Predicting Soil erosion based on the machine learning in Loess Plateau of China 

2021 ◽  
Author(s):  
Chenlu Huang ◽  
Qinke Yang
Keyword(s):  
Machine Learning ◽  
Soil Erosion ◽  
Loess Plateau ◽  
Evaluation Method ◽  
Interpolation Method ◽  
Dynamic Change ◽  
Control Factor ◽  
Change Analysis ◽  
Slope Length ◽  
Entire Study

<p>Soil erosion is one of the global ecological and environmental problems, which is an important factor leading to land degradation. To scientifically and effectively control soil erosion, it’s necessary to improve soil erosion evaluation methods that can obtain the actual rates of soil erosion, rather than potential erosion. For this, about 300 sampling units deployed in the Loess Plateau used as the basic data in our study, combining the seven soil erosion factors (rainfall-runoff erosivity factor, soil erodibility factor, slope length and steepness factor, biological-control factor, engineering-control factor, tillage practices factor) involved in the CSLE model and 50 soil erosion covariates related to climate, soil, topography, vegetation, human activities, etc. Using machine learning methods to establish an optimal model, and spatially predict the soil erosion rate and make a soil erosion mapof the entire study area. The prediction results show that the explanation degree of the random forest spatial prediction model is 73%. Among the selected optimal characteristic parameters, terrain and vegetation-related variables are the most important factors affecting soil erosion, from high to low, the order is LS > B > NDVI (May to September). Compared to previous studies with USLE/RUSLE/CSLE and GIS integrated mapping methods, or sampling survey based interpolation method, improvements in this paper can be concluded to : (1) the use of machine learning instead of simple multiply by soil erosion factors (linear regression), (2) higher resolution interpretation results supported by the project of “Pan-Third Pole Project”, which provide soil erosion that closed to the actual rates of soil erosion. (3) considerate additional related covariates such as population density, precipitation, soil conservation measures and so on. Further development of soil erosion prediction could provide a more accurate soil erosion evaluation method. This method can not only monitor and evaluate soil erosion in real time, and provide the possibility for the dynamic change analysis? of soil erosion in the future, but also help decision makers take effective measures in the process of mitigating soil erosion risk.</p>

A new RUSLE slope length factor and its application to soil erosion assessment in a Loess Plateau watershed

2018 ◽  
Vol 182 ◽  
pp. 10-24 ◽  
Author(s):  
Wei Qin ◽  
Qiankun Guo ◽  
Wenhong Cao ◽  
Zhe Yin ◽  
Qinghong Yan ◽  
...  
Keyword(s):  
Soil Erosion ◽  
Loess Plateau ◽  
Slope Length ◽  
Slope Length Factor ◽  
Soil Erosion Assessment

scholarly journals Soil Erosion and Controls in the Slope-Gully System of the Loess Plateau of China: A Review

2021 ◽  
Vol 9 ◽  
Author(s):  
Bingbing Zhu ◽  
Zhengchao Zhou ◽  
Zhanbin Li
Keyword(s):  
Soil Erosion ◽  
Loess Plateau ◽  
Overland Flow ◽  
Vegetation Pattern ◽  
Sediment Source ◽  
The Loess Plateau ◽  
Erosion Process ◽  
Slope Length ◽  
Base Level ◽  
Erosion Processes

The Loess Plateau has long been suffering from serious soil erosion of which erosion from the slope-gully system is now dominant. The slope-gully system is characterized with distinctive erosion distribution zones consisting of inner and inter gully areas wherein erosion patterns spatially vary, acting as both sediment source and the dominant sediment and water transport mechanism. In this paper, a substantial body of research is reviewed concentrating on the soil erosion processes and control practices in the slope-gully system. The inner gully area is identified as the main sediment source while runoff and sediment from the inter-gully upland is found to significantly affect down slope erosion processes. Correspondingly, the protective vegetation pattern and coverage should be strategically designed for different erosion zones with an emphasis on the critical vegetation cover and pattern to reduce sediment yield of the whole slope-gully system. Check-dam could change the base level of erosion and reduce the slope length of the gully side, which will further decrease the possibility and magnitude of gravity erosion. We concluded that understanding the erosion processes and implementing erosion practices for the slope-gully system are of importance and require more research efforts that emphasize: 1) the influence of upland runoff on erosion processes at downslope; 2) the relationship between hydraulic characteristics of overland flow and erosion process at a slope-gully system scale; 3) physical mechanisms of different vegetation patterns on the slope-gully erosion process.

