Geospatial digital mapping of soil organic carbon using machine learning and geostatistical methods in different land uses

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 6, 2025

Improper management of soil resources leads to the destruction organic carbon (SOC) stock and, as a result, reduction quality, well accelerating process climate change through release SOC into atmosphere. This study was conducted evaluate potential different simulation models map spatial variability affected by land use in area Qarasu watershed Kermanshah province, west Iran. Map sampling points prepared using Latin hypercube method. A total 168 observation were selected and profile dug described these points. The samples taken horizon determine content laboratory. mapped kriging geostatistical method area. changes simulated multivariate analysis machine learning methods including generalized linear model (GLM), additive (LAM), cubist, random forest (RF), support vector (SVM) models. Comprehensive measurement data is utilized develop validate predictive Predictor variables included 16 topographic two vegetation, six parent material, four climatic variables. In-depth statistical analyses are proposed performance. results showed that ranged from 0.19 8.44 percent uses. spherical variogram with MAE = 0.41 best fits interpolate ordinary LAM estimated wider range (SOC 0.18–4.82%) among model. However, RF (R2 0.64 RMSE 0.58%) most accurate predicting quantity comparing other It can be used reliable predict similar semiarid regions West Asia Among predictor variables, material's intrinsic properties topography had greatest effect variability.

Language: Английский

---

Digital mapping of soil pH and carbonates at the European scale using environmental variables and machine learning

The Science of The Total Environment, Journal Year: 2022, Volume and Issue: 856, P. 159171 - 159171

Published: Sept. 30, 2022

Language: Английский

---

Evaluation of Agricultural Bare Soil Properties Retrieval from Landsat 8, Sentinel-2 and PRISMA Satellite Data

Remote Sensing, Journal Year: 2022, Volume and Issue: 14(3), P. 714 - 714

Published: Feb. 2, 2022

The PRISMA satellite is equipped with an advanced hyperspectral Earth observation technology capable of improving the accuracy quantitative estimation bio-geophysical variables in various Science Applications and particular for soil science. purpose this research was to evaluate ability imager estimate topsoil properties (i.e., organic carbon, clay, sand, silt), comparison current multispectral sensors. To investigate expectation, a test carried out using data collected Italy following two approaches. Firstly, PRISMA, Sentinel-2 Landsat 8 spectral simulated datasets were obtained from resampling laboratory library. Subsequently, bare reflectance experimental areas Italy, real satellites images, at dates close each other. models calibrated employing both Partial Least Square Regression Cubist algorithms. results study revealed that best accuracies retrieving by data, datasets. Indeed, resampled spectra provided Ratio Performance Inter-Quartile distance (RPIQ) clay (4.87), sand (3.80), carbon (2.59) estimation, library For imagery, higher level prediction RPIQ ± SE values 2.32 0.07 3.85 0.19 silt, 3.51 0.16 carbon. imagery performance SOC. same better estimated PLSR case data. statistical retrieval SOC potential actual satellite. supported expected good properties.

Language: Английский

---

The Role of Remote Sensing Data and Methods in a Modern Approach to Fertilization in Precision Agriculture

Remote Sensing, Journal Year: 2022, Volume and Issue: 14(3), P. 778 - 778

Published: Feb. 7, 2022

The precision fertilization system is the basis for upgrading conventional intensive agricultural production, while achieving both high and quality yields minimizing negative impacts on environment. This research aims to present application of modern prediction methods in by integrating agronomic components with spatial component interpolation machine learning. While were a cornerstone soil past decades, new challenges process larger more complex data have reduced their viability present. Their disadvantages lower accuracy, lack robustness regarding properties input sample values requirements extensive cost- time-expensive sampling addressed. Specific (ordinary kriging, inverse distance weighted) learning (random forest, support vector machine, artificial neural networks, decision trees) evaluated according popularity relevant studies indexed Web Science Core Collection over decade. As shift towards increased accuracy computational efficiency, an overview state-of-the-art remote sensing improving precise was completed, accent open-data global satellite missions. State-of-the-art techniques allowed hybrid predict sampled supported such as high-resolution multispectral, thermal radar or unmanned aerial vehicle (UAV)-based imagery analyzed studies. representative approaches performed based 121 samples phosphorous pentoxide (P2O5) potassium oxide (K2O) common parcel Croatia. It visually quantitatively confirmed superior retained local heterogeneity approach. concludes that significant role agriculture today will be increasingly important future.

Language: Английский

---

Exploring the capability of Gaofen-5 hyperspectral data for assessing soil salinity risks

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2022, Volume and Issue: 112, P. 102969 - 102969

Published: Aug. 1, 2022

Soil salinization has hampered the achievement of sustainable development goals (SDGs) in many countries worldwide. Several have recently launched hyperspectral remote sensing satellites, opening new avenues for accurate soil-salinity monitoring. Among them, Gaofen-5 (GF-5) from China a high comprehensive performance, including spectral resolution 5 nm, 330 bands, and signal-to-noise ratio 700. However, potential GF-5 estimating soil salinity is not well understood. In this study, we proposed strategy that includes bootstrap methods, fractional order derivative (FOD) techniques decision-level fusion models to exploit diagnostic information reduce estimation uncertainty Ebinur Lake oasis northwestern China. The results showed data were suitable assessing salinity. FOD technique enhanced correlation between spectra, identified more improved accuracy estimation, reduced model uncertainty. low-order outperformed high-order FOD. spectra processed by 0.9 most correlated with (r = −0.76). driven 0.8 produced optimal estimated (R2 0.95, root mean square error (RMSE) 3.20 dS m−1 performance interquartile distance (RPIQ) 5.96). had less than based on original integer-order (first- second- derivatives) spectra. This study provides reference using framework low accuracy. great environmental problems facilitating further SDGs.

