Search Results (511)

Optical vegetation and soil indices are widely used in Earth observation, although their estimation is strongly affected by cloud coverage and illumination variability. Synthetic-aperture radar (SAR) has therefore attracted increasing interest as an alternative source for spectral index prediction. Most existing studies focus on directly estimating a single index from SAR observations. In this work, we investigate a more flexible formulation in which Sentinel-2 multispectral bands are first reconstructed from Sentinel-1 SAR data and subsequently used to derive multiple spectral indices. Experiments are conducted on the SEN12TP dataset, exploiting near-synchronous paired Sentinel-1 and Sentinel-2 acquisitions together with auxiliary elevation and land-cover information. Three SAR-to-multispectral reconstruction strategies are compared, namely, Efficient-UNet, Pix2Pix, and a conditional flow matching model. The resulting indices are then evaluated against those obtained through dedicated index-specific reconstruction models. The results show that Efficient-UNet achieves the best overall multispectral reconstruction performance among the evaluated architectures. Moreover, indices derived from reconstructed multispectral bands achieve performance comparable to dedicated index-specific models while offering substantially greater flexibility, as multiple indices can be computed within a single framework without retraining task-specific models. At the same time, the experiments highlight important intrinsic limitations of SAR-based spectral reconstruction. Although the reconstructed products preserve the large-scale spatial organization of the scenes, they do not fully recover fine spectral and vegetation-sensitive details. Consequently, SAR-derived spectral indices should be regarded as approximate proxies of optical observations rather than direct substitutes, particularly in applications requiring accurate biophysical interpretation. Full article

Estimation of forest biomass and surface fuels is essential for wildfire risk assessment, carbon accounting, and ecosystem management in semi-arid forests. This study evaluated Sentinel-2 spectral indices and Gamma generalized linear models (GLMs) for estimating aboveground live biomass, forest floor biomass, and downed woody debris in oak–pine forests of Northern Mexico. Spectral predictors were grouped according to chlorophyll activity, vegetation vigor, physiological condition, and soil/background correction effects. Measurements of live biomass, forest floor biomass, and downed woody debris were linked to Sentinel-2 spectral information. Aboveground live biomass was most strongly associated with the chlorophyll-related index CIre8A (p < 0.0001), whereas both forest floor biomass and downed woody debris showed stronger relationships with indices associated with red–NIR reflectance gradients, vegetation senescence, and soil/background correction effects, including the Normalized Difference Vegetation Index (NDVI), the Alpha-weighted Red–NIR Vegetation Index (PVIα), the Normalized Pigment Chlorophyll Index (NPCI), and the Atmospherically Resistant Vegetation Index (ARVI) (p < 0.05). Single-index Gamma GLMs showed significant predictive relationships (p < 0.05), supporting the use of individual Sentinel-2 indices as practical predictors of biomass and surface fuels. In contrast, multivariate models combining several spectral indices were affected by collinearity and parameter instability. Principal Component Analysis (PCA) integrated four representative indices into a single orthogonal predictor. The first principal component (Prin1) explained 90.30%–95.25% of total spectral variance, eliminated collinearity effects (VIF = 1), and produced highly significant Gamma GLMs across all biomass components (p ≤ 0.0009). PCA-based models also yielded lower AIC and BIC values, providing the most parsimonious framework when multiple spectral predictors were combined. These results indicate that individual spectral indices offer practical alternatives for biomass estimation, whereas PCA provides a robust approach for integrating complementary spectral information while maintaining model stability. Full article

Early detection of contrasting water-stress types is important for understanding plant stress responses and improving water-stress management in forest seedlings. In this study, three-year-old Pinus tabuliformis seedlings were exposed to control (CK, 75–85% of field capacity), drought (D, 25–35% of field capacity), waterlogging (WL, water level flush with the soil surface), and water flooding (WF, water level 2 cm above the soil surface) treatments. Each treatment included 15 independent seedlings at each sampling date, resulting in 420 seedlings across four treatments and seven sampling dates. Needle water potential (Ψw), total chlorophyll content (Chl), maximum photochemical efficiency (Fv/Fm), indole-3-acetic acid (IAA), abscisic acid (ABA), and electrical impedance spectroscopy (EIS) parameters were measured to compare the temporal responses of physiological and electrical traits under different water conditions. EIS showed clear treatment-dependent changes under all stress treatments, with larger changes observed under WF and WL than under D at the treatment levels and durations evaluated in this study. Based on the present discrete sampling schedule, the real (Re) and imaginary (Im) components showed statistical separation among the four treatments in all six pairwise comparisons of the four treatments on Day 7, whereas ABA and Ψw showed separation on Day 11, Fv/Fm on Day 18, and IAA and Chl on Day 25. These results indicate that Re and Im showed treatment-dependent divergence at earlier evaluated sampling dates than the selected physiological variables. Exploratory CLAFIC analysis further indicated group-level spectral separation between CK and WL/WF on Day 3 and between CK and D on Day 7. Overall, EIS-derived features may provide useful information on early electrical responses of P. tabuliformis seedlings to contrasting water conditions. However, further validation using independent datasets, standardized measurement procedures, and field or nursery conditions is required before EIS can be applied as a practical tool for water-stress assessment. Full article