scholarly journals Evaluation of Three Different Machine Learning Methods for Object-Based Artificial Terrace Mapping—A Case Study of the Loess Plateau, China

2021 ◽  
Vol 13 (5) ◽  
pp. 1021
Author(s):  
Hu Ding ◽  
Jiaming Na ◽  
Shangjing Jiang ◽  
Jie Zhu ◽  
Kai Liu ◽  
...  
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Loess Plateau ◽  
Water Conservation ◽  
Nearest Neighbor ◽  
Gradient Boosting ◽  
K Nearest Neighbor ◽  
The Loess Plateau ◽  
Object Based ◽  
Extreme Gradient Boosting

Artificial terraces are of great importance for agricultural production and soil and water conservation. Automatic high-accuracy mapping of artificial terraces is the basis of monitoring and related studies. Previous research achieved artificial terrace mapping based on high-resolution digital elevation models (DEMs) or imagery. As a result of the importance of the contextual information for terrace mapping, object-based image analysis (OBIA) combined with machine learning (ML) technologies are widely used. However, the selection of an appropriate classifier is of great importance for the terrace mapping task. In this study, the performance of an integrated framework using OBIA and ML for terrace mapping was tested. A catchment, Zhifanggou, in the Loess Plateau, China, was used as the study area. First, optimized image segmentation was conducted. Then, features from the DEMs and imagery were extracted, and the correlations between the features were analyzed and ranked for classification. Finally, three different commonly-used ML classifiers, namely, extreme gradient boosting (XGBoost), random forest (RF), and k-nearest neighbor (KNN), were used for terrace mapping. The comparison with the ground truth, as delineated by field survey, indicated that random forest performed best, with a 95.60% overall accuracy (followed by 94.16% and 92.33% for XGBoost and KNN, respectively). The influence of class imbalance and feature selection is discussed. This work provides a credible framework for mapping artificial terraces.

scholarly journals Effect of Forest Management Operations on Aggregate-Associated SOC Dynamics Using a 137Cs Tracing Method

Forests ◽  
2021 ◽  
Vol 12 (7) ◽  
pp. 859
Author(s):  
Geng Guo ◽  
Xiao Li ◽  
Xi Zhu ◽  
Yanyin Xu ◽  
Qiao Dai ◽  
...  
Keyword(s):  
Forest Management ◽  
Soil Erosion ◽  
Management Practices ◽  
Negative Impact ◽  
Dynamic Change ◽  
Aggregate Size ◽  
Nutrient Loss ◽  
Vegetation Coverage ◽  
Forest Conversion ◽  
Mountain Areas

Although forest conversions have long been a focus in carbon (C) research, the relationship between soil erosion and the dynamic change of soil organic carbon (SOC) has not been well-quantified. The objective of this study was to investigate the effects of converting CBF (coniferous and broad-leaved mixed forests) to economic forests, including CF (chestnut forest), HF (hawthorn forest), and AF (apple forest), on the soil structure and nutrient loss in the Huaibei Rocky Mountain Areas, China. A 137Cs tracer method was used to provide soil erosion data in order to quantify the loss of aggregate-associated SOC. The results showed that forest management operations caused macro-aggregates to decrease by 1.69% in CF, 4.52% in AF, and 3.87% in HF. Therefore, the stability of aggregates was reduced. The SOC contents in each aggregate size decreased significantly after forest conversion, with the largest decreases occurring in AF. We quantified the loss of 0.15, 0.38, and 0.31 Mg hm−2 of aggregate-associated SOC after conversion from CBF to CF, AF, and HF, respectively. These results suggest that forest management operations have a negative impact on soil quality and fertility. CF has better vegetation coverage and less human interference, making it more prominent among the three economic forests species. Therefore, when developing forest management operations, judicious selection of tree varieties and appropriate management practices are extremely critical. In addition, measures should be taken to increase surface cover to reduce soil erosion and achieve sustainable development of economic forests.

scholarly journals Machine Learning Empowered Trust Evaluation Method for IoT Devices

IEEE Access ◽  
2021 ◽  
Vol 9 ◽  
pp. 65066-65077
Author(s):  
Wei Ma ◽  
Xing Wang ◽  
Mingsheng Hu ◽  
Qinglei Zhou
Keyword(s):  
Machine Learning ◽  
Evaluation Method ◽  
Trust Evaluation ◽  
Iot Devices

scholarly journals Effects of Vegetation Restoration on Soil Erosion on the Loess Plateau: A Case Study in the Ansai Watershed

