Soil Loss and Sedimentation Rates in a Sub-Catchment of the Yellow River Basin in China

doi:10.21203/rs.3.rs-2510450/v1

Download PDF

Research Article

Soil Loss and Sedimentation Rates in a Sub-Catchment of the Yellow River Basin in China

https://doi.org/10.21203/rs.3.rs-2510450/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 03 Dec, 2023

Read the published version in International Soil and Water Conservation Research →

Version 1

posted

You are reading this latest preprint version

Soil loss by water erosion is one of the main threats to soil health and food production in intensively used agricultural areas. To assess its significance to the overall sediment production we applied the Water and Tillage Erosion Model/Sediment Delivery model (WaTEM/SEDEM) to the Luoyugou catchment, a sub-catchment of the Yellow River basin within the Chinese Loess plateau. WaTEM/SEDEM considers rill and inter-rill erosion and deposition rates, resulting in sediment yield rates leaving the catchment. Although terraces were established in the 1990s to reduce soil loss, no further soil erosion modeling has been published for this area. Applying 1000 Monte Carlo simulations of the WaTEM/SEDEM, modeled average soil loss by rill and inter-rill erosion for 2020 was 12.4 ± 11.8 t ha^-1 yr^-1, with sediment yield at the outlet of 52949 ± 11215 t yr^-1. Terracing reduced erosion rates, while land cover changes, mainly conversion of forests and grassland partly counteracted the mitigation (combined effect: 50% reduction). Modeled sediment loads by rill and inter-rill erosion account for 17.1% of the total long-term sediment production recorded by flow discharge measurements. Other processes not considered by the model like landslides, gully erosion, riverbank erosion, and sediment production by construction seem to dominate overall sediment yield. Considering years with baseline sediment production only, measured and modeled sediment yield compared well, indicating that the latter processes contribute mainly during extreme events.

Geomorphology

Geographic Information Systems

Environmental Engineering

Agricultural Engineering

WaTEM/SEDEM

erosion modeling

soil erosion rates

sediment load

risk assessment

terraces

Land use and land management have tremendous effects on soils by triggering processes like soil detachment and dislocation (Dotterweich, 2013). Soil that is detached and mobilized through water erosion is usually transported downslope and deposited within the catchment or routed to the catchment outlet.

Soil can be mobilized as rill, inter-rill, gully, tillage erosion, or land sliding (Gilley, 2005). While gully erosion and land sliding might be successfully mapped with remote sensing tools even in arable areas without permanent vegetation cover (Vanmaercke et al., 2021), rill, inter-rill, and tillage erosion can only be assessed by measurement (only possible for small-scale) or modeling.

The most widely used soil erosion by water model is the Revised Universal Soil Loss Equation (RUSLE), which estimates the on-site soil loss rates by rill and inter-rill erosion (Alewell, Borrelli, Meusburger, & Panagos, 2019). RUSLE is based on a multiplication of six factors that assess the natural conditions (R: rainfall erosivity, K: soil erodibility, L: slope length, S: slope steepness) and the land use and land management specific variables (C: land cover and management, P: protection measures). For China, an adapted version of RUSLE, the Chinese Soil Loss Equation (CSLE, B. Liu, Zhang, & Xie, 2002) has been developed, where the C and the P factor were substituted with three factors considering biology practices for trees (B), engineering (E) and tillage practices (T). However, both modeling approaches do not consider the deposition of sediments within the modeled area or sediment delivery to the outlet of the modeled catchment and are thus not directly comparable to measured sediment yield at catchment outlets.

Studies in Europe have shown that a mean soil erosion rate of 4.62 t ha^-1 yr^-1 with a total estimated sediment yield by rill and inter-rill erosion of 116 Billion t yr^-1 is predicted for the European Union’s area (Borrelli et al., 2018). For China, such a national or regional (e.g. for the whole Loess Plateau) assessment of the ratio between erosion and deposition is not available and often impeded due to the high spatial heterogeneity of climate, soils, and topography in China. According to a global modeling assessment, China is one of the hot spots of soil erosion with about 6.3% of the area facing erosion rates of > 20 t ha^-1 yr^-1 (Borrelli et al., 2017). The fourth national soil erosion survey in 2011 for China estimates that 13.5% of the soil loss in China exceeds the soil loss tolerance (B. Liu et al., 2020).

A sub-catchment of the Yellow River, the Luoyugou catchment, was intensively studied by R. Wang et al. (2008) who observed an average soil erosion rate of 50 t ha^-1 yr^-1. That study showed that a single flood event could contribute to a fraction of 60% of the total annual sediment load. Furthermore, an experimental site of the Tian Shui Experimental Site of Soil and Water Conservation was established within the catchment in the 1940s resulting in long-time hydrological measurements (Yang, 2013). The Luoyugou catchment is ideally suited for soil erosion assessments as the changes in vegetation structural coverage and its influence on runoff (Wu, Liu, Xia, & Li, 2022) as well as the spatial variability in soil organic carbon (Xin, Qin, & Yu, 2016) have been investigated over years. Regarding large-scale soil erosion models in the area, research using subjective Bayesian probabilities and GIS was already conducted in 2011 (Shen, 2011) with a basic evaluation of the soil risk (probability of soil erosion occurrence) but no soil loss rate assessment was done. We applied the Water and Tillage Erosion Model/Sediment Delivery model (WaTEM/SEDEM), which combines the on-site sediment production strength with a downslope sediment routing, deposition, and/or transportation. Thus, soil loss rates by water as well as the rates and spatial patterns of sedimentation within the catchment including the total sediment yield by rill and inter-rill erosion that leaves the catchment can be assessed. WaTEM/SEDEM was inter alia applied catchment-wise in the Baiquan Country in northeastern China (Fang, 2017; Fang & Sun, 2017) and the Loess Plateau (Feng, Wang, Chen, Fu, & Bai, 2010; Y. Liu & Fu, 2016; Zhao et al., 2015). By applying the approach to the selected catchment, we aim to (i) better understand the erosion and sedimentation dynamics in the catchment, (ii) compare the model results to measured long-term sediment yields and thus assess the contribution of rill and inter-rill processes to the overall sediment production, and (iii) evaluate the change of erosion to deposition rates according to the land use change from 1986 to 2020.

