Junru Yu , Longcai Zhao , Yanfu Liu , Qingrui Chang , Na Wang
{"title":"Automatic crop type mapping based on crop-wise indicative features","authors":"Junru Yu , Longcai Zhao , Yanfu Liu , Qingrui Chang , Na Wang","doi":"10.1016/j.jag.2025.104554","DOIUrl":"10.1016/j.jag.2025.104554","url":null,"abstract":"<div><div>Mapping crop types with remote sensing imagery is crucial for timely acquisition of crop planting information. However, large-scale crop type mapping is often hindered by the absence of ground truth data and spatiotemporal heterogeneity of available cloud-free optical images. In this study, the conception of daily crop-wise indicative features (DCIF) was proposed for automatic crop mapping to address this issue. Crop-wise indicative feature (CIF) was defined as a comprehensive feature that can distinctly differentiates the target crop and all other classes, with the CIF value showing a polarization trend with target crop close to 1 and others close to 0. Logistic Regression (LR) model was utilized to learn CIF of target crop on each single date, referred to as CIF extractor. The time series analysis method (i.e., seasonal trend decomposition) and imputation method was then utilized for the discrete time series of CIF extractors, which were separately trained based on training samples on each day. This process yielded the DCIF extractor that can capture the unique feature pattern on any given day during the entire growing period. The comprehensive features produced by DCIF extractor later served as the input of Otsu algorithm to automatic distinguish target crop from other crops. Our results in Heilongjiang, China, Iowa, USA, and Bas-Rhin, France show that the DCIF extractor effectively overcomes abovementioned challenges in crop mapping by applying time series analysis method at model-level. The overall accuracy exceeded 90 % in Heilongjiang and Iowa for both 2021 and 2022. In Bas-Rhin, where fragmented land parcels and uneven crop distribution are prevalent, the overall accuracy surpassed 87 %. The proposed method provides a scientific basis for crop acreage estimation and field management, contributing to more informed agricultural practices.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104554"},"PeriodicalIF":7.6,"publicationDate":"2025-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143877391","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}
Hamid Kamangir , Brent S. Sams , Nick Dokoozlian , Luis Sanchez , J. Mason Earles
{"title":"Predicting crop yield lows through the highs via binned deep imbalanced regression: A case study on vineyards","authors":"Hamid Kamangir , Brent S. Sams , Nick Dokoozlian , Luis Sanchez , J. Mason Earles","doi":"10.1016/j.jag.2025.104536","DOIUrl":"10.1016/j.jag.2025.104536","url":null,"abstract":"<div><div>Crop yield estimation is vital for agricultural management but often struggles with predicting extreme values that can significantly impact operations and markets. Traditional models face challenges with these extremes, leading to biased and inaccurate predictions. To address this challenge, our study introduces two innovative strategies. First, we propose a cost-sensitive loss function, ExtremeLoss, designed to better capture and represent less frequent yield values by giving greater importance to extreme cases during training. Second, we develop a conditional deep learning model that enhances feature representation by conditioning on a binned yield observation map. This approach encourages smoother and more coherent input feature maps across different segments of the yield value range by leveraging similarities within and across yield bins, ultimately improving the model’s ability to generalize and distinguish between subtle variations in yield. This approach creates ”yield zone maps,” grouping yields into classes (e.g., low extreme, common, high extreme) to improve the identification of yield variability, which can be removed during inference. Our model was tested on a comprehensive grape yield dataset from 2016 to 2019, covering 2,200 hectares and 42 blocks of eight cultivars. We compared its performance against advanced techniques such as Focal-R loss, label distribution smoothing, dense weighting, and class-balanced methods under two validation scenarios: block-hold-out (BHO) and year-block-hold-out (YBHO). Our approach outperforms existing models in R-squared <span><math><mrow><mo>(</mo><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>)</mo></mrow></math></span>, Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Notably, it reduces MAE by +2.98 and +14.45 (t/ha) for low and high extremes in the BHO scenario and by +7.18 and +11.05 (t/ha) in the YBHO scenario. It also significantly decreases MAPE by +19.09% and +23.94% in the BHO scenario and by +33.76% and +19.61% in the YBHO scenario. Our model shows a marked improvement in capturing spatial variability and significantly advances spatio-temporal yield estimation, particularly for extreme values in complex agricultural settings like vineyards.