2021 ◽  
Vol 18 (12) ◽  
pp. 6266
Author(s):  
Hui Wei ◽  
Wenwu Zhao ◽  
Han Wang
Keyword(s):  
Land Use ◽  
Soil Erosion ◽  
Loess Plateau ◽  
Water Conservation ◽  
Large Scale ◽  
Land Use Changes ◽  
Vegetation Restoration ◽  
The Loess Plateau ◽  
Local Soil ◽  
Erosion Environment

Large-scale vegetation restoration greatly changed the soil erosion environment in the Loess Plateau since the implementation of the “Grain for Green Project” (GGP) in 1999. Evaluating the effects of vegetation restoration on soil erosion is significant to local soil and water conservation and vegetation construction. Taking the Ansai Watershed as the case area, this study calculated the soil erosion modulus from 2000 to 2015 under the initial and current scenarios of vegetation restoration, using the Chinese Soil Loess Equation (CSLE), based on rainfall and soil data, remote sensing images and socio-economic data. The effect of vegetation restoration on soil erosion was evaluated by comparing the average annual soil erosion modulus under two scenarios among 16 years. The results showed: (1) vegetation restoration significantly changed the local land use, characterized by the conversion of farmland to grassland, arboreal land, and shrub land. From 2000 to 2015, the area of arboreal land, shrub land, and grassland increased from 19.46 km2, 19.43 km2, and 719.49 km2 to 99.26 km2, 75.97 km2, and 1084.24 km2; while the farmland area decreased from 547.90 km2 to 34.35 km2; (2) the average annual soil erosion modulus from 2000 to 2015 under the initial and current scenarios of vegetation restoration was 114.44 t/(hm²·a) and 78.42 t/(hm²·a), respectively, with an average annual reduction of 4.81 × 106 t of soil erosion amount thanks to the vegetation restoration; (3) the dominant soil erosion intensity changed from “severe and light erosion” to “moderate and light erosion”, vegetation restoration greatly improved the soil erosion environment in the study area; (4) areas with increased erosion and decreased erosion were alternately distributed, accounting for 48% and 52% of the total land area, and mainly distributed in the northwest and southeast of the watershed, respectively. Irrational land use changes in local areas (such as the conversion of farmland and grassland into construction land, etc.) and the ineffective implementation of vegetation restoration are the main reasons leading to the existence of areas with increased erosion.

scholarly journals Comparison of Ensemble Machine Learning Methods for Soil Erosion Pin Measurements

2021 ◽  
Vol 10 (1) ◽  
pp. 42
Author(s):  
Kieu Anh Nguyen ◽  
Walter Chen ◽  
Bor-Shiun Lin ◽  
Uma Seeboonruang
Keyword(s):  
Machine Learning ◽  
Soil Erosion ◽  
Ensemble Methods ◽  
Machine Learning Algorithms ◽  
Multivariate Adaptive Regression Splines ◽  
Gradient Boosting ◽  
Support Vector ◽  
Ensemble Machine Learning ◽  
Boosting Method ◽  
Bagging Method

Although machine learning has been extensively used in various fields, it has only recently been applied to soil erosion pin modeling. To improve upon previous methods of quantifying soil erosion based on erosion pin measurements, this study explored the possible application of ensemble machine learning algorithms to the Shihmen Reservoir watershed in northern Taiwan. Three categories of ensemble methods were considered in this study: (a) Bagging, (b) boosting, and (c) stacking. The bagging method in this study refers to bagged multivariate adaptive regression splines (bagged MARS) and random forest (RF), and the boosting method includes Cubist and gradient boosting machine (GBM). Finally, the stacking method is an ensemble method that uses a meta-model to combine the predictions of base models. This study used RF and GBM as the meta-models, decision tree, linear regression, artificial neural network, and support vector machine as the base models. The dataset used in this study was sampled using stratified random sampling to achieve a 70/30 split for the training and test data, and the process was repeated three times. The performance of six ensemble methods in three categories was analyzed based on the average of three attempts. It was found that GBM performed the best among the ensemble models with the lowest root-mean-square error (RMSE = 1.72 mm/year), the highest Nash-Sutcliffe efficiency (NSE = 0.54), and the highest index of agreement (d = 0.81). This result was confirmed by the spatial comparison of the absolute differences (errors) between model predictions and observations using GBM and RF in the study area. In summary, the results show that as a group, the bagging method and the boosting method performed equally well, and the stacking method was third for the erosion pin dataset considered in this study.

scholarly journals The application of137Cs and210Pbexmethods in soil erosion research of Titel loess plateau, Vojvodina, Northern Serbia