2.1. Study area

The Luoyugou catchment (Fig. 1) is a sub-catchment of the Yellow River which is the second-longest river in China. That catchment covers an area of approximately 73 km² with a main stream channel length of 21.8 km and a width of approx. 3.4 km (Fig. 1). It is a headwater catchment of the Ji river that is a primary tributary of the Wei river, the largest tributary of the Yellow River. Mountains with an average slope of 18° dominate the catchment. The elevation ranges from 1194 m a.s.l. to 1897 m a.s.l. (mean: 1532 m a.s.l.). The Luoyugou catchment is situated at the border of the Longxi Loess Plateau hill-ravine region and the Longnan Mountains.

The climate is categorized as a continental monsoon climate. The multiannual average precipitation was 553 mm during the period 1986–2010, of which > 80% fell during the wet season from June until September (Rui Jie Qin, Li, Xiao, & An, 2015; R. J. Qin, Lu, & An, 2019). The annual average temperature was 10.7°C during the same period (Hui, Yongbo, Junzhi, & A-Xing, 2014). Dominant soil types are Calcaric Cambisols (for more than 90% of the catchment area), followed by Haplic Greyzems and Calceric Fluvisols (S. Liu & Li, 1989). Intensive agricultural activities in the mixed land use catchment with a focus on dryland agriculture have led to severe soil erosion in the area (R. Wang et al., 2008). In 2020, about half of the catchment area is cultivated by terraces (Fig. 2) which were mainly created in the 1990s as a soil conservation measure. For an impression of the study area, see a drone video (A1) of the catchment captured by DJI Mavic Pro 2 flight in the Appendix.

2.2. Datasets and Software

Climate data is represented by a total of 25 automated gauging stations operated by the Yellow River Conservancy Commission (see Fig. 1). Six stations at or just outside the border of the catchment area are considered to better regionalize the catchment and avoid edging effects. Daily records of precipitation cover the period from 1985 to 2010 with some data gaps due to instrument failure. We used the long-term annual mean precipitation values of each station to calculate the R factor. Higher resolution data from, e.g., national gauging stations (maintained by China Meteorological Administration) are rare (n = 1) in the catchment or its near surroundings.

Soil texture and soil organic matter to calculate the K factor originate from the ISRIC SoilGrids (Hengl et al., 2017; Poggio & Sousa, 2020). Although local qualitative soil data exists for the Luoyugou catchment (acquired by the Yellow River Conservancy Commission Tian Shui Experimental Station of Soil and Water Conservation), it does not provide the required physical and quantitative soil parameters. A digital elevation model (DEM, spatial resolution of 10 m) that was digitized from a 1:10000 elevation map by F. C. Qin (2006) was used for terrain analysis (LS factor). The 10m resolution is an appropriate compromise for our approach as higher spatial resolutions of DEMs do not necessarily result in better results (Bircher, Liniger, & Prasuhn, 2019).

The land use information is available for the years 1986, 1995, 2001, 2008, 2015, and 2020. The six years from 1986 to 2020 enable focus on temporal soil erosion dynamics under the objective of land use change in the study area. The land use information was manually interpreted with expert knowledge of the study area by different satellite images (Landsat, Sentinel-2, and SPOT) partly processed in Google Earth Engine (for 2015 and 2020). Randomly ground validation was conducted during various field visits over the last decades. Daily flow discharge and annual sediment loads have been measured at the main outlet of the catchment (Zuojiachang hydrometric station, N 34°35’29 E105°43’6, Fig. 1) for the same period as precipitation data was recorded. All used datasets are listed in Table 1. Background data of the map figures are from various resources like ESRI, FAO, NOAA, USGS, Earthstar Geographics, and Maxar Technology.

The modeling was realized in RStudio v.2022.02.3 (“RStudio: Integrated Development for R,” 2020) with R version 4.2.0 (R Foundation for Statistical Computing, 2022) and the integrated RSAGA (Brenning, Bangs, Becker, Schratz, & Polakowski, 2018) package based on SAGA-GIS v2.2.2 (Conrad et al., 2015). Outcomes were visualized in ESRI ArcGIS Pro 2.9 (ESRI Environmental Systems Research Institute, 2020).