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104536"},"PeriodicalIF":7.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869211","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":"Estimating characteristics of planted forests’ relative yield index using low pulse density LiDAR and satellite remote sensing","authors":"Asahi Hashimoto , Shodai Inokoshi , Chen-Wei Chiu , Yuichi Onda , Takashi Gomi , Yoshimi Uchiyama","doi":"10.1016/j.jag.2025.104558","DOIUrl":"10.1016/j.jag.2025.104558","url":null,"abstract":"<div><div>The increased demand for forest management, particularly thinning, in Japan is a direct consequence of aging planted forests. However, forest inventories (FI) in Japan lack crucial details regarding the developmental stage or ecological status of forests, often only providing tree species, age, and owner information. The relative yield index (<em>R<sub>y</sub></em>) is a forest density index widely used in the forestry industry in Japan. It can be combined with tree height data to calculate tree density, diameter at breast height, timber volume, and basal area at breast height as the stand scale. Although <em>R<sub>y</sub></em> is a valuable indicator for forest management, no studies have been reported on its estimation over a large spatial scale. Therefore, in this study, we aimed to estimate the <em>R<sub>y</sub></em> of planted Japanese cedar and cypress forests at the stand scale over a large area by combining satellite imagery and airborne light detection and ranging (LiDAR) data, which offer excellent vertical resolution.</div><div>Data on surface temperature, which is sensitive to differences in forest density, was obtained from Landsat8 satellite imagery. Considering that surface temperature is highly dependent on topography, we developed a topography-aware normalized surface temperature index (<em>Ω<sub>ST</sub></em>) using surface temperature data and a digital elevation model. The leaf area index (LAI), which was positively correlated with <em>R<sub>y</sub></em>, was estimated from the enhanced vegetation index obtained from Landsat. A normalized LAI (<em>Ω<sub>LAI</sub></em>) was developed to address differences in LAI attributable to tree height. The <em>R<sub>y</sub></em> estimation index (<em>R<sub>y_estimated</sub></em>) calculated using <em>Ω<sub>ST</sub></em> and <em>Ω<sub>LAI</sub></em> was correlated with the <em>R<sub>y</sub></em> estimated from LiDAR data (correlation coefficient; <em>r</em> = 0.61–0.65), confirming its high accuracy (root mean square error; RMSE = 0.07–0.11). By applying this method to a 3,650 km<sup>2</sup> area of planted Japanese cedar and cypress forests in the Kanto region of Japan, large-scale and detailed information on various forest characteristics was obtained. This method derives tree height data from LiDAR and extracts forest density information from satellite imagery. The combination of LiDAR data and satellite imagery potentially enhances the accuracy of forest-based estimates, reduces data acquisition costs, and improves the efficiency of creating and updating FIs.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104558"},"PeriodicalIF":7.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869210","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}
Maxence Dodin , Florent Levavasseur , Antoine Savoie , Lucie Martin , Emmanuelle Vaudour
{"title":"Farm-scale mapping of compost and digestate spreadings from Sentinel-2 and Sentinel-1","authors":"Maxence Dodin , Florent Levavasseur , Antoine Savoie , Lucie Martin , Emmanuelle Vaudour","doi":"10.1016/j.jag.2025.104555","DOIUrl":"10.1016/j.jag.2025.104555","url":null,"abstract":"<div><div>According to few recent studies, exogenous organic matters (EOM) can be detectable on either emerging vegetation or bare soil using optical and radar remote sensing techniques. Nevertheless, these image processing approaches considered one single EOM, one season and/or year only and were limited to one surface condition prior to spreading. So far no method addressed the simultaneously tracking of both liquid and solid EOM applications using satellite imagery, for different years, seasons and surface conditions. Relying on Support Vector Machine (SVM) classifier, this study aimed to track applications of both composted manure and liquid digestate over three seasons of successive years (late winter of 2019; spring of 2020 and 2021) in agricultural fields on a farm scale with distinct surface conditions (grassland, winter crop, bare soil). Within-field reference areas were delineated based on both the observed amendment practices, crops and soil map and randomly selected with replacement to train/ validate SVM with several iterations. Various feature sets composed of bands, signals and specific spectral indices from either Sentinel-2 and/or Sentinel-1 data served to compute SVM in a bootstrapping approach in order to produce a series of map results, to assess the final mode class and the uncertainty