Integrating machine learning with spectral data provides an effective and cost-efficient scheme for estimating cadmium (Cd) contents in soils compared with labor-intensive laboratory analyses. However, conventional machine learning–based spectral estimation methods often yield unsatisfactory performance because they rely on an unrealistic assumption that the functional relationships across geographically distinct subregions are homogeneous, thereby reducing the accuracy and stability of soil Cd estimation. To address this issue, a geospatial information-guided XGBoost (GIGS) model was developed, in which a spatial weighting module was incorporated into the XGBoost framework to account for spatial heterogeneity in functional relationships among sampling locations. The spatial heterogeneity of each soil spectral sample was quantified and assigned a corresponding spatial weight, which was subsequently incorporated during model calibration. The optimal principal component spectra (PCS) derived from spectral data were used as model inputs, with measured soil Cd content as the dependent variable, thereby establishing a robust spectral estimation model for soil Cd. Results indicated that the GIGS model exhibited satisfactory performance in the spectral estimation of soil Cd content, with R², RMSE, and RPIQ values of 0.78, 0.04, and 2.02, respectively. Compared to commonly used spectral estimation models (e.g., XGBoost, random forest, support vector regression), the GIGS model achieved a maximum performance improvement of approximately 27.87% and a minimum improvement of approximately 16.42% (with reference to the R² value). Five PCS of soil spectral data were extracted as predictors for the model. PCS-1 was found to be closely associated with iron oxides, while PCS-2 primarily reflects the spectral characteristics of clay minerals. The spectral bands at 600 nm and 815 nm contribute most strongly to PCS-3, which is linked to soil organic matter and indirectly reflects the soil Cd status. PCS-4 and PCS-5 represent mixed spectral information derived from materials associated with soil Cd. An integrated framework combining the GIGS model and PCS data developed in this study provides an accurate and reliable tool for spectral estimation and mapping of soil Cd, thereby supporting cost-effective soil management and environmental sustainability worldwide. Full article

Field-based phenotyping of water-related traits for precision irrigation in tropical agroecosystems poses a persistent methodological challenge, driven by high climatic variability and the complex water-use physiology of Crassulacean Acid Metabolism (CAM) crops such as pineapple (Ananas comosus var. MD2). We developed and validated a Physics-Informed Machine Learning (PIML) framework that integrates high-resolution UAV multispectral imagery, IoT-based microclimatic records, and a mechanistic soil water balance based on the FAO-56 Penman–Monteith standard to predict plot-scale soil moisture depletion as a proxy of plant water status. A six-month field campaign (March–August 2022) across 25 georeferenced commercial pineapple plots in the Colombian Orinoquia piedmont yielded a spatiotemporally balanced dataset of $N=150$ observations. Soil-adjusted vegetation indices (OSAVI, MSAVI) outperformed standard NDVI for capturing water-related canopy traits, effectively decoupling spectral responses from substrate noise. A Gradient Boosting regressor achieved $R^2=0.842$ and $RMSE=0.0705$ on a normalized target scale, corresponding to a $7.05\%$ error over the prediction range, while the traffic-light Decision Support System (DSS) for irrigation scheduling reached $91.1\%$ accuracy (Cohen’s Kappa $=0.91$). Incorporating daily soil moisture depletion as a mechanistic feature improved predictive accuracy over a spectral-only baseline ($\mathrm{\Delta }{R}^2=+0.052$) and anchored predictions within a physically consistent framework based on the FAO-56 water balance, with no false negatives observed for water deficit detection in the hold-out validation set. This framework advances high-throughput, population-scale phenotyping of water-related traits in open-canopy CAM crops, establishing a transferable methodology for operational precision irrigation under tropical savanna conditions. Full article

