Improving the spatial prediction of topsoil properties in a typical grazing area using multi-season PlanetScope spectral covariates and data mining techniques

Earth Science Informatics, Год журнала: 2025, Номер 18(2)

Опубликована: Фев. 1, 2025

Abstract Understanding the spatial distribution of topsoil properties in grassland ecosystems is essential for improving soil ecosystem services, quality, and erosion resilience. The availability free, high-resolution satellite imagery advanced data mining techniques offers new opportunities efficient property assessment. This study aimed to evaluate potential utility multi-season PlanetScope predict organic carbon (SOC), pH, calcium carbonate (CaCO 3 ). Using random sampling, 121 samples (0–30 cm depth) were collected with an auger across grasslands, bare soil, eroded areas within a typical grazing land use. Three techniques: forest (RF), extreme gradient boosting (XGB), support vector machines (SVM), applied evaluated using 10-fold cross-validation. results indicated that spectral covariates considerably improved accuracy target compared single-season imagery. SVM was most effective algorithm predicting SOC, achieving root mean square error (RMSE) 0.52%, absolute (MAE) 0.24%, R² 0.92. RF best-performing pH (RMSE = 0.22, MAE 0.17, 0.97) CaCO 0.55%, 0.42%, 0.96). While XGB failed capture variability other models generated interpretable maps accurately represented different cover categories. green-red vegetation index (GRVI) critical covariate while elevation topographic wetness (TWI) key predictors , respectively. underscores recommends conducting similar studies diverse geographical settings validate these findings develop more generalizable models.

Язык: Английский

---

Evaluating Airborne Hyperspectral Scanner (AHS) for the mapping of soil organic matter and clay in a Mediterranean forest ecosystem

CATENA, Год журнала: 2025, Номер 252, С. 108889 - 108889

Опубликована: Март 4, 2025

Язык: Английский

---

Soil Organic Carbon Assessment Using Remote-Sensing Data and Machine Learning: A Systematic Literature Review

Remote Sensing, Год журнала: 2025, Номер 17(5), С. 882 - 882

Опубликована: Март 1, 2025

In the current global change scenario, valuable tools for improving soils and increasing both agricultural productivity food security, together with effective actions to mitigate impacts of ongoing climate trends, are priority issues. Soil Organic Carbon (SOC) acts on these two topics, as C is a core element soil organic matter, an essential driver fertility, becomes problematic when disposed in atmosphere its gaseous form. Laboratory methods measure SOC expensive time-consuming. This Systematic Literature Review (SLR) aims identify techniques alternative ways estimate using Remote-Sensing (RS) spectral data computer process this database. SLR was conducted Meta-Analysis (PRISMA) methodology, highlighting use Deep Learning (DL), traditional neural networks, other machine-learning models, input were used SOC. The concludes that Sentinel satellites, particularly Sentinel-2, frequently used. Despite limited datasets, DL models demonstrated robust performance assessed by R2 RMSE. Key data, such vegetation indices (e.g., NDVI, SAVI, EVI) digital elevation consistently correlated predictions. These findings underscore potential combining RS advanced artificial-intelligence efficient scalable monitoring.

Язык: Английский

---

National-scale mapping topsoil organic carbon of cropland in China using multitemporal Sentinel-2 images

Geoderma, Год журнала: 2025, Номер 456, С. 117272 - 117272

Опубликована: Март 30, 2025

Язык: Английский

---

Spatial modelling and drivers of soil organic carbon across successional communities in tropical deciduous forests: insights from Northwest Himalayan foothills

Environmental Monitoring and Assessment, Год журнала: 2025, Номер 197(5)

Опубликована: Апрель 7, 2025

Язык: Английский

---

Individual And Pairwise Representativeness Of Sampling Points In Interpolation Tasks Of Heavy Metals Distribution In The Topsoil

GEOGRAPHY ENVIRONMENT SUSTAINABILITY, Год журнала: 2025, Номер 18(1), С. 6 - 13

Опубликована: Апрель 7, 2025

The optimization of environmental soil monitoring based on representative selection a training subset for an artificial neural network is unresolved problem in the tasks interpolation distribution metals topsoil. survey data, often used as input modeling, are datasets at irregular points. Usually, division data into and test subsets carried out randomly ratio 70% to 30% points, respectively. question individual collective representativeness local sampling points element content given area remains beyond scope problems. In this work, plays crucial role reducing ANN error enhancing correlation between results model calculations natural measurements when part subset. When evaluating pairwise representativeness, we found two types effects: synergy anti-synergy. was achieved with increase accuracy pair entered anti-synergy manifested decrease informativeness point modeling. various locations have different information unequal meaning feature interpolation. scale-free structures were by RMSE .

