classification method combined with the green chlorophyll (GCl), MSAVI2 temporal index to handle the complexity and uncertainty inherent in the phenological data. Owing to its lesser variance on the testing target species over the other classes, the 21 optimum temporal combinations of GCl images were chosen as a benchmark for comparison of the output with the Probabilistic Support Vector Machine (PSVM) with Radial Basis Function (RBF) kernel approach machine learning classifier which is well known for its ability to handle probabilistic information and high-dimensional data. In this study, a diverse dataset of tree species phenological observations has been employed to evaluate the performance of both classifiers. Key metrices such as overall accuracy and F1-score were utilized for the comparison of different models. The MPCM classifier achieved notable performance, with 92 % overall accuracy and an F1-score of 0.93 when utilizing the 21-temporal GCI database. In contrast, a single-date output resulted in only 65% overall accuracy and an F1-score of 0.74. When compared to PSVM model, which exhibits an F-score of 0.88 and an overall accuracy of 82 %, the utilization of MPCM with combined 21 temporal GCI indices demonstrated superior classification performance. Additionally, this research provides insights into how various evolutionary strategies and algorithms can enhance the classifiers’ adaptability to changing data distributions.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101667"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144750405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decoding urban complexity: Deep learning-based terrain-specific building segmentation for Indian cities","authors":"Akshit Koduru ,&nbsp;Reedhi Shukla","doi":"10.1016/j.rsase.2025.101673","DOIUrl":"10.1016/j.rsase.2025.101673","url":null,"abstract":"<div><div>Accurate building segmentation from satellite imagery is essential for urban planning, disaster management, and environmental monitoring. This paper presents a novel approach utilizing the UNET architecture for deep learning-based building segmentation, focusing on diverse terrains in Indian cities. Indian cities are uniquely complex in their urban complexity because of a highly densely packed urban landscape and patterns of buildings, thus making segmentation a challenging task. Our method includes meticulously performed data preprocessing and exhaustive validation to achieve high accuracy and adaptability in our trained terrain-based model. Very high-resolution satellite imagery with a 0.5-m spatial resolution was utilized for model training. Specialized models were developed for different terrain types—urban, coastal, and hilly—resulting in significant improvements in segmentation performance compared to generalist models. We reduce human effort and increase efficiency as the proposed system automates segmentation. Such research will, therefore, scale the solution very well in building segmentation. Its application will be practical to aspects of urban planning and disaster response while developing the smart city, and further work will be oriented towards expanding the dataset and generalizing and further developing the model.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101673"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144841965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"High precision building detection from satellite imagery with a novel SDBN-HCWO method","authors":"Md Helal Miah ,&nbsp;Shuanggen Jin ,&nbsp;Mayin Uddin Jubaid ,&nbsp;Md Altab Hossin","doi":"10.1016/j.rsase.2025.101698","DOIUrl":"10.1016/j.rsase.2025.101698","url":null,"abstract":"<div><div>Detecting buildings from satellite imagery presents challenges related to computational efficiency, model adaptation, and occlusion. This paper introduces a novel method called the Secant Deep Belief Network-Hyperbolic Cosine Whale Optimization (SDBN-HCWO) for building detection in satellite images. The research utilizes the SDBN-HCWO method to enhance building detection accuracy in satellite images. It addresses challenges like computational efficiency, occlusion, and dataset adaptation. The method integrates a multi-layer structure, including Hyperbolic Cosine Prey Encircling for edge identification, Shrinking Encircle for optimal edge linking, and Secant Object Detection for accurate identification. Additionally, a Densely Connected Convolutional Network (DCCN) and Depth-wise Separable Convolution (DSC) optimize feature extraction, reducing computational costs. The model is evaluated on both quantitative and qualitative metrics, ensuring high accuracy and low false positive rates. The research findings demonstrate that the SDBN-HCWO method significantly improves building detection accuracy in satellite imagery. It enhances detection efficiency by integrating Discrete Latent Deep Reinforcement Learning and a bubble-net mechanism, reducing false positives by 58 %. The model outperforms conventional approaches, achieving an 18 % increase in PSNR, 34 % rise in CA, and 19 % reduction in training time. High AP scores (90.40 %–92.67 %) confirm its reliability, though challenges persist in medium-damage detection. It surpasses YOLOv3, YOLOv4, and Faster R-CNN in accuracy and efficiency. This research significantly advances building detection in satellite imagery, facilitating more accurate urban planning, disaster response, and environmental monitoring.