Cadmium (Cd) contamination in soil poses a serious threat to crop quality. Biochar is widely regarded as an effective amendment that can reduce Cd bioavailability and limit Cd uptake by crops. However, studies on the rapid and nondestructive evaluation of crop physiological responses under biochar-mediated alleviation of Cd stress remain insufficient. Spectral modeling methods can enable rapid and nondestructive monitoring of crop physiological status. In this preliminary experiment, Zhengdan 958 maize seedlings grown in Cd-contaminated soil were subjected to five biochar application rates: 0, 10, 30, 50, and 70 g/pot, designated as CK, A1, A3, A5, and A7, respectively. The study established a non-destructive spectral detection model for relative chlorophyll content expressed as SPAD values of maize leaves to achieve spectral inversion of leaf physiological information. The alleviating effect of biochar on Cd stress was evaluated by analyzing SPAD values and Cd accumulation in roots, stems, and leaves. The original spectral data underwent preprocessing steps including multivariate scattering correction, standard normal variable transformation, normalization, trend removal, first-order derivative transformation, and second-order derivative transformation. The effectiveness of different preprocessing methods was compared using partial least squares regression. Feature bands were identified via Pearson correlation analysis, and support vector regression models were established based on genetic algorithm (GA), particle swarm optimization (PSO), and grid search optimization. The results demonstrated that biochar application significantly increased the SPAD values of corn leaves (r = 0.879) and reduced the proportion of bioavailable Cd in soil, with the A7 treatment showing the most substantial decrease (30%). This indicates that biochar effectively mitigates Cd’s inhibitory effect on chlorophyll synthesis, with the alleviation effect enhancing as biochar application rates increased. Validation of the partial least squares regression model revealed that detrended spectra achieved optimal predictive performance (R²c = 0.94, RMSEC = 0.82, R²p = 0.88, RMSEP = 1.15), leading to the development of three optimized support vector regression models: GA-SVR, PSO-SVR, and GS-SVR. The GA-SVR model with a sigmoid kernel demonstrated the best internal validation performance for predicting SPAD values in maize leaves (R²c = 0.95, RMSEC = 0.24; R²p = 0.75, RMSEP = 1.63). This study provides preliminary theoretical support and technical reference for rapid spectral detection of the physiological status of maize under biochar-mediated mitigation of cadmium stress. Full article

Soil salinization severely restricts ecosystem stability and the sustainable development of agricultural productivity. However, current understanding of the spectral–salinity quantitative relationships under the influence of surface cracking still remains limited. To address this gap, this study collected hyperspectral reflectance data (350–2500 nm) from salt-affected soil in both cracked and uncracked surface conditions across the Songnen Plain, and applied fractional-order differentiation (FOD) processing with orders ranging from 0 to 2 and a step size of 0.1. Based on this, 14 types of FOD spectral indices were constructed, incorporating one-dimensional (1D), two-dimensional (2D), and three-dimensional (3D) structures. For each spectral index, the optimal fractional order and corresponding band combinations were first selected through Pearson correlation analysis for pH and EC under both surface conditions; subsequently, feature selection was performed using XGBoost-SHAP explainable analysis among the 14 optimal indices across different dimensions. Furthermore, the predictive performance of four modeling methods, including partial least squares regression (PLSR), Gaussian process regression (GPR), support vector regression (SVR), and random forest regression (RFR), was evaluated. The results showed that FOD transformations significantly enhanced correlations with EC and pH compared to raw reflectance. All prediction models demonstrated higher prediction accuracy under cracked surface conditions than uncracked surface conditions, indicating that desiccation cracks positively modulate spectral signals to enhance salinity information expression. Across different surface states, model performance generally followed the ranking: PLSR > GPR > SVR > RFR, with PLSR achieving the best predictions for EC and pH under cracked surfaces (R² of 0.88 and 0.76, RMSE of 0.29 dS/m and 0.35). This study not only deepens the understanding of fractional-order spectral response mechanisms in saline–alkali soils but also provides methodological support for regional monitoring of soil salinization. Full article