Table 1

Datasets for WaTEM/SEDEM modeling of the Luoyugou in northwestern China.
Dataset Name	Information	Scale/Resolution	Source
Precipitation data	Daily precipitation data, aggregated to monthly and annual precipitation means	25 gauging stations Daily, monthly, annual	Yellow River Conservancy Commission Tian Shui Experimental Station of Soil and Water Conservation
Qualitative soil data	Chinese description of soil types	/	Yellow River Conservancy Commission Tian Shui Experimental Station of Soil and Water Conservation
Quantitative soil data	Textural data of soils	250 m	Hengl et al., 2017; Poggio & Sousa, 2020
Digital Elevation Model	Terrain information	10 m	F. C. Qin, 2006
Land Use	Land use information	1986, 1995, 2001, 2008, 2015, 2020	Manual interpretation from Landsat, Sentinel2, and SPOT images
Flow discharge	Daily flow discharge measurements at the catchment outlet	Daily, monthly, annual	Yellow River Conservancy Commission Tian Shui Experimental Station of Soil and Water Conservation
Road network	Road network of the catchment	/	OpenStreetMap, 2022

2.3. Model structure of WaTEM/SEDEM

The WaTEM/SEDEM, developed by Van Oost, Govers, and Desmet; van Rompaey, Verstraeten, Van Oost, Govers, and Poesen; Verstraeten, Oost, Rompaey, Poesen, and Govers (2000; 2001; 2002) is a spatially distributed sediment delivery model that models the detachment of soil by water (rill and inter-rill erosion) and describes the sediment routing and deposition as long-term annual averages. Thus, WaTEM/SEDEM enables the modeling of sediment production, transport, and deposition within the catchment on a grid cell scale. The model is also able to simulate tillage erosion, however, this form of soil erosion is not part of the present study as the required data was not available.

The first step of WaTEM/SEDEM, namely the estimation of sediment production or the total amount of gross soil loss (SL), is based on the empirical soil erosion model RUSLE (Renard, Foster, Weesies, McCool, & Yoder, 1997) by including the factors rainfall erosivity (R in MJ mm ha^-1 h^-1 yr^-1), soil erodibility (K in t h MJ^-1 mm^-1), two-dimensional slope length and steepness (LS_2D, dimensionless) according to Desmet and Govers (1996), cover and management (C, dimensionless) and support practices (P, dimensionless). The here used basic equation of RUSLE is the following:

$$SL=R\text{*}K\text{*}{LS}_{2D}\text{*}C\text{*}P$$

Each factor of RUSLE was exploited individually based on the listed datasets in Table 1 and explained in detail in section 2.4.

Secondly, the modeled soil loss is routed downslope of the relief and is either redeposited in the landscape or finally transported to the riverine systems. Sediments that leave the catchment by the outlet through the river systems are called total river export or sediment yield of rill and inter-rill erosion. Sediments are routed downslope as long as the transport capacity overpass the sediment supply or when sediments leave the catchments in the river (van Rompaey et al., 2001). Thus, transport capacity (TC in Mg yr^-1) per unit width needs to be estimated (Notebaert et al., 2006) (Eq. 2). TC is controlled by a transport capacity coefficient (k_tc in m) and the RUSLE factors R, K, and LS_2D. Additionally, an inter-rill slope gradient (S_IR in m m^-1) is added. That gradient can be substituted by a slope gradient (S_g in m m^-1) that can be derived from the digital elevation model (van Rompaey et al., 2001).

$$TC= {k}_{TC}\text{*}R\text{*}K\text{*}\left({LS}_{2D}-0.6\text{*}{S}_{IR}\right)={k}_{TC}\text{*}R\text{*}K\text{*}({LS}_{2D}-0.6\text{*}6.86\text{*}{S}_{g}^{0.8})$$

The transport capacity coefficient (k_tc) is a calibrated constant that describes the “theoretical upslope distance that is needed to produce enough sediment to reach the transport capacity at the grid cell, assuming a uniform slope and discharge” (Van Rompaey et al., 2001). As it is difficult to obtain data for k_tc, we used multiple random combinations of constants in a set of 1000 Monte Carlo simulations. The range of constants values is oriented by studies of Batista, Fiener, Scheper, and Alewell (2022) and Borrelli et al. (2018). Furthermore, parcel connectivities (P_con) were simulated. P_con is a reduction parameter that describes reductions of flows at parcel borders if land use changes from forest or grassland to arable land (Notebaert et al., 2006). If a forest or grassland grid cell touches an arable grid cell it receives a specific P_con. A P_con of e.g., 0.75 matters a reduction of the contributing area of the forest or grassland parcel by 75% at the border pixel. Bordering pixels of forests and grasslands are identified in ArcGIS Pro 2.9. (ESRI Environmental Systems Research Institute, 2020) by defining features containing all arable parcels separated from those containing forest and grassland parcels. In the next step, the forest and grassland polygon features are converted into lines and split into line segments in between vertices subsequently. All line segments that are intersected with arable parcels are selected and transformed into raster cells (in alignment with the DEM). Furthermore, parcel trap efficiencies (PTEF) that are relevant for defining the contribution of an upstream cell to the next downstream cell are also simulated. They describe the detention of sediments, e.g., PTEF of 0.75 to a detention of sediments by 75%, e.g. a grid cell of 10 m x 10 m (100 m²) will only contribute 25 m² to the downstream pixel. Following long-term observations in the study catchment and the approach described by Batista et al. (2022), we assumed, that roads serve as pathways for runoff and sediments to follow the road or reach the next field by crossing the infrastructure. According to this assumption, roads have a very high transport capacity, and no deposition can occur on the road surfaces.