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101698"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144903184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The use of Unmanned Aerial Vehicle (UAV) remotely sensed data and biophysical variables to predict maize Above-Ground Biomass (AGB) in small-scale farming systems","authors":"Celuxolo Michal Dlamini ,&nbsp;John Odindi ,&nbsp;Trylee Nyasha Matongera ,&nbsp;Onisimo Mutanga","doi":"10.1016/j.rsase.2025.101706","DOIUrl":"10.1016/j.rsase.2025.101706","url":null,"abstract":"<div><div>Considering the current and projected increase in human population, approaches to optimize crop productivity to meet the rising demand are paramount. Timely and accurate maize Above Ground Biomass (AGB) measurements allow for development of models that can precisely predict yield prior to harvesting, useful for managing cropping systems and food production. The development of Unmanned Aerial Vehicles (UAVs) as a new generation of robust remote sensing platforms, mounted with high-resolution sensors has allowed timely and accurate prediction of maize AGB in pursuit of sustaining food security. Hence, this study aimed to predict maize crop AGB in small-scale farming systems using UAV-remotely sensed data and landscape biophysical variables. The DJI Matrice 300 UAV mounted with a MicaSense multispectral camera was used to acquire high-resolution images at four phenological stages that covered the vegetative (V8 &amp;V12) and reproductive stages (R2 &amp; R5). Furthermore, in-situ plant biophysical measurements and landscape variables were acquired and combined with UAV-remotely sensed derived vegetation indices to model maize AGB using a Deep Neural Network (DNN) model. The optimal model for predicting maize AGB was achieved by combining optimal vegetation indices, with Leaf Area Index (LAI), leaf chlorophyll content, slope, aspect, and soil moisture across all phenological stages. Results showed that the V12 phenological stage yielded a better overall prediction accuracy (R² = 0.75, RMSE = 0.07 kg/m², rRMSE = 6.12 %) than the V8 (R² = 0.71 RMSE = 0.10 kg/m², rRMSE = 8.02 %), R2 (R² = 0.73 RMSE = 0.09 kg/m², rRMSE = 7.86 %), and R5 (R² = 0.70 RMSE = 0.10 kg/m², rRMSE = 8.51 %) growth phases. The study concludes that the V12 and R2 phenological stages are optimum for estimating maize AGB. This study contributes to a better understanding of maize crop monitoring efforts for improved production and food security.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101706"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144920067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"When ratoon longevity matters: Spatial and temporal analysis of sugarcane plant population patterns using features extracted from UAV images","authors":"Leonardo Felipe Maldaner ,&nbsp;José Paulo Molin ,&nbsp;Carlos Tadeu dos Santos Dias ,&nbsp;Eudocio Rafael Otavio da Silva","doi":"10.1016/j.rsase.2025.101703","DOIUrl":"10.1016/j.rsase.2025.101703","url":null,"abstract":"<div><div>Monitoring the spatial and temporal dynamics of plant populations in sugarcane fields is essential for site-specific management and for sustaining high yields over time. However, to the phenological characteristics of sugarcane make plant detection and mapping particularly challenging. This study aimed to analyze spatial and temporal changes in plant populations within sugarcane ratoon fields using unmanned aerial vehicle (UAV) imagery. The goal was to improve management in commercial plantations and to map susceptibility to plant reduction over time, based on features extracted from UAV data: terrain slope, sugarcane row (path), path angle, gap length, and plant population. UAV imagery was collected over two successive seasons, 2019 and 2020. RGB mosaics were split into tiles (40,000 square pixels) and then into 50 × 50-pixel windows, subsequently used for L∗a∗b∗-based K-means segmentation, identifying sugarcane clumps via centroid extraction and mask filtering, as well as gaps along the rows. Nineteen plots (representing diverse slopes and paths) were analyzed, comparing image-derived and manual plant counts. To assess susceptibility to plant reduction over time, principal component analysis (PCA) and cluster analysis were applied for classification and mapping. The K-means segmentation achieved 91.00 % accuracy in detecting sugarcane plants. Overall, the plant population decreased by 16.00 %, with a 0.70 m increase in gap length over the study period. Regions with terrain slopes of 5.00–8.00 % and above 8.00 % with curved paths had fewer plants compared to flatter areas. Higher terrain slopes correlated with a greater probability of plant population reduction over time. The susceptibility patterns were mapped, providing insights to support management decisions, including the identification of areas requiring replanting and planning for field renovation.