The Al-Ahsa Oasis region is an important agricultural area; however, continuous spatial–temporal monitoring is essential to assess and mitigate the impacts of climate change and land use change. The current study examines environmental and land cover changes in the Al-Ahsa Oasis region from 1990 to 2025 by utilizing spectral indices derived from multiple satellites. Multi-temporal Landsat imagery (Landsat 5, 8, and 9) was processed in Google Earth Engine (GEE) to derive key biophysical indicators, including the Normalized Difference Vegetation Index (NDVI), land surface temperature (LST), and bare soil index (BSI). Supervised classification techniques were employed to generate LULC maps for each time step, enabling the assessment of spatiotemporal land cover dynamics. In addition, a random forest (RF) machine learning algorithm was applied to accurately quantify and map the distribution of palm trees across the study area. The results showed that NDVI values fluctuated between −0.19 and 0.75 during the period from 1990 to 2025. Higher vegetation density was observed in central and eastern areas, with maximum values of −0.44–0.75 in 2025. The higher LST was observed in 2025, with a range of 34.7 to 54.6 °C, and the lower LST was observed in 1990 with a range 28.7 to 48.34 °C. BSI values decreased from −0.40 to 0.46 between 1990 and 2025 to a more variable range of −0.27 to 0.36, indicating reduced soil exposure. The classification of LULC numerical data shows a rapid rise in urban development of 67.19% and a 25% decrease in vegetation area. Furthermore, the results of the RF model indicate that palm tree area increased by 16.23% from 1990 to 2025, with overall accuracy of 98.15, and kappa coefficient of 0.962. This research highlights that urban expansion impacts environmental indicators such as LST, while the increasing trend of NDVI could support the palm trees expansion. This study finds valuable information for policymakers and land use planners to develop sustainable urban growth strategies, protect agricultural lands, and enhance oasis ecosystem resilience. Combined remote-sensing-based monitoring into regional planning frameworks can inform decision making for balancing urban development, environmental protection, and long-term agricultural sustainability in the Al-Ahsa Oasis. Full article

Coastal Louisiana continues to experience rapid wetland loss, increasing the exposure of marsh-creation containment dikes to storm-driven waves, erosion, and sediment loss. This study evaluated offshore-to-nearshore wave transformation, erosion risk reduction, wave runup, and hydrodynamic loading at a representative marsh-creation site in Plaquemines Parish, Louisiana. A 25-year return-period offshore wave condition was derived from long-term Wave Information Study hindcast data and propagated using the SWAN spectral wave model. Two idealized foreshore conditions were examined: a bare-bed case and a marsh-roughened shallow water case represented through enhanced bottom friction. Web Soil Survey data were used to characterize the local soil context of the containment-dike zone. The results show strong wave attenuation across the inner shelf and marsh platform. Relative to the bare-bed case, marsh roughness reduced dike toe significant wave height by 16.1–27.4% and decreased the Hs²-based erosion exposure proxy by 29.6–47.4% across three still-water levels. These reductions produced 15.4–26.4% lower 2% exceedance runup and 28.5–45.8% lower quasi-hydrostatic loading on the containment dike. The results indicate that marsh-induced dissipation can help reduce erosion potential and support sustainable coastal restoration infrastructure management. Full article

Populus pruinosa Schrenk is a keystone species in arid riparian ecosystems, where its physiological status is critical for biodiversity and soil stabilization. In this study, spectral reflectance, leaf chlorophyll density (CHD), and leaf water content (LWC) were measured for Populus pruinosa in the Tarim River headwater region and Awati County, Xinjiang, from July to October 2023. The aim was to estimate CHD using hyperspectral data combined with machine learning and to evaluate the effect of LWC on model accuracy. Raw spectra were preprocessed using Savitzky–Golay (SG) smoothing and continuous wavelet transform (CWT). A two-step feature selection strategy comprising Random Frog and iterative retaining informative variables (IRIV) was applied to extract characteristic bands. Three machine learning models—support vector machine (SVM), random forest (RF), and extreme gradient boosting (XGBoost)—were developed for CHD estimation with and without LWC as an additional input. Incorporating LWC consistently improved the predictive performance of all models. Without LWC, the RF model achieved the best accuracy (training R² = 0.842, test R² = 0.830), whereas after LWC integration, XGBoost reached the optimal performance (training R² = 0.871, test R² = 0.865). SHAP analysis identified the 687 nm wavelength and its interaction with LWC as the most important predictors. These results indicate that combining spectral information with LWC effectively improves the accuracy and stability of CHD estimation for Populus pruinosa, providing a reliable non-destructive approach for assessing forest ecosystem physiological status—a key contribution to the sustainable management of arid riparian forests. Full article