An executable software version of WaTEM/SEDEM is provided by the developers and can be downloaded at https://ees.kuleuven.be/eng/geography/modelling/watemsedem/index.html. However, we decided to run WaTEM/SEDEM in R to be more flexible in adjusting the factors and integrating the Monte Carlo simulation. We adopted the R code of WaTEM/SEDEM that is published by Batista et al. (2022). Unlike Batista et al. (2022) we did not simulate the whole parameter space but used defined R, K, C, and P factors. Only P_con, PTEF, and thus LS_2D, and k_tc randomly change during the simulation runs. P_con and PTEF parameters were derived from a uniform distribution between 0 (equals to 0%, no reduction of sediments at border transition resp. no retention of sediments) and 1 (equal to 100%, complete reduction of sediments at border transition resp. complete detention of sediments); uniformly distributed k_tc values were separated in a high k_tc value (k_tc high) between 1 and 200 in steps by 1 for land use classes with a C greater than 0.01 and a low k_tc value (k_tc low) between 1 and 100 in steps by 1 for all other classes.

In the final step, we aggregated the simulations by overlaying the 1000 simulated maps cell-wise. A mean value out of the 1000 simulations is subsequently calculated per cell. This final, aggregated, and averaged simulation is used for interpreting the results regarding spatial clusters and general statistics. It represents the mean catchment's soil loss and sedimentation rates over all simulations. This process was repeated for each of the six years (1986, 1995, 2001, 2008, 2015, and 2020) where land use data were available. As the land use data for 2020 is the most recent data, this year serves as the baseline year to describe the dimension and spatial clusters of soil erosion and deposition. The results of the 2020 modeling are presented and discussed in Chap. 3.1. Besides that baseline modeling, long-term temporal patterns, and trends of the erosion and erosion-deposition rates are the subject of a comparison of the six different land use situations (see Chap. 3.3). As we specifically focus on the temporal soil erosion and sediment dynamics under changing land uses, only the C and P factors changed in the model. We’ve decided to assume stable R, K, and LS factors and a stable road network to avoid the simultaneous variation of multiple variables. That assumption enables a better evaluation of the effects of land use and terrain changes on soil erosion. As terraces were implemented in the 1990s to reduce soil erosion, the aim is to assess the effects of that planned erosion control measure. Terraces are known as an effective measure to protect soils that are cultivated in hilly environments (Chen, Wei, & Chen, 2017; Xiaoyu Liu, Xin, & Lu, 2021) and these effects are captured by the P factor that reduces the overall soil erosion rates.

2.4. Specific calculation of the RUSLE parameters

Rainfall erosivity and soil erodibility were calculated as a long-term mean and are used for all considered years as a stable factor. Rainfall erosivity (R in MJ mm ha^-1 h^-1 yr^-1) is calculated based on 1986–2010 annual means of precipitation as 10-minute precipitation data was not available for the study region. Thus, we choose the equation of Renard and Freimund (1994) (Eq. 3) that is suitable for such a resolution:

$$R=0.04830\text{*}{Prec}^{1.1610}$$

whereas Prec is the long-term mean annual precipitation (mm). In the second step, the rainfall erosivities of the 25 gauging stations within and outside of the catchment are regionalized by Empirical Bayesian Kriging (EBK, Krivoruchko & Gribov, 2019) by including topography as a covariate. 100 simulations were used for log-transformed data for an exponential semivariogram.

The K factor (t ha h ha^-1 MJ^-1 mm^-1) is based on Eq. 4 by Römkens et al. (1997) as it is suggested to use if “(1) data are limited (for instance, no information about the very-fine-sand fraction or organic-matter content) and (2) the textural composition is given in a different classification system.” (Römkens et al., 1997). As we use the ISRIC SoilGrids250m, no information about very fine-sand fractions nor organic-matter contents is available. The geometric mean diameter (D_g in mm) can be derived as proposed by Shirazi and Boersma (1984) in Eq. 5. F_i corresponds to the weight percentage of the particle size fraction (%) and m_i is the arithmetic mean of the particle size limits (mm). n is the number of particle size fractions.

$$K=0.0034+0.0387\text{*}exp\left[-\frac{1}{2}{\left(\frac{{log}_{10}\left({D}_{g}\right)+1.533}{0.7671}\right)}^{2}\right]$$

$${D}_{g}=exp\left(0.01\text{*}\sum _{i=1}^{n}{f}_{i}\text{*}ln\text{*}{m}_{i}\right)$$

LS_2D is calculated with the package RSAGA in R according to Eq. 6 (Desmet & Govers, 1996; McCool, Brown, Foster, Mutchler, & Meyer, 1987).

$${L}_{i,j}=\frac{{\left({A}_{i,j-in }+{D}^{2}\right)}^{m+1}-{A}_{i,j-in}^{m+1}}{{D}^{m+2} \text{*} {X}_{i,j }^{m}\text{*} {22.13}^{m}}$$

where A_i,j-in is the flow accumulation in m² at the inlet of a grid cell (i,j). D is the grid cell size in m and ${X}_{i,j}$ equals to $sin{a}_{i,j}+cos{a}_{i,j}$ where a_i,j is the aspect of the grid cell (i,j). The coefficient m (Eq. 7) can be calculated using the β-value (Eq. 8) and represents the ratio of rill and inter-rill erosion where θ is the slope angle in degrees. M ranges between 0 (ratio of rill to inter-rill erosion close to 0) and 1.

$$m=\frac{\beta }{\beta +1}$$

$$\beta =\frac{\frac{sin\theta }{0.0896}}{\left[0.56+3\text{*}{\left(sin\theta \right)}^{0.8}\right]}$$

Two different S factor equations related to a slope steepness threshold of 9% are used (McCool et al., 1987). Eq. 9 is used for slopes < 9% and Eq. 10 is used for slopes > 9%. Radians are used for slope steepness (s).

$$S=10.8s+0.03$$

$$S=16.8s-0.50$$

As the LS factor is influenced by the catchment area, which is influenced itself by the simulated PTEF, the LS is recalculated for each model simulation.