Язык: Английский

---

Variability analysis of soil organic carbon content across land use types and its digital mapping using machine learning and deep learning algorithms

Environmental Monitoring and Assessment, Год журнала: 2025, Номер 197(5)

Опубликована: Апрель 10, 2025

Язык: Английский

---

A Novel Method for Estimating Soil Organic Carbon Density Using Soil Organic Carbon and Gravel Content Data

Sustainability, Год журнала: 2025, Номер 17(8), С. 3533 - 3533

Опубликована: Апрель 15, 2025

Soil organic carbon density (SOCD) is crucial for assessing soil (SOC) storage, but its estimation remains challenging when bulk (BD) data are unavailable. Traditional methods substituting missing BD data, including using the mean, median, and pedotransfer functions (PTFs), introduce varying degrees of uncertainty in SOCD estimation: (1) The mean median ignore effects type, environmental conditions, land use changes on BD. They also heavily rely representativeness samples, which may lead to systematic bias. (2) accuracy PTFs depends modeling approaches, variable selection, dataset characteristics, differences among biases SOCD. To overcome this challenge, we analyzed 443 profiles from Yangtze River Delta region China developed an innovative approach that estimates only SOC gravel content data. By formulating linear, polynomial, power function regression models, directly estimated per centimeter horizon i (SOCDicm) under conditions with without available followed by calculation. results indicated a strong correlation between SOCDicm, three models direct SOC-based SOCDicm yielding consistently high accuracy. Neglecting overall resulted overestimation 7.01–9.45%. After incorporating as correction factor, new method estimating was improved, prediction set achieving R² values 0.927–0.945, RMSE 0.819–0.949 kg m−2, RPIQ 4.773–5.533. surpassed comparable PTF method, thus enabling reliable estimation. This study introduces developing regional estimate rapid samples information historical provides methodology calculating global stocks. contributes improving stock estimation, supporting management cycle research, providing scientific evidence sustainable agricultural development climate change mitigation strategies.

Язык: Английский

---

Prediction of soil organic matter using Landsat 8 data and machine learning algorithms in typical karst cropland in China

European Journal of Agronomy, Год журнала: 2024, Номер 160, С. 127323 - 127323

Опубликована: Авг. 27, 2024

Язык: Английский

---

A dynamic normalized difference index for estimating soil organic matter concentration using visible and near-infrared spectroscopy

Ecological Indicators, Год журнала: 2023, Номер 147, С. 110037 - 110037

Опубликована: Фев. 17, 2023

Accurate, rapid, and non-destructive estimation of soil organic matter (SOM) is crucial for fertility diagnosis precision farming. Due to the complicated unstable spectral characteristics SOM, few SOM indexes have been proposed widely used. In this paper, a new dynamic normalized difference index (DNDI) was constructed estimate using visible near-infrared spectroscopy. A correction factor α used adjust optimal wavelength range obtain more robust features SOM. Different pre-processing methods were applied compared. The support vector machine (SVM) model Partial least square regression (PLSR) calibrated based on DNDI To end, total 111 samples collected in southern coastal plain Laizhou Bay. results showed that by optimization could higher correlation with than two-dimensional (NDI). had maximum 0.88 from first derivative reflectance, NDI correlations most improved standard normal variate transform (SNV), reaching 0.81. For models, exhibited better performance, yielding validation R2, RMSE, RPD 0.78, 0.17 g·kg−1, 2.01, respectively. Our algorithm has strong application potential estimating other properties enhancing ground- satellite-based sensing.

Язык: Английский

---