Synergetic Use of Bare Soil Composite Imagery and Multitemporal Vegetation Remote Sensing for Soil Mapping (A Case Study from Samara Region’s Upland) DOI Creative Commons
А. В. Чинилин, Nikolai Lozbenev, P. M. Shilov

et al.

Land, Journal Year: 2024, Volume and Issue: 13(12), P. 2229 - 2229

Published: Dec. 20, 2024

This study presents an approach for predicting soil class probabilities by integrating synthetic composite imagery of bare with long-term vegetation remote sensing data and survey data. The goal is to develop detailed maps the agro-innovation center “Orlovka-AIC” (Samara Region), a focus on lithological heterogeneity. Satellite were sourced from cloud-filtered collection Landsat 4–5 7 images (April–May, 1988–2010) 8–9 (June–August, 2012–2023). Bare surfaces identified using threshold values NDVI (<0.06), NBR2 (<0.05), BSI (>0.10). Synthetic generated calculating median reflectance across available spectral bands. Following adoption no-till technology in 2012, average additionally calculated assess condition agricultural lands. Seventy-one sampling points within classified both Russian WRB classification systems. Logistic regression was applied pixel-based prediction. model achieved overall accuracy 0.85 Cohen’s Kappa coefficient 0.67, demonstrating its reliability distinguishing two main classes: agrochernozems agrozems. resulting map provides robust foundation sustainable land management practices, including erosion prevention use optimization.

Language: Английский

Unveiling year-round cropland cover by soil-specific spectral unmixing of Landsat and Sentinel-2 time series DOI Creative Commons
Felix Lobert, Marcel Schwieder, Jonas Alsleben

et al.

Remote Sensing of Environment, Journal Year: 2025, Volume and Issue: 318, P. 114594 - 114594

Published: Jan. 9, 2025

Language: Английский

Citations

2
Towards Explainable AI: Interpreting Soil Organic Carbon Prediction Models Using a Learning‐Based Explanation Method DOI Creative Commons
Nafiseh Kakhani, Ruhollah Taghizadeh–Mehrjardi, Davoud Omarzadeh

et al.

European Journal of Soil Science, Journal Year: 2025, Volume and Issue: 76(2)

Published: Feb. 24, 2025

ABSTRACT An understanding of the key factors and processes influencing variability soil organic carbon (SOC) is essential for development effective policies aimed at enhancing storage in soils to mitigate climate change. In recent years, complex computational approaches from field machine learning (ML) have been developed modelling mapping SOC various ecosystems over large areas. However, order understand that account ML models serve as a basis new scientific discoveries, predictions made by these data‐driven must be accurately explained interpreted. this research, we introduce novel explanation approach applicable any model investigate significance environmental features explain across Germany. The methodology employed study involves training multiple using content measurements LUCAS dataset incorporating derived Google Earth Engine (GEE) explanatory variables. Thereafter, an applied elucidate what learned about relationship between supervised manner. our approach, post hoc trained estimate contribution specific inputs outputs models. results indicate different classes rely on interpretable but distinct variability. Decision tree‐based models, such random forest (RF) gradient boosting, highlight importance topographic features. Conversely, chemical information, particularly pH, crucial performance neural networks linear regression Therefore, interpreting studies requires carefully structured guided expert knowledge deep being analysed.

Language: Английский

Citations

1
Assessing the potential of multi-source remote sensing data for cropland soil organic matter mapping in hilly and mountainous areas DOI
Peng Li, Xiaobo Wu,

C Feng

et al.

CATENA, Journal Year: 2024, Volume and Issue: 245, P. 108312 - 108312

Published: Aug. 12, 2024

Language: Английский

Citations

5
Satellite Soil Observation (Satsoil): Extraction of Bare Soil Reflectance for Soil Organic Carbon Mapping on Google Earth Engine DOI
Morteza Khazaei, Preston Sorenson, Ramata Magagi

et al.

Published: Jan. 1, 2025

Language: Английский

Citations

0
Time series of Landsat-based bimonthly and annual spectral indices for continental Europe for 2000–2022 DOI Creative Commons
Xuemeng Tian, Davide Consoli, Martijn Witjes

et al.

Earth system science data, Journal Year: 2025, Volume and Issue: 17(2), P. 741 - 772

Published: Feb. 26, 2025

Abstract. The production and evaluation of the analysis-ready cloud-optimized (ARCO) data cube for continental Europe (including Ukraine, UK, Türkiye), derived from Landsat dataset version 2 (ARD V2) produced by Global Land Analysis Discovery (GLAD) team covering period 2000 to 2022, is described. consists 17 TB at a 30 m resolution includes bimonthly, annual, long-term spectral indices on various thematic topics, including surface reflectance bands, normalized difference vegetation index (NDVI), soil adjusted (SAVI), fraction absorbed photosynthetically active radiation (FAPAR), snow (NDSI), water (NDWI), tillage (NDTI), minimum (minNDTI), bare (BSF), number seasons (NOS), crop duration ratio (CDR). was developed with intention provide comprehensive feature space environmental modeling mapping. quality time series assessed (1) assessing accuracy gap-filled bimonthly artificially created gaps; (2) visual examination artifacts inconsistencies; (3) plausibility checks ground survey data; (4) predictive tests, examples organic carbon (SOC) land cover (LC) classification. reconstruction demonstrates high accuracy, root mean squared error (RMSE) smaller than 0.05, R2 higher 0.6, across all bands. indicates that product complete consistent, except winter periods in northern latitudes high-altitude areas, where cloud density introduce significant gaps hence many remain. check further shows logically statistically capture processes. BSF showed strong negative correlation (−0.73) coverage data, while minNDTI had moderate positive (0.57) Eurostat practice data. detailed temporal characteristics provided different tiers predictors this proved be important both regression LC classification experiments based 60 723 LUCAS observations: (tier 4) were particularly valuable mapping SOC LC, coming out top variable importance assessment. Crop-specific (NOS CDR) limited value tested applications, possibly due noise or insufficient quantification methods. made available https://doi.org/10.5281/zenodo.10776891 (Tian et al., 2024) under CC-BY license will continuously updated.