of map results. Classification performance was higher for pre- and post-application image pairs compared to post-application images alone and slightly improved when adding Sentinel-1 data. While the areal percentage of the highest uncertainty class covered less of 10% of the mapped area regardless of the year, the best models showed accuracies higher than 93% in 2020 and 2021. In 2019, the overall accuracy did not reach more than 79%, probably due to rainfall events and considerable time lags between the image pairs. This study underscores, not only the potential of Sentinel-2 and 1 for monitoring EOM applications, but also the requirement of better understanding the spectral behaviour of the EOM spreadings, in line with a thorough characterization of the sequence of crop technical management.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104555"},"PeriodicalIF":7.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869209","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}
Heng Tang , Xiaoping Rui , Jiarui Li , Ninglei Ouyang , Yiheng Xie , Xiaodie Liu , Yiming Bi
{"title":"SLIP-flood: Soft-combination of Swin Transformer and Lightweight Language-Image Pre-training for Flood Images Classification","authors":"Heng Tang , Xiaoping Rui , Jiarui Li , Ninglei Ouyang , Yiheng Xie , Xiaodie Liu , Yiming Bi","doi":"10.1016/j.jag.2025.104543","DOIUrl":"10.1016/j.jag.2025.104543","url":null,"abstract":"<div><div>Flood monitoring is a complex task involving multimodal data mining and multitask collaboration. In order to leverage the role of multimodal data in flood management, conducting visual-language pretraining (VLP) in the field of flood disaster monitoring and obtaining foundational pretraining models that are suitable for multiple downstream flood-related tasks is an urgent problem that needs to be addressed. This paper introduces SLIP-Flood, an innovative VLP framework supporting flood image classification, image-text retrieval, and auxiliary text classification. To overcome the limitations of existing cross-modal models that rely on small datasets and lack robustness, we have constructed two specialized datasets for the first time: 1) FloodMulS for the Flood Image Classification Model (FICM), and 2) FloodIT for the Flood Text-Image Retrieval Model (FTIRM). Traditional models employ “Hard Categorization Strategy (HC)” for image classification, neglecting the impacts of “Categorization Ambiguity.” To improve performance, we propose a “Soft Categorization Strategy.” Furthermore, traditional models focus on unimodal (image) information, not fully utilizing joint image-text information. We address this with a “Soft Combination” to integrate FICM and FTIRM, termed SCSC. Experimental results show SCSC improves SLIP-Flood’s performance: a 7.62% increase in the F1 score on FICM compared to HC, and a 0.35% increase in FTIRM’s F1 score based on FICM. SLIP-Flood also achieves a maximum recall of 89.24% in image-text retrieval and shows promise in auxiliary flood text classification. Relevant resources are available at https://github.com/muhan-yy/SLIP-Flood.git.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104543"},"PeriodicalIF":7.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869208","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":"LogiTide2DEM: A method for reconstructing intertidal topography in complex tidal flats using logistic regression with multi-temporal Sentinel-2 and Landsat imagery","authors":"Yi-Chin Chen, Su-Fen Wang","doi":"10.1016/j.jag.2025.104561","DOIUrl":"10.1016/j.jag.2025.104561","url":null,"abstract":"<div><div>Intertidal zones are transitional areas between marine and terrestrial systems and are significant for ecological systems, socio-economic activities, and coastal erosion mitigation. However, intertidal topography is often degraded and lost by natural coastal processes and anthropogenic impacts. Accurate and frequent mapping of intertidal topography is essential for effective coastal management. Previous studies have relied on ground-based surveys or shoreline method using satellite imagery to measure intertidal topography, but these approaches are often limited in spatial extent and temporal resolution. To address these limitations, we propose LogiTide2DEM, a novel method for reconstructing intertidal topography. This method uses logistic regression to determine optimal tide heights for water-land differentiation as surface elevation using multi-temporal Sentinel-2 and Landsat imagery. The advantages of LogiTide2DEM include: a pixel-based approach rather than shoreline methods, elevation constraints using tidal height records, adaptive elevation estimation using logistic regression, and integrated quality assessment of topographic data. We applied and validated LogiTide2DEM on the Changhua coast in central-western Taiwan, reconstructing 22 years (2002–2023) of intertidal topography. Results show root-mean-square differences in elevation ranging from 0.24 m to 0.59 m, with the ratio of mean absolute error to the DEM elevation range between 11 % and 13 %, and correlation coefficients of 0.69 to 0.81. Despite a moderate omission error in water-land classification, LogiTide2DEM demonstrates robust performance in reconstructing intertidal topography. This study shows the capability of LogiTide2DEM to generate high-resolution, reliable topography spanning past decades, providing valuable insights for coastal management and research.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104561"},"PeriodicalIF":7.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873721","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}