The C and P factors are calculated for the years (1986, 1995, 2001, 2008, 2015, and 2020) based on the specific land use information for each year. The factors rely on literature values for comparable and adjacent catchments (Al-Quraishi, 2003; Fu et al., 2005). P factors for terraces, which are the only soil erosion mitigation measure present in the study area follow global means presented in Ebabu et al. (2022). Table 2 shows a list of used C and P factors in the present study.

Table 2

List of RUSLE C and P factors by literature used in the study catchment.
Land use type	C factor according to Al-Quraishi; Fu et al. (2003; 2005)	P factor according to Ebabu et al. (2022)
Highway	0	1
Construction land	0	1
River	0	1
Forest	0.09	1
Grassland	0.12	1
Young plantations	0.15	1
Orchard	0.3	1
Terrace	0.3	0.3
Slope farmland	0.3	1
Bare land	1	1

3.1. Soil loss and deposition rates in the catchment for the most recent year 2020

Total averaged erosion rates out of the 1000 simulations (by a cell-wise overlay of 1000 simulation results per investigated year) for the most recent year 2020 are spatially highly variable between 0 t ha^-1 yr^-1 and 247.0 t ha^-1 yr^-1. The mean erosion rate for the catchment is 12.4 ± 11.8 t ha^-1 yr^-1 and exceeds the tolerable soil loss (10 t ha^-1 yr^-1) for the Chinese Loess Plateau (S. J. Zhang & Jiao, 2011). Other studies in the Loess Plateau by Y. Liu and Fu (2016) and D. P. Yue and Dong (2010) reported average soil erosion up to 41.4 t ha^-1 yr^-1 for a comparable catchment. The global modeling of RUSLE (coarse resolution of 25km) of Borrelli et al. (2017) clipped to the catchment extents resulted in a mean erosion rate of 18.19 t ha^-1 yr^-1 which is the same order of magnitude as estimated here. Based on the CSLE, B. Liu et al. (2020) estimated an average erosion rate of 5 t ha^-1 yr^-1 for the territory of China with an average rate of 18.2 t ha^-1 yr^-1 for Chinese sloped, cultivated croplands.

Spatial patterns of soil loss do not necessarily follow the relief as hot spots of soil loss are particularly at the north-western mountain range of the catchment with some isolated hotspots at the southern hillslopes (Fig. 3). The maximum total soil loss rates are in the north-western hotspot. The average deposition rate within the catchment for all simulations in 2020 is 4.8 t ha^-1 yr^-1. High deposition occurs parallel to the depression lines of the catchment.

Expanding the common RUSLE approach to the WaTEM/SEDEM model, the budgeting of average erosion and deposition rates identified that erosion outweighs deposition in the catchment. The average net soil loss rate (gross erosion minus deposition) of all 1000 scenarios after subtracting deposition rates is 7.7 t ha^-1 yr^-1. Simulations result in the highest net soil loss dynamics for the northern and especially the north-western part of the catchment (Fig. 3). That region is mainly characterized by high elevation ranges and steep slopes in addition to high climatic impacts regarding rainfall intensities. Furthermore, intense land use on the hillslopes that are modified by terraces is dominant in that region. At some spots, deposition occurs but has only a minor effect on net soil loss.

As deposition rates are generally lower than soil loss rates, the catchment experiences a net soil loss. Thus, sediments are routed eventually to the outlet of the catchment located in the southeast and are either sedimented in the Wei river or transported further to the Yellow river. On average from all modeled scenarios, a total of 52949 ± 11215 t yr^-1 leave the catchment in 2020 which indicated a substantial soil loss within the Luoyugou catchment.

Regarding the simulated constants of P_con, PTEF, and k_tc, field experiments to calibrate such values and constants would be highly recommended. Advances in assessing one of these constants are already presented (Jung, Jang, & Kim, 2017a, 2017b) for South Korea but are not transferable to the conditions of the here presented catchment in China. A simulation of the constants seemed to be the most appropriate way to assess parameter uncertainty (Bezak, Rusjan, Petan, Sodnik, & Mikoš, 2015; Lizaga Villuendas, Latorre, Gaspar, & Navas, 2022). For future investigations in the study area, the determination of k_tc is planned to at least reduce the uncertainty of the transport capacity coefficient. As signs of erosion are very well observable by UAVs in the study area, P_con could be captured and estimated via drone imagery, and PTEF may be determined by the installation of sediment traps.

3.2. Comparing modeled sediment yield of rill and inter-rill erosion with sediment load measurements at the catchment outlet

The long-term modeled averaged total sediment yield of rill and inter-rill erosion by considering the variation in C and P factors due to the land use change for all timesteps from 1986–2020 is 58440 ± 11015 t yr^-1. Measured annual sediment loads are only available from 1986 to 2010 (average 338314 t yr^-1). Even though modeled and measured values do not cover completely the same period, the overlap of 24 years indicates that modeled annual erosion sediment yield from rill and inter-rill erosion is much lower than measurements of total sediment production (Fig. 4). In fact, modeled long-term mean sediment yields of rill and inter-rill erosion for the overlapping period (1986, 1995, 2001, and 2008) are by a factor of 5.8 smaller than the measured long-term mean value of total sediment production. WaTEM/SEDEM modeling only predicts sediment production from rill and inter-rill erosion, while other processes like landslides, gully, tillage, and riverbank erosion or construction work are not captured. Xu, Zhu, Zhang, and Gao (2020) state that 12% of the total soil loss of the catchment stems from landslides. Additionally, 44 active erosion gullies with a total area of 1.12 km² (1.5% of the catchment area) are observed within the catchment (W. Zhang & Liu, 2019).