Accurate soil salinization monitoring in arid oases is crucial for agricultural sustainability and ecological security. However, existing deep learning-based approaches often suffer from insufficient use of multi-scale information and inadequate modeling of feature interactions, limiting their accuracy for retrieving complex salinity patterns. To address these limitations, we propose a multi-scale attention-based optimized hybrid deep learning model that integrates multi-scale 1D convolutional neural networks (1D-CNN), bidirectional gated recurrent units (Bi-GRU), and Transformer mechanisms (termed SMS–1D-CNN–Bi-GRU–Transformer). In this study, “scale” refers to the receptive-field scale formed by different 1D convolutional kernel sizes. The model employs a multi-scale feature extraction module to capture remote sensing signals across different scales, a multi-scale attention mechanism to adaptively weight the most informative features, and a Bi-GRU–Transformer module to explore complex sequential and global feature relationships. The proposed framework is applied to an oasis irrigation zone in Weili County, Xinjiang, using hyperspectral data from the ZY-1E satellite, topographic indices, and spectral-derived variables. The proposed method outperforms conventional 1D-CNN, GRU–Transformer, and other benchmark models on the test set—showing improvements of 2.8% in the coefficient of determination (0.952) and 18.9% in the root mean square error (0.867 g·kg⁻¹), demonstrating practical utility for precision land management and salinity monitoring in vulnerable irrigated ecosystems. Full article

The spatial heterogeneity and dynamic changes in soil organic matter (SOM) are key indicators for assessing cultivated land quality and the carbon cycle. Currently, large-scale SOM monitoring relies primarily on limited ground sampling, making it difficult to capture continuous spatiotemporal variation patterns. Taking Anhui Province, China as the study area, this research integrates multi-source remote sensing and geostatistical methods to construct a multi-source collaborative SOM inversion model and analyze its spatiotemporal evolution patterns, thereby achieving high-precision, continuous spatiotemporal monitoring of SOM. A total of 3026 sampling points in Huangshan, Chuzhou and Fuyang cities in Anhui Province were selected as model training samples. The study divided the terrain into three elevation zones (<20 m, 20–40 m, >40 m) and employed the Synthetic Minority Oversampling Technique (SMOTE) method to optimize sample distribution. Based on MODIS data, this study screened spectral bands and key phenological periods significantly correlated with SOM. By integrating spectral information from Landsat 8/9 OLI imagery, meteorological data and topographic factors, a random forest (RF) inversion model incorporating multi-source environmental variables was constructed. The results indicate that (1) the RF-based SOM inversion model exhibits moderate predictive accuracy acceptable for regional-scale SOM mapping, with a coefficient of determination (R²) of 0.55 and a root-mean-square error (RMSE) of 3.3 g/kg, effectively enabling the quantitative estimation of SOM at a regional scale. (2) The model’s inversion results reflect the spatial distribution of SOM in cultivated land in Anhui Province for the years 2019, 2022 and 2024. The provincial average SOM value shows an upward trend, with SOM content exhibiting a pattern of higher levels in the south and lower levels in the north, higher levels in the west and lower levels in the east, as well as a tendency to cluster. (3) Analysis using GeoDetector indicates that topography and precipitation are the primary drivers influencing SOM distribution, and the interaction between these two factors provides significantly greater explanatory power for SOM distribution than either factor alone. Through the integration of multi-source remote sensing data and model optimization, this study has validated the feasibility of multi-scale remote sensing-based SOM inversion, revealed the spatial differentiation characteristics and driving mechanisms of SOM in Anhui Province’s cultivated land, and provided a scientific basis for improving cultivated land quality and soil carbon sink management. Full article