2020 ◽  
Vol 12 (1) ◽  
pp. 11-24
Author(s):  
Kristina S. Kalkan ◽  
Sofija Forkapić ◽  
Slobodan B. Marković ◽  
Kristina Bikit ◽  
Milivoj B. Gavrilov ◽  
...  
Keyword(s):  
Soil Erosion ◽  
Loess Plateau ◽  
Alternative Methods ◽  
Delivery Ratio ◽  
Undisturbed Soil ◽  
Migration Model ◽  
Erosion Rates ◽  
Soil Erosion Rates ◽  
Profile Distribution ◽  
And Migration

AbstractSoil erosion is one of the largest global problems of environmental protection and sustainable development, causing serious land degradation and environmental deterioration. The need for fast and accurate soil rate assessment of erosion and deposition favors the application of alternative methods based on the radionuclide measurement technique contrary to long-term conventional methods. In this paper, we used gamma spectrometry measurements of 137Cs and unsupported 210Pbex in order to quantify the erosion on the Titel Loess Plateau near the Tisa (Tisza) River in the Vojvodina province of Serbia. Along the slope of the study area and in the immediate vicinity eight representative soil depth profiles were taken and the radioactivity content in 1 cm thick soil layers was analyzed. Soil erosion rates were estimated according to the profile distribution model and the diffusion and migration model for undisturbed soil. The net soil erosion rates, estimated by 137Cs method range from −2.3 t ha−1 yr−1 to −2.7 t ha−1 yr−1, related to the used conversion model which is comparable to published results of similar studies of soil erosion in the region. Vertical distribution of natural radionuclides in soil profiles was also discussed and compared with the profile distribution of unsupported 210Pbex measurements. The use of diffusion and migration model to convert the results of 210Pbex activities to soil redistribution rates indicates a slightly higher net erosion of −3.7 t ha−1 yr−1 with 98% of the sediment delivery ratio.

scholarly journals A Semi-Automated Object-Based Gully Networks Detection Using Different Machine Learning Models: A Case Study of Bowen Catchment, Queensland, Australia

Sensors ◽  
2019 ◽  
Vol 19 (22) ◽  
pp. 4893 ◽  
Author(s):  
Hejar Shahabi ◽  
Ben Jarihani ◽  
Sepideh Tavakkoli Piralilou ◽  
David Chittleborough ◽  
Mohammadtaghi Avand ◽  
...  
Keyword(s):  
Machine Learning ◽  
Image Segmentation ◽  
Vegetation Index ◽  
Scale Parameter ◽  
Slope Aspect ◽  
Support Vector ◽  
Topographic Wetness Index ◽  
Slope Length ◽  
Optimal Scale ◽  
Object Based

Gully erosion is a dominant source of sediment and particulates to the Great Barrier Reef (GBR) World Heritage area. We selected the Bowen catchment, a tributary of the Burdekin Basin, as our area of study; the region is associated with a high density of gully networks. We aimed to use a semi-automated object-based gully networks detection process using a combination of multi-source and multi-scale remote sensing and ground-based data. An advanced approach was employed by integrating geographic object-based image analysis (GEOBIA) with current machine learning (ML) models. These included artificial neural networks (ANN), support vector machines (SVM), and random forests (RF), and an ensemble ML model of stacking to deal with the spatial scaling problem in gully networks detection. Spectral indices such as the normalized difference vegetation index (NDVI) and topographic conditioning factors, such as elevation, slope, aspect, topographic wetness index (TWI), slope length (SL), and curvature, were generated from Sentinel 2A images and the ALOS 12-m digital elevation model (DEM), respectively. For image segmentation, the ESP2 tool was used to obtain three optimal scale factors. On using object pureness index (OPI), object matching index (OMI), and object fitness index (OFI), the accuracy of each scale in image segmentation was evaluated. The scale parameter of 45 with OFI of 0.94, which is a combination of OPI and OMI indices, proved to be the optimal scale parameter for image segmentation. Furthermore, segmented objects based on scale 45 were overlaid with 70% and 30% of a prepared gully inventory map to select the ML models’ training and testing objects, respectively. The quantitative accuracy assessment methods of Precision, Recall, and an F1 measure were used to evaluate the model’s performance. Integration of GEOBIA with the stacking model using a scale of 45 resulted in the highest accuracy in detection of gully networks with an F1 measure value of 0.89. Here, we conclude that the adoption of optimal scale object definition in the GEOBIA and application of the ensemble stacking of ML models resulted in higher accuracy in the detection of gully networks.

Sign in / Sign up

Export Citation Format

Share Document