Measurements of sediment load are highly driven by extreme events that transport sediments to the outlet via flash floods (Ralston, Warner, Geyer, & Wall, 2013; Ran et al., 2020; Tu et al., 2020). Nevertheless, very high sediment load was measured in 1988 (1623000 t yr^-1) and 2001 (1163000 t yr^-1) for the Luoyugou catchment, but there is no clear trend detectable over time. Neglecting these two extreme years would result in a long-term average sediment load of 246602 t yr^-1 instead of 338314 t yr^-1, which is still 4 times the average modeled sediment load. However, neglecting years with extreme erosion events and comparing measured to modeled values in “base line sediment yields” (see Fig. 4, red circles and purple line), measured (1986, 1995, 2000, 2002, 2008) to modeled (1986, 1995, 2001, 2008) rates are in the same order of magnitude with 106682 t yr^-1 and 61128 t yr^-1, respectively.

Relating measured annual river sediments at the catchment outlet from 1986 to 2010 to the modeled long-term mean sediment yield of rill and inter-rill erosion at the same location from 1996 to 2020, only 17.3% accounts for rill and inter-rill erosion. If only “base line” years are compared, 57.3% of the total annual river sediments originate from rill and inter-rill erosion.

3.3. Changes in modeled erosion and deposition rates from 1986 to 2020

The six modeled years that cover the land use patterns from 1986 to 2020 have captured a clear shift of slope farmland to terrace cultivation in the transition from 1986 to 1995 (Fig. 5). In 1986, only 4.0% of the catchment’s area was used by terrace cultivation but 68.8% by slope farming. In 1995, a share of 54.5% of the area was identified to be terraced farmland and only 15.5% was left to slope farmland (Fig. 5). Terraces were explicitly installed to reduce soil erosion. Modeled erosion rates (without considering deposition) for each timestep under respect to changing land use reflected by specific C and P factors change drastically from 1986 (22.3 t ha^-1 yr^-1) to 1995 (13.1 t ha^-1 yr^-1) and further down to a minimum in 2015/2020 (12.3 t ha^-1 yr^-1 for both years).

This nearly 50% decrease in soil erosion in time from 1986 to 2020 is a combined mitigation effect of terracing with simultaneous changes in land use - land cover coupled with ambiguous effects. Although terrace cultivation has increased over time completely replacing sloped farmland, a substantial decrease in forest and grassland is simultaneously registered (Fig. 5). Both land use types (forest and grassland) are typically connected to lower erosion rates. Young plantations and orchards are replacing the formerly forested and grassland areas. Depending on the soil cover beneath plantation and orchard trees, both land use types might or might not be prone to erosion.

3.4. RUSLE parameters in the Luoyugou catchment

Rainfall erosivity in the Luoyugou catchment ranges between 958 and 1409 MJ mm ha^-1 h^-1 yr^-1 with a mean R factor of 1189 MJ mm ha^-1 h^-1 yr^-1. The northern and southern margins of the catchment that represent the highest altitudes (up to 1897 m a.s.l.) are the regions with the highest values. The valley bottom and the outlet of the catchment in the East have the lowest R factors (Fig. 6). T. Yue, Yin, Xie, Yu, and Liu (2022) predicted a mean R value of 781 MJ mm ha^-1 h^-1 yr^-1 for the same region. Global predictions by Panagos et al. (2017) based on the Global Rainfall Erosivity Database are averaged to 831 MJ mm ha^-1 h^-1 yr^-1 for the study area and 1600 MJ mm ha^-1 h^-1 yr^-1 for the whole Chinese territory. T. Yue et al. (2022) predicted an even higher mean R for China of 2200 MJ mm ha^-1. The here presented rainfall erosivity only relies on annual means. With that, no extreme events are captured as they are laminated by the coarse temporal resolution. As the aim is to assess the rainfall erosivity by a high quantity of rainfall stations, we relied on stational records from the Yellow River Conservancy Commission and the Tian Shui Experimental Station of Soil and Water Conservation instead of the use of data from the China Meteorological Administration due to a lack of stations in the study area.

K factors are relatively coarse (250 m) due to the resolution of SoilGrids250m (Hengl et al., 2017). However, spatial patterns are obvious (Fig. 7) as the highest factors are in the eastern valley bottom and scattered at the south-eastern edge of the catchment with a factor up to 0.0278 t ha h ha^-1 MJ^-1 mm^-1. The north-western catchment edge, which represents the most elevated slopes, has the lowest K factors. The minimum K factor is 0.004 t ha h ha^-1 MJ^-1 mm^-1. As such, the K factors can be classified as low to moderate (according to the classification proposed by Foster, Weesies, McCool, Yoder, & Renard, 1999) with a catchment-wide average of 0.023 t ha h ha^-1 MJ^-1 mm^-1 compared to 0.0290 t ha h ha^-1 MJ^-1 mm^-1 modeled by Borrelli et al. (2017) for the Luoyugou catchment in a global study. Regarding the study by B. Wang, Zheng, and Guan (2016) for China, only the eastern part of the country is covered in the study with no data for the Luoyugou catchment. Please note that the effects of the stone cover were not included in the here presented K factor (Panagos, Meusburger, Ballabio, Borrelli, & Alewell, 2014).