D. García-Rodríguez , A. Pérez-Hoyos , B. Martínez , Pablo Catret Ruber , J. Javier Samper-Zapater , E. López-Baeza , J.J. Martínez Durá
{"title":"Comparative analysis of different algorithms for VAS station land cover classification with limited training points","authors":"D. García-Rodríguez , A. Pérez-Hoyos , B. Martínez , Pablo Catret Ruber , J. Javier Samper-Zapater , E. López-Baeza , J.J. Martínez Durá","doi":"10.1016/j.jag.2025.104537","DOIUrl":"10.1016/j.jag.2025.104537","url":null,"abstract":"<div><div>The Valencian Anchor Station (VAS) (Spain) is an outstanding site operating as a central location for calibrating and validating numerous remote sensing instruments and products. Hence, an accurate characterization of its land cover is required. This research conducts a land cover classification within the VAS station (10 × 10 km<sup>2</sup>) and its surrounding area (30 × 30 km<sup>2</sup>) for 2021 using multi-temporal imagery from Sentinel-2 Multispectral Instrument (MSI). Several aspects of land cover classification have been evaluated, including <em>i)</em> the feature selection, <em>ii)</em> the temporal resolution of time series (i.e., monthly, seasonal), <em>iii)</em> the performance of six Machine Learning algorithms (i.e., CART, GTB, k-NN, NB, RF, and SVM, alongside three deep learning models (FC-NN, MLP-ED, and ResCNN) and <em>iv)</em> the optimization of classifier tuning parameters. Furthermore, the study assesses the impact of reducing sample sizes on classifying similar areas, extending the classification to three buffers (1 km, 5 km and 10 km) without increasing reference data. ResCNN emerged as the best-performing model, yielding superior classification metrics (96 % overall accuracy and 95 % kappa score) in July, coinciding with the peak vineyard phenology. Producer’s and user’s accuracy values generally exceeded 90 % for most land cover classes, with some exceptions in more challenging categories such as artificial surfaces and non-irrigated arable land, which showed lower accuracies due to inter-class similarity. Overall, the findings underscore the robust performance of all models in land cover classification, demonstrating the feasibility of achieving high-quality classification with a robust methodology and limited training data.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104537"},"PeriodicalIF":7.6,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869207","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":"Assessing habitat fragmentation in highly urbanized area: Implications for mitigating landscape dissection and habitat restoration","authors":"Wei Hou , Jixian Zhang , Zihao Peng , Ulrich Walz","doi":"10.1016/j.jag.2025.104557","DOIUrl":"10.1016/j.jag.2025.104557","url":null,"abstract":"<div><div>Effective measurement and monitoring of habitat fragmentation are essential for achieving the United Nations’ Sustainable Development Goals, especially given the ongoing global expansion of urbanization. However, current researches primarily address habitat connectivity or landscape dissection separately, while neglecting the impact of landscape dissection on habitat pattern. In this study, we adopt Road network tiles (RNTs, created through the dissection of the landscape by road networks) as analytical units and develop an integrated index of effective habitat network area to assess habitat fragmentation (structural connectivity), with particular emphasis on landscape dissection. Furthermore, a connection efficiency index is proposed to quantify the ratio of current habitat connectivity to its maximum potential. Collectively interpreting these indices reflects both the adverse effects of road expansion on habitat patterns and the effectiveness of mitigation efforts in spatial planning. These indices were applied to the Beijing-Tianjin-Hebei (BTH) region of China. The results show that the effective area of habitat networks in the BTH region declined from 887.66 km2 to 488.50 km2 from 2007 to 2018, despite an overall increase in total habitat area. Nevertheless, connection efficiency for most cities in the BTH region has improved, suggesting that habitat connectivity in 2018 was more efficient and closer to its maximum value compared to 2007, potentially due to effective landscape management practices. By incorporating urban permeation into the joint analysis with RNTs, key zones for habitat conservation and restoration can be identified. This study offers a novel and efficient approach to quantifying habitat fragmentation, providing valuable insights for decision-making aimed at mitigating the negative impacts of urbanization on habitat pattern.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104557"},"PeriodicalIF":7.6,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143864357","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}
