Digital soil mapping of heavy metals using multiple geospatial data: Feature identification and deep neural network

Ecological Indicators, Год журнала: 2023, Номер 154, С. 110863 - 110863

Опубликована: Сен. 2, 2023

Monitoring the spatial distribution and sources of heavy metals (HM) in soil is essential for avoiding health risks achieving sustainable utilization. Multiple geospatial data, including remote sensing, climate, topography were used to extract environmental covariates. Additionally, scene was employed as alternative data land use/land cover describe urban functions human activity intensity more detail. After converting a uniform resolution 30 m, these covariates adopted characterize several common HM soil, copper (Cu), chromium (Cr), lead (Pb), nickel (Ni), zinc (Zn). The RReliefF algorithm identify important variables. quantification models established using back-propagation neural network (BPNN) deep (DNN). Besides, impact distance from scenes on analyzed. result demonstrated that key covariate estimating soil. Compared with BPNN, DNN model provided better accuracy (R2 = 0.67–0.75) estimation five elements. Therefore, map concentrations at grid scale m. highest risk pollution are industrial areas, residential road, commercial concentration negatively correlated scenes. effective distances areas about 2000 road 500

Язык: Английский

---

Continental-scale mapping of soil pH with SAR-optical fusion based on long-term earth observation data in google earth engine

Ecological Indicators, Год журнала: 2024, Номер 165, С. 112246 - 112246

Опубликована: Июнь 14, 2024

Язык: Английский

---

Exploiting Soil and Remote Sensing Data Archives for 3D Mapping of Multiple Soil Properties at the Swiss National Scale

Remote Sensing, Год журнала: 2024, Номер 16(15), С. 2712 - 2712

Опубликована: Июль 24, 2024

Soils play a central role in ecosystem functioning, and thus, mapped soil property information is indispensable to supporting sustainable land management. Digital Soil Mapping (DSM) provides framework spatially estimate properties. However, broad-scale DSM remains challenging because of non-purposively sampled data, large data volumes for processing extensive covariates, high model complexities due varying soil–landscape relationships. This study presents three-dimensional Switzerland, targeting the properties clay content (Clay), organic carbon (SOC), pH value (pH), potential cation exchange capacity (CECpot). The approach based on machine learning comprehensive exploitation remote sensing archives. Quantile Regression Forest was applied link sample from national base with covariates derived LiDAR-based elevation model, climate raster multispectral time series satellite imagery. covariate set comprises multiscale terrain attributes, patterns their temporal variation, temporarily use features, spectral bare signatures. predictions were evaluated respect different landcovers depth intervals. All reference sets found be clustered towards croplands, showing an increasing density lower upper According R2 independent overall accuracy amounts 0.69 Clay, 0.64 SOC, 0.76 pH, 0.72 CECpot. Reduced accuracies accompanied by limited sizes (e.g., CECpot), uneven statistical distributions SOC), low spatial densities woodland subsoils). Multiscale highly influential all models; particularly important Clay model; showed enhanced importance modeling pH; reflectance major driver SOC CECpot models.

Язык: Английский

---

Retrieving heavy metal concentrations in urban soil using satellite hyperspectral imagery

International Journal of Applied Earth Observation and Geoinformation, Год журнала: 2024, Номер 132, С. 104079 - 104079

Опубликована: Авг. 1, 2024

Efficient prediction and precise depiction of heavy metal concentrations in urban soil are essential for mitigating non-point source pollution safeguarding public health. Therefore, this research investigated the estimation derived from Gaofen-5 (GF-5) hyperspectral images calibrated by direct standardization (DS) algorithm. The inversion strategy response to two-dimensional spectral index (2D-SSI) was proposed coupling Pearson correlation coefficient (r) competitive adaptive reweighting algorithm (CARS) feature selection. results indicated that optimal models based on 2D-SSI outperform calibrated, filtered original bands. For Pb, Cu, Cd, Hg, model determination coefficients validation data set (RV2) were 0.871 (SVM), 0.883 (BPNN), 0.834 (PLSR), 0.907 respectively. features highlighted space, predicted distribution aligned with observed ground measurements. This study revealed DS-corrected GF-5 AHSI constructed can serve as a reliable technical approach prevention.

Язык: Английский

---

Mapping the soil C:N ratio at the European scale by combining multi-year Sentinel radar and optical data via cloud computing

Soil and Tillage Research, Год журнала: 2024, Номер 245, С. 106311 - 106311

Опубликована: Сен. 24, 2024

Язык: Английский

---