The LS factors are simulated for each year according to the changes in parcel trap efficiency (PTEF) per iteration. However, due to the cell-wise averaging of the 1000 simulations, no significant statistical nor spatial differences are obvious in the factors for 1986, 1995, 2001, 2008, 2015, and 2020. The average LS factor of the 1000 simulations for the baseline year 2020 results in a mean factor of 3.3 with a range from 0.03 to 9.4. Borrelli et al. (2017) modeled an LS factor of 4.1 for the study area. Spatial patterns (Fig. 8) show the highest LS values in the northern mountain ranges and the lowest values at the southern edge of the catchment as well as in the valley bottom. The mean LS of the catchment is by a factor of 2 higher than for the European Union (Panagos, Borrelli, & Meusburger, 2015) but much lower than for Switzerland (14.8; note that an adapted factor was developed for the steep alpine slopes of Switzerland; Schmidt, Tresch, & Meusburger, 2019).

C factors of the year 2020 follow the land use classification. The average C factor for the whole catchment is 0.22 which shows agreement with the coarse C factor of 0.14 for the same region by Borrelli et al. (2017). The most frequent class of cultivated land (on hillslopes with/without terraces) has a C value of 0.3. The northern mountain range, a region with wild grassland and shrublands, examines less influence by human activities. Agricultural practices on terraces can be seen as an efficient measure to reduce soil loss and thus a P value (0.3) was used for such units as proposed in the literature by Ebabu et al. (2022) for terraces. The protective effect of terraces has been proved multiple times in the literature (Chen et al., 2017; Xiaoyu Liu et al., 2021; Xiaobing Liu, Zhang, Zhang, Ding, & Cruse, 2011; Sang-Arun, Mihara, Horaguchi, & Yamaji, 2006; Wen & Zhen, 2020). As such, the combined CP value (Fig. 9) of cultivated land on terraces is lower (0.09) than on cultivated land on hillslopes (0.3). All other land use classes are assigned with a P value of 1 (equivalent to no protection measure). As the trend from 1986 to 2020 shows (Fig. 5), the effect of the shift in the landscape from the cultivation of hillslopes to terraces is represented in the CP factor. The average of the combined CP factor for 2020 is 0.19. High CP factors (≥ 0.3) undergo a spatial change from ubiquitous coverage in 1986 to larger patches in the west in 1995 and 2001 and finally patches in the center and east (2005, 2015, and 2020) temporal dynamics of the CP factor from 1986 to 2020.

The multiplication of the five factors and averaging by the 1000 Monte Carlo simulations of the year 2020 results in the soil erosion map without considering deposition as presented in Fig. 9. Spatial patterns of Fig. 10 are in good accordance with the soil erosion occurrence probability map by Shen (2011). Some small parts of the catchments are missing land use data and therefore cannot be part of the modeling approach (e.g. southern border).

The WaTEM/SEDEM application for the Luoyugou catchment resulted in high soil erosion rates due to rill and inter-rill erosion (22.3 t ha^-1 yr^-1) with a deposition rate within the catchment of 12.0 t ha^-1 yr^-1 for the year 1986. Converting sloped farmland into terraces with simultaneous changes in land use/ land cover in the catchment reduced erosion rates to a minimum of 12.4 t ha^-1 yr^-1 resp. a deposition rate of 4.8 t ha^-1 yr^-1 (both modeled years of 2015 and 2020). Comparing the total modeled sediment yield of rill and inter-rill erosion of the catchment with sediment load measurements of a comparable timeframe, revealed that modeled erosion contributes only about 17% to the total measured sediment loads. This is consistent with observations that processes like gullying, landslides, construction works, and riverbank erosion contribute considerably to the total sediment load. However, comparing measured years without extreme events (and thus with likely high contribution from gullying, land sliding, or construction work) to modeled values, yielded comparable erosion rates (106682 t yr^-1 and 61128 t yr^-1, respectively).

Temporal trends in soil erosion and deposition dynamics from 1986 to 2020 show the expected reduction in soil erosion loss due to the installation of terraces although land use transition had simultaneously reduced favorable land use classes such as forests and grasslands. The measurement and thus determination of the constants P_con, PTEF, and k_tc might reduce modeling uncertainty as fewer assumptions have to be made. For this reason, we suggest developing strategies to parameterize and measure these values and add them to an updated version of the model.

Author Contributions: Conceptualization, Simon Scheper and Christine Alewell; methodology, Simon Scheper and Christine Alewell; software, Simon Scheper; validation, all authors; formal analysis, Simon Scheper; investigation, Simon Scheper; resources, Zhongbao Xin, Chunyue Liu, and Lishan Ran; data curation, all authors; writing—original draft preparation, all authors; writing—review & editing, all authors; visualization, Simon Scheper and Zhongbao Xin; supervision, Christine Alewell; project administration, Christine Alewell; funding acquisition, Simon Scheper, Lishan Ran and Christine Alewell. All authors have read and agreed to the published version of the manuscript.

Funding: This work was partly supported by the National Natural Science Foundation of China (grant number 42177319).

Data Availability Statement: Datasets are available upon request from the authors.

Acknowledgments: The authors would like to thank all data providers. The here-used code is an adoption of the code published by Batista et al. (2022). We like to thank the Yellow River Conservancy Commission Tian Shui and Wang Ronghua for providing the climate, soil, and land use data.

Conflicts of Interest: The authors declare no conflict of interest.