Language: Английский

---

Improving pedotransfer functions for predicting soil mineral associated organic carbon by ensemble machine learning

Geoderma, Journal Year: 2022, Volume and Issue: 428, P. 116208 - 116208

Published: Oct. 24, 2022

Language: Английский

---

Improving soil organic carbon predictions from a Sentinel–2 soil composite by assessing surface conditions and uncertainties

Geoderma, Journal Year: 2022, Volume and Issue: 429, P. 116128 - 116128

Published: Nov. 10, 2022

Soil organic carbon (SOC) prediction from remote sensing is often hindered by disturbing factors at the soil surface, such as photosynthetic active and non–photosynthetic vegetation, variation in moisture or surface roughness. With increasing amount of freely available satellite data, recent studies have focused on stabilizing reflectance building composites using time series images. Although composite imagery has demonstrated its potential SOC prediction, it still not well established if resulting spectra mirror fingerprint optimal conditions to predict topsoil properties (i.e. a smooth, dry bare soil). We collected 303 photos surfaces Belgian loam belt where five main classes were distinguished: smooth seeded soils, crusts, partial cover growing crop, moist soils crop residue cover. Reflectance then extracted Sentinel–2 images coinciding with date photos. After was removed an NDVI < 0.25, Normalized Burn Ratio (NBR2) calculated characterize threshold NBR2 0.05 found be able separate unfavorable i.e. wet covered residues. Additionally, we that normalizing dividing each band mean all spectral bands) allows for cancelling albedo shift between crusts seed–bed conditions. built exposed southern Belgium part Noord-Holland Flevoland Netherlands (covering spring periods 2016–2021). used per pixel content means Partial Least Squares Regression Model (PLSR) 10–fold cross–validation. The uncertainty models assessed via interval ratio (PIR). cross validation model gave satisfactory results (mean 100 bootstraps: efficiency coefficient (MEC) = 0.48 ± 0.07, RMSE 3.5 0.3 g C kg–1, RPD 1.4 0.1 RPIQ 1.9 0.3). maps show decreases when number scenes increases, reaches minimum least six are PIR pixels 12.4 while predicted 14.1 kg–1). against independent data set showed median difference 0.5 kg–1 2.8 measured (average 13.5 kg–1) contents field scale. Overall, this compositing method shows both realistic within regional patterns.

Language: Английский

---

Assessing Machine Learning-Based Prediction under Different Agricultural Practices for Digital Mapping of Soil Organic Carbon and Available Phosphorus

Agriculture, Journal Year: 2022, Volume and Issue: 12(7), P. 1062 - 1062

Published: July 20, 2022

Predicting soil chemical properties such as organic carbon (SOC) and available phosphorus (Ava-P) content is critical in areas where different land uses exist. The distribution of SOC Ava-P influenced by both natural anthropogenic factors. This study aimed at (1) predicting a piedmont plain Northeast Iran using the Random Forests (RF) Cubist mathematical models hybrid (Regression Kriging), (2) comparing models’ results, (3) identifying key variables that influence spatial dynamics under agricultural practices. machine learning were trained with 201 composite surface samples 24 ancillary data, including climate (C), organism (O), topography- relief (R), parent material (P) features (S) according to SCORPAN digital mapping framework, which can predictively represent formation factors spatially. Clay, one most well-known relationship SOC, was important predictor followed open-access multispectral satellite images-based vegetation indices. had similar set effective variables. Hybrid approaches did not improve model accuracy significantly, but they reduce map uncertainty. In validation set, calculated RF algorithm normalized root mean square (NRMSE) 96.8, while an NRMSE 94.2. These values change when technique for Ava-P; however, changed just 1% SOC. management supply activities be guided maps. Produced maps scientist plays active role used identify concentrations are high need protected, uncertainty sampling required further monitoring.

Language: Английский

---

Stoichiometry of soil carbon, nitrogen, and phosphorus in farmland soils in southern China: Spatial pattern and related dominates

CATENA, Journal Year: 2022, Volume and Issue: 217, P. 106468 - 106468

Published: June 21, 2022

Language: Английский

---

Integrating additional spectroscopically inferred soil data improves the accuracy of digital soil mapping

Geoderma, Journal Year: 2023, Volume and Issue: 433, P. 116467 - 116467

Published: April 6, 2023

Digital soil mapping has been increasingly advocated as an efficient approach to deliver fine-resolution and up-to-date information in evaluating ecosystem services. Considering the great spatial heterogeneity of soils, it is widely recognized that more representative observations are needed for better capturing variation thus increase accuracy digital maps. In reality, budget field work laboratory analysis commonly limited due its high cost low efficiency. last two decades, being alternative wet chemistry, spectroscopy, such visible-near infrared (Vis-NIR), mid-infrared (MIR) spectroscopy developed measuring a rapid cost-effective manner enable collect (DSM). However, spectroscopically inferred (SI) data subject higher uncertainties than reference analysis. Many DSM practices integrated SI with into modelling while few studies addressed key question whether these non-errorless improve map DSM. this study, French Soil Monitoring Network (RMQS) Land Use Coverage Area frame Survey (LUCAS Soil) datasets were used evaluate potential from Vis-NIR MIR properties (i.e. organic carbon, clay, pH) at national scale. Cubist quantile regression forests spectral predictive modelling, respectively. For both RMQS LUCAS dataset, different scenarios regarding varying proportions tested models models. Repeated (50 times) external validation suggested adding additional can performance regardless (gain R2 proportion 3–19%) when (≤50%). Lower model led greater improvement Our results also showed lowered prediction intervals which may result underestimation uncertainty. The determination threshold on use needs be explored future studies.

Language: Английский

---