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101703"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144920068","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of nature-based solutions (NBS) on urban surface temperatures and land cover changes using remote sensing and machine learning","authors":"Paloma Carollo Toscan ,&nbsp;Kijin Seong ,&nbsp;Junfeng Jiao ,&nbsp;Carlos Alexandre Lopes Rodrigues Ribeiro ,&nbsp;Francisco André Costa Carvalho ,&nbsp;Marcos L.S. Oliveira ,&nbsp;Eduardo B. Pereira","doi":"10.1016/j.rsase.2025.101721","DOIUrl":"10.1016/j.rsase.2025.101721","url":null,"abstract":"<div><div>Urban areas are increasingly vulnerable to climate change, with rising urban heat driven by the replacement of vegetation with impervious surfaces. Nature-Based Solutions (NBS) provide promising strategies to mitigate urban heat while promoting environmental sustainability. This study analyzes the spatiotemporal dynamics of Land Cover (LC) and Land Surface Temperature (LST) in Guimarães, Portugal, from 2013 to 2023, and forecasts scenarios for 2028 using advanced machine learning techniques.</div><div>Key methodologies included supervised LC classification via Random Forest (RF), LC prediction using the MOLUSCE plugin, and LST prediction using ensemble models such as XGBoost, Bagging, and AdaBoost, with XGBoost demonstrating the highest performance (R² = 0.9543). The results highlight significant transitions from barren and built-up areas to vegetation, reflecting localized environmental recovery. NBS interventions, such as green roofs and urban gardens, achieved measurable cooling effects, reducing temperatures by up to 2.49 °C in their surroundings. Projections for 2028 indicate a slight decline in vegetation (−0.35 %), underscoring the urgent need for strengthened conservation efforts. Identified thermal hotspots, particularly in urban and industrial zones, further emphasize the importance of expanding NBS strategies.</div><div>This research advances the integration of remote sensing and machine learning for urban climate analysis, offering practical insights for urban planning and climate mitigation policies. Future studies should incorporate additional variables to refine prediction models, assess large-scale impacts of distributed NBS, and leverage high-resolution data for broader applications. These findings provide a scalable framework for sustainable urban development worldwide.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101721"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145019184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A machine learning algorithm to retrieve the red peak of phytoplankton absorption spectra from ocean-colour remote sensing","authors":"Mohammad Ashphaq,&nbsp;Shovonlal Roy","doi":"10.1016/j.rsase.2025.101702","DOIUrl":"10.1016/j.rsase.2025.101702","url":null,"abstract":"<div><div>Light absorption by microscopic phytoplankton in marine ecosystems is a crucial process underpinning biological production and global biogeochemical cycles. Accurate estimation of phytoplankton absorption coefficients, an inherent optical property of ocean water, can improve remote sensing applications including spectral photosynthesis models and assessments of ocean health, biodiversity, and climate change impacts. However, considerable uncertainty exists in current satellite retrievals of phytoplankton absorption coefficients, particularly for <em>ɑ</em>_<em>ph</em>(676) - the phytoplankton absorption peak at red wavelengths near 676 nm - which is an input to several novel and advanced satellite algorithms. This uncertainty hinders operational use of algorithms for assessing phytoplankton physiology, size structure and oceanic carbon pools from space. We aimed to improve satellite-based estimation of <em>ɑ</em>_<em>ph</em> (676) using advanced machine learning (ML) techniques. We compiled a comprehensive <em>in situ</em> dataset (n = 1576) of <em>ɑ</em>_<em>ph</em>(676) from published databases and matched with remote-sensing reflectance <em>Rrs</em> at six wavelengths (412, 443, 490, 510, 560, and 665 nm) from the Ocean Colour Climate Change Initiative. We extensively evaluated multiple base ML algorithms: Random Forest (RF), Gradient Boosting Machines, and Linear Regression; and implemented ensemble ML models: RF with Grid Search Cross-Validation, eXtreme Gradient Boosting Ensembled Model, Ensemble Forecast, Stacked Voting, Optimised Ensemble and Meta Stacking, integrating the base models through cross-validated hyperparameter tuning. Meta Stacking outperformed individual ML models in predictive accuracy across temporal resolutions, showing best results with daily composites. Our study addresses key limitations of previous models, including small training datasets, inconsistent performances, and lack of ensemble comparisons. We present a robust, extensively trained and validated ensemble ML model that significantly improves <em>ɑ</em>_<em>ph</em>(676) estimation and opens the possibility of routinely using in ocean colour remote sensing.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"39 ","pages":"Article 101702"},"PeriodicalIF":4.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144988697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}