Alewell, C., Borrelli, P. [Pasquale], Meusburger, K., & Panagos, P. [Panos] (2019). Using the USLE: Chances, challenges and limitations of soil erosion modelling. International Soil and Water Conservation Research, 7(3), 203–225. https://doi.org/10.1016/j.iswcr.2019.05.004
Al-Quraishi, A. M. F. (2003). Soil Erosion Risk Prediction with RS and GIS for the Northwestern Part of Hebei Province, China. Journal of Applied Sciences, 3(10), 659–669. https://doi.org/10.3923/jas.2003.659.669
Batista, P. V. G., Fiener, P., Scheper, S., & Alewell, C. (2022). A conceptual-model-based sediment connectivity assessment for patchy agricultural catchments. Hydrology and Earth System Sciences, 26(14), 3753–3770. https://doi.org/10.5194/hess-26-3753-2022
Bezak, N., Rusjan, S., Petan, S., Sodnik, J., & Mikoš, M. (2015). Estimation of soil loss by the WATEM/SEDEM model using an automatic parameter estimation procedure. Environmental Earth Sciences, 74(6), 5245–5261. https://doi.org/10.1007/s12665-015-4534-0
Bircher, P., Liniger, H. P., & Prasuhn, V. (2019). Comparing different multiple flow algorithms to calculate RUSLE factors of slope length (L) and slope steepness (S) in Switzerland. Geomorphology, 346, 106850. https://doi.org/10.1016/j.geomorph.2019.106850
Borrelli, P. [Pasquale], Robinson, D. A., Fleischer, L. R., Lugato, E., Ballabio, C., Alewell, C., . . . Panagos, P. [Panos] (2017). An assessment of the global impact of 21st century land use change on soil erosion. Nature Communications, 8(1), 2013. https://doi.org/10.1038/s41467-017-02142-7
Borrelli, P. [Pasquale], Van Oost, K. [Kristof], Meusburger, K., Alewell, C., Lugato, E., & Panagos, P. [Panagos] (2018). A step towards a holistic assessment of soil degradation in Europe: Coupling on-site erosion with sediment transfer and carbon fluxes. Environmental Research, 161, 291–298. https://doi.org/10.1016/j.envres.2017.11.009
Brenning, A., Bangs, D., Becker, M., Schratz, P., & Polakowski, F. (2018). Package 'RSAGA' (Version 1.3.0) [Computer software]. Retrieved from https://cran.r-project.org/web/packages/RSAGA/RSAGA.pdf
Chen, D., Wei, W., & Chen, L. (2017). Effects of terracing practices on water erosion control in China: A meta-analysis. Earth-Science Reviews, 173, 109–121. https://doi.org/10.1016/j.earscirev.2017.08.007
Conrad, O., Bechtel, B., Bock, M., Dietrich, H., Fischer, E., Gerlitz, L., . . . Böhner, J. (2015). System for Automated Geoscientific Analyses (SAGA) v. 2.1.4. Geoscientific Model Development, 8(7), 1991–2007. https://doi.org/10.5194/gmd-8-1991-2015
Desmet, P. [P.J.J.], & Govers, G. [G.] (1996). A GIS procedure for automatically calculating the USLE LS factor on topographically complex landscape units. Journal of Soil and Water Conservation, 51(5), 427–433. Retrieved from https://www.jswconline.org/content/51/5/427
Dotterweich, M. (2013). The history of human-induced soil erosion: Geomorphic legacies, early descriptions and research, and the development of soil conservation—A global synopsis. Geomorphology, 201, 1–34. https://doi.org/10.1016/j.geomorph.2013.07.021
Ebabu, K., Tsunekawa, A., Haregeweyn, N., Tsubo, M., Adgo, E., Fenta, A. A., . . . Poesen, J. [Jean] (2022). Global analysis of cover management and support practice factors that control soil erosion and conservation. International Soil and Water Conservation Research, 10(2), 161–176. https://doi.org/10.1016/j.iswcr.2021.12.002
ESRI Environmental Systems Research Institute (2020). ArcGIS Pro (Version 2.7.3) [Computer software]. Redlands. Retrieved from www.esri.com
Fang, H. (2017). Impact of Land Use Change and Dam Construction on Soil Erosion and Sediment Yield in the Black Soil Region, Northeastern China. Land Degradation & Development, 28(4), 1482–1492. https://doi.org/10.1002/ldr.2677
Fang, H., & Sun, L. (2017). Modelling soil erosion and its response to the soil conservation measures in the black soil catchment, Northeastern China. Soil and Tillage Research, 165, 23–33. https://doi.org/10.1016/j.still.2016.07.015
Feng, X., Wang, Y., Chen, L., Fu, B., & Bai, G. (2010). Modeling soil erosion and its response to land-use change in hilly catchments of the Chinese Loess Plateau. Geomorphology, 118(3-4), 239–248. https://doi.org/10.1016/j.geomorph.2010.01.004
Foster, G. R., Weesies, G., McCool, D. K., Yoder, D., & Renard, K. (1999). Revised Universal Soil Loss Equation - User's Manual. Tucson.
Fu, B. J., Zhao, W. W., Chen, L. D., Zhang, Q. J., Lü, Y. H., Gulinck, H., & Poesen, J. [J.] (2005). Assessment of soil erosion at large watershed scale using RUSLE and GIS: a case study in the Loess Plateau of China. Land Degradation & Development, 16(1), 73–85. https://doi.org/10.1002/ldr.646
Gilley, J. E. (2005). EROSION | Water-Induced. In Daniel Hillel (Ed.), Encyclopedia of Soils in the