Guanya Wang , Zhiwei Li , Han Gao , Jun Hu , Mi Jiang , Peng Ren , Jie Zhang
{"title":"Adaptive sequential estimator for InSAR time series phase estimation","authors":"Guanya Wang , Zhiwei Li , Han Gao , Jun Hu , Mi Jiang , Peng Ren , Jie Zhang","doi":"10.1016/j.jag.2025.104552","DOIUrl":"10.1016/j.jag.2025.104552","url":null,"abstract":"<div><div>The coherence estimation errors in phase linking can be mitigated through the weighted alignment of interferometric pairs and the intermediate filtering of data subsets. The Sequential Estimator (SE) serves as a representative method. It divides the coherence-weighted matrix into smaller subsets, using image compression and recursive estimation to enhance phase linking. However, the SE method has inherent limitations due to its dependence on fixed subset size and manual parameter setting, which hinder its application in complex, natural scenarios. In such environments, the distributions of coherent and low-coherence signals are often unpredictable. To address such limitations, this paper proposes an Adaptive Sequential Estimator (ASE) method. First, an adaptive coherence-weighted matrix partitioning method is proposed. Utilizing Otsu’s algorithm and a local subset merging algorithm, it adaptively generates data subsets which are dynamically tailored to the coherence distribution. Second, a modified sequential estimator is proposed. It selects the optimal subsets from the list with multiple merging degrees, to guide image compression and recursive phase estimation. Based on these, the ASE method adaptively prioritizes coherent information while minimizing the impact of decorrelation noise, thereby improving phase estimation accuracy. Experimental evaluation is conducted using 30 Radarsat-2 SAR images with VV polarization, including the quantitative and visual comparisons between the ASE method and existing methods. The results indicate that the ASE method outperforms other methods, and is particularly well-suited to handling the variable coherence matrix in natural scenarios. Compared to SE, the ASE method increases the distributed scatterer point density with 7%.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104552"},"PeriodicalIF":7.6,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143864358","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}
Wenkai Cui , Lin Yang , Lei Zhang , Chenconghai Yang , Chenxu Zhu , Chenghu Zhou
{"title":"A novel approach of generating pseudo revisited soil sample data based on environmental similarity for space-time soil organic carbon modelling","authors":"Wenkai Cui , Lin Yang , Lei Zhang , Chenconghai Yang , Chenxu Zhu , Chenghu Zhou","doi":"10.1016/j.jag.2025.104542","DOIUrl":"10.1016/j.jag.2025.104542","url":null,"abstract":"<div><div>Soil Organic Carbon (SOC) is vital for the global carbon cycle, agricultural sustainability, and climate change. Process-based models like Rothamsted carbon model (RothC) simulate SOC dynamics, but their accuracy relies on revisited soil samples for calibration, which are often scarce, especially in environmentally heterogeneous regions. This study proposes a novel approach to generate pseudo revisited samples using environmental similarity, addressing the lack of revisited samples. For each intervening-year sample, pseudo SOC stocks in unsampled years were constructed by calculating environmental similarity with existing samples and applying weighted averaging. These pseudo SOC stocks served as revisited samples for model calibration. Bayesian optimization was used to adjust RothC’s microbial activity parameters. Validation results showed the approach significantly improved predictive accuracy, with an RMSE of 5.28 t/ha (31.6 % lower than global parameter optimization and 10.7 % lower than spatiotemporal Random Forest) and an R2 improved from 0.319 (spatiotemporal Random Forest) to 0.456. The approach enhances SOC model accuracy by leveraging environmental similarity and parameter optimization, offering a practical solution for regions lacking revisited samples and improving long-term SOC dynamics simulations. This approach not only addresses data scarcity but also provides more reliable predictions for climate and agricultural management.</div></div>","PeriodicalId":73423,"journal":{"name":"International journal of applied earth observation and geoinformation : ITC journal","volume":"139 ","pages":"Article 104542"},"PeriodicalIF":7.6,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859548","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}