Search

Soil organic carbon (SOC) and total nitrogen (TN) influence physical, chemical and biological properties and process in soils that determine soil fertility and enhance or maintain agricultural productivity and food security. Knowledge on the spatial distribution of SOC and TN is necessary for sustainable soil resource management. In this study, we applied boosted regression trees (BRT) model to predict and map the spatial distribution of SOC and TN in the northeastern coastal areas of China, and C:N ratio map was generated from the predicted maps of SOC and TN. A total of 149 topsoil (0–20 cm) samples and 12 environmental variables that included topography, landuse and remote sensing indices, and climatic variables were selected. The performance of the models was evaluated based on cross-validation using mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R 2 ). The BRT model was run for 100 iterations where the average of 100 SOC and TN maps was considered as final predicted maps and its standard deviation as prediction uncertainty. The BRT model showed a good predictive performance due to its higher R 2 and lower MAE, and RMSE indices. The model explained approximately 64% and 56% of the total SOC and TN variability. Topographic variables showed a maximum influence in the prediction of SOC and TN, followed by vegetation and climate. The SOC and TN contents were reduced from the northwest towards the southeast of the study area; average predicted SOC and TN contents were 15.4 g kg −1 and 1.01 g kg −1 , respectively. The spatial distribution of C:N ratio was closely related to landuse types in a decreasing order: woodland > orchard > cultivated land > grassland. We found topographic variables as main environmental indicators of SOC and TN distribution, and recommend to include them in future SOC and TN mapping studies in the coastal agroecosystems in China. [ABSTRACT FROM AUTHOR]

In digital soil mapping (DSM), soil maps are usually produced in a univariate manner, that is, each soil map is produced independently and therefore, when multiple soil properties are mapped the underlying dependence structure between these soil properties is ignored. This may lead to inconsistent predictions and simulations. For example, soil organic carbon (SOC) and total nitrogen (TN) maps produced independently may show unrealistic carbon–nitrogen (C:N) ratios. In the last decade the production of soil maps with machine learning models has become increasingly popular as these models are able to capture complex non-linear relationships between soil properties and environmental covariates. However, producing soil maps with multivariate machine learning models is still lacking and requires much investigation in DSM. In this paper we present the combined modelling of multiple soil properties with a multivariate random forest (MRF) model. We applied this model to mapping SOC and TN, and we compared it with results of two separate univariate random forest (RF) models. The comparison was done by means of stochastic simulations determined by sampling from the conditional distributions of the soil properties, given the covariates, as estimated by quantile regression forest. The results show that the MRF model is superior in terms of maintaining the dependence structure between SOC and TN, and consequently, is also able to produce more realistic C:N ratios. The models were also compared on the basis of prediction accuracy using commonly used accuracy metrics such as the root mean square error (RMSE). We found that the accuracy of the MRF model (RMSE-SOC = 40. 04 , RMSE-TN = 2. 26 , RMSE-CN = 3. 58) is comparable to that of the univariate RF models (RMSE-SOC = 39. 76 , RMSE-TN = 2. 26 , RMSE-CN = 3. 65). We performed the same comparisons between a regression co-kriging model and two separate regression kriging models, and made similar conclusions. • Multivariate RF accounts for the correlation structure between soil properties. • Simulations produced with multivariate RF are superior compared to those of RF. • In digital soil assessment it is vital to account for correlations in soil properties. • Multivariate RF C:N ratio simulations outperformed those of regression co-kriging. • Prediction accuracy of multivariate RF is comparable with that of RF. [ABSTRACT FROM AUTHOR]

Soil organic carbon (SOC) and soil carbon-to-nitrogen ratio (C:N) are the main indicators of soil quality and health and play an important role in maintaining soil quality. Together with Landsat, the improved spatial and temporal resolution Sentinel sensors provide the potential to investigate soil information on various scales. We analyzed and compared the potential of satellite sensors (Landsat-8, Sentinel-2 and Sentinel-3) with various spatial and temporal resolutions to predict SOC content and C:N ratio in Switzerland. Modeling was carried out at four spatial resolutions (800 m, 400 m, 100 m and 20 m) using three machine learning techniques: support vector machine (SVM), boosted regression tree (BRT) and random forest (RF). Soil prediction models were generated in these three machine learners in which 150 soil samples and different combinations of environmental data (topography, climate and satellite imagery) were used as inputs. The prediction results were evaluated by cross-validation. Our results revealed that the model type, modeling resolution and sensor selection greatly influenced outputs. By comparing satellite-based SOC models, the models built by Landsat-8 and Sentinel-2 performed the best and the worst, respectively. C:N ratio prediction models based on Landsat-8 and Sentinel-2 showed better results than Sentinel-3. However, the prediction models built by Sentinel-3 had competitive or better accuracy at coarse resolutions. The BRT models constructed by all available predictors at a resolution of 100 m obtained the best prediction accuracy of SOC content and C:N ratio; their relative improvements (in terms of R2) compared to models without remote sensing data input were 29.1% and 58.4%, respectively. The results of variable importance revealed that remote sensing variables were the best predictors for our soil prediction models. The predicted maps indicated that the higher SOC content was mainly distributed in the Alps, while the C:N ratio shared a similar distribution pattern with land use and had higher values in forest areas. This study provides useful indicators for a more effective modeling of soil properties on various scales based on satellite imagery. Unlabelled Image • The performance of Landsat-8, Sentinel-2 and 3 data in mapping SOC and C:N ratio was compared. • Our results revealed that the model type, modeling resolution and sensor selection greatly influenced outputs. • National scale SOC content and C:N ratio were mapped in Switzerland. • Remote sensing variables were the best predictors for our soil prediction models. [ABSTRACT FROM AUTHOR]