Built-up area extraction is important for monitoring urban development and land-use change. Index-based methods are widely used for extracting built-up areas from Landsat imagery because of their simplicity and efficiency. However, conventional built-up indices often enhance bare land together with built-up areas due to their similar spectral characteristics, which reduces extraction accuracy and limits automatic threshold selection. To address this problem, this study proposes a built-up area extraction method based on multi-index synthesis and principal component analysis (PCA). First, NDBI (Normalization Differential Building Index), SAVI (Soil-Adjusted Vegetation Index), MNDWI (Modified Normalized Difference Water Index), and the brightness, greenness, and wetness components of the Tasseled Cap transformation were stacked to construct a six-band synthetic index image, enhancing the contrast among built-up areas, bare land, vegetation, and water bodies. PCA was then applied to the synthetic image using both correlation and covariance matrices, and the second principal component was used to enhance built-up area information. The resulting CorPC2 and CovPC2 methods were evaluated and compared with conventional built-up indices. The results showed that both PC2-based methods improved the separability between built-up areas and background features, while CovPC2 achieved the best performance by more effectively suppressing bare-land interference without requiring an additional bare-land mask. In the main experimental area, CovPC2 achieved higher accuracy than the comparison methods, and its Otsu-based result remained close to the optimal-threshold result. Validation in three typical cities further demonstrated the applicability of the proposed method across different Landsat sensors and urban environments. The proposed PC2-based method, particularly CovPC2, provides an effective and more automated approach for Landsat-based built-up area extraction under bare-land interference. Additionally, by using a threshold optimizing algorithm, built-up areas can be automatically extracted with high accuracy. Full article

Accurate estimation of soil total nitrogen (TN) and soil organic matter (SOM) is important for sustainable soil fertility assessment and precision nutrient management. Visible–near-infrared hyperspectral sensing provides a rapid and non-destructive solution, but its inversion accuracy is strongly affected by spectral preprocessing, especially under small-sample conditions. To reduce dependence on a manually selected preprocessing operator, this study proposes a homologous preprocessing representation fusion framework based on greedy concatenation (HPRF–GC). The framework constructs multiple homologous spectral views from the same raw spectrum, selects informative views through cross-validation-guided greedy forward selection, and concatenates the selected views before random forest or support vector regression. A self-built in situ hyperspectral dataset was collected from two representative black calcareous Mollisol farms in Heilongjiang Province, China, including 200 composite samples measured with a GaiaField Pro V10 imager at 5 m height under midday illumination using white reference calibration. On this dataset, HPRF–GC reduced RMSE by 3.61% for TN–RF, 9.94% for TN–SVR, 0.87% for SOM–RF, and 7.15% for SOM–SVR compared with the strongest single-preprocessing baseline, while introducing only a modest training-time overhead. On the public LUCAS 2015 dataset, HPRF–GC achieved competitive TN prediction performance, with an $R^2$ of 0.890 and an RMSE of 1.191 under RF. These results indicate that HPRF–GC provides a lightweight, interpretable and reproducible strategy for reducing preprocessing selection sensitivity in small-sample soil hyperspectral inversion. Full article

Rapid and cost-effective estimation of soil total nitrogen (TN) is essential for soil fertility assessment and nutrient management. However, the performance of laboratory visible–near-infrared (Vis–NIR) models is shaped not only by preprocessing and modeling strategy but also by sample preparation and the soil’s compositional background. In this study, TN prediction was evaluated using 376 topsoil samples from two contrasting datasets: Mollisols from the black-soil region of Northeast China and Ultisols from Qiyang County, Hunan Province, southern China. Spectra acquired over 350–2500 nm for three particle-size fractions were preprocessed using Savitzky–Golay smoothing combined with standard normal variate (SNV), first-derivative, or second-derivative transformations, and modeled using partial least squares regression (PLSR), support vector regression (SVR), and extreme gradient boosting (XGBoost). Model development used a 5 × 5 nested cross-validation followed by evaluation on a sample-grouped held-out test set. Among all combinations, XGBoost with first-derivative preprocessing on the 0.25 mm fraction produced the best performance, with test R² values of 0.91 for Mollisol and 0.78 for Ultisol. Shapley additive explanations (SHAP) and principal component analysis (PCA) consistently identified informative spectral regions at 430–480 and 1330–1450 nm for Mollisol and at 585–635, 820–900, and 2180–2240 nm for Ultisol. Prediction errors were larger in the sampled Ultisol dataset and increased with DCB-extractable Fe and mineral backgrounds. A second-stage log-domain residual correction incorporating ancillary soil properties further reduced the Ultisol RMSE from 0.30 to 0.27 g kg⁻¹. These findings support the 0.25 mm, first-derivative, XGBoost workflow as a robust laboratory Vis–NIR approach for TN prediction and indicate that composition-aware residual correction can improve prediction in oxide- and mineral-rich soils. Full article