Language: Английский

Citations

0
Impact of data density and endmember definitions on long-term trends in ground cover fractions across European grasslands DOI Creative Commons
Katarzyna Ewa Lewińska, Akpona Okujeni, Katja Kowalski

et al.

Remote Sensing of Environment, Journal Year: 2025, Volume and Issue: 323, P. 114736 - 114736

Published: April 5, 2025

Language: Английский

Citations

0
Soil Organic Carbon Prediction and Mapping in Morocco Using PRISMA Hyperspectral Imagery and Meta-Learner Model DOI Creative Commons
Yassine Bouslıhım, Abdelkrim Bouasria, Budiman Minasny

et al.

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(8), P. 1363 - 1363

Published: April 11, 2025

Accurate mapping of soil organic carbon (SOC) supports sustainable land management practices and accounting initiatives for mitigating climate change impacts. This study presents a novel meta-learner framework that combines multiple machine learning algorithms spectra processing to optimize SOC prediction using the PRISMA hyperspectral satellite imagery in Doukkala plain Morocco. The employs two-layer structure models. first layer consists Random Forest (RF), Support Vector Regression (SVR), Partial Least Squares (PLSR). These base models were configured data smoothing, transformation, spectral feature selection techniques, based on 70/30% split. second utilizes ridge regression model as integrate predictions from Results indicated RF SVR performance improved primarily with selection, while PLSR was most influenced by smoothing. approach outperformed individual models, achieving an average relative improvement 48.8% over single R2 0.65, RMSE 0.194%, RPIQ 2.247. contributes development methodologies predicting properties data.

Language: Английский

Citations

0
Contribution of the Sentinel-2 spring seedbed spectra to the digital mapping of soil organic carbon concentration DOI Creative Commons
Fien Vanongeval, Jos Van Orshoven, Anne Gobin

et al.

Geoderma, Journal Year: 2024, Volume and Issue: 449, P. 116984 - 116984

Published: Aug. 1, 2024

Soil organic carbon (SOC) is central to the functioning of terrestrial ecosystems, has climate mitigation potential and provides several benefits for soil health. Understanding spatial distribution SOC can help formulate sustainable management practices. Digital mapping (DSM) uses advanced statistical geostatistical methods estimate properties across large areas. DSM integrates data, topographic features, geology, legacy maps, land remote sensing data. Bare spectra may reflect presence particular components, making satellite derived suitable predictors SOC. from Sentinel-2 were used concentration (SOC%) granulometric fractions in plough layer (0–30 cm) agricultural parcels northern Belgium. Thereafter, estimation performance SOC% was compared three models: one with bare spectra, environmental covariates (topography, granulometry vegetation), a combined model covariates. The sand, silt clay using spring seedbed (R2: 0.53–0.74; RPD: 1.49–2.05; RPIQ: 1.52–2.39) higher than that 0.16; 1.08; 1.32). highest obtained including all 0.28; 1.18; 1.44), but contribution containing small. results provide valuable insights refining property spectral

Language: Английский

Citations

2
An explainable spatial interpolation method considering spatial stratified heterogeneity DOI
Shifen Cheng, Wenhui Zhang,

Peng Luo

et al.

International Journal of Geographical Information Science, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 27

Published: Nov. 14, 2024

Spatial interpolation is essential for handling sparsity and missing spatial data. Current machine learning-based methods are subject to the statistical constraints of stratified heterogeneity (SSH), normally involving separate modeling each stratum simple weighted averaging integrate intra-stratum inter-strata features. However, these models overlook different contributions features locations within a (heterogeneous associations, HIA) explanation effects on process, leading suboptimal unreliable outcomes. This article proposes novel explainable method considering SSH (X-SSHM). environmental utilized describe information, which fed into random forest-based learners achieve high-level semantic feature mapping. Geographically regression employed unified expression HIA, obtaining final result. Shapley (GSHAP) proposed decompose marginal Model performance evaluated simulated soil organic matter datasets. X-SSHM outperformed five baselines regarding accuracy. Moreover, validated X-SSHM's ability elucidate mechanisms by SSH, autocorrelation HIA affect model process.

Language: Английский

Citations

2
A Comprehensive Evaluation of Machine Learning Algorithms for Digital Soil Organic Carbon Mapping on a National Scale DOI Creative Commons
Dorijan Radočaj, Danijel Jug, Irena Jug

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(21), P. 9990 - 9990

Published: Nov. 1, 2024

The aim of this study was to narrow the research gap ambiguity in which machine learning algorithms should be selected for evaluation digital soil organic carbon (SOC) mapping. This performed by providing a comprehensive assessment prediction accuracy 15 frequently used SOC mapping based on studies indexed Web Science Core Collection (WoSCC), basis algorithm selection future studies. Two areas, including mainland France and Czech Republic, were 2514 400 samples from LUCAS 2018 dataset. Random Forest first ranked (mainland) then Republic regarding accuracy; coefficients determination 0.411 0.249, respectively, accordance with its dominant appearance previous WoSCC. Additionally, K-Nearest Neighbors Gradient Boosting Machine regression indicated, relative their frequency WoSCC, that they are underrated more considered Future consider areas not strictly related human-made administrative borders, as well interpretable ensemble approaches.

Language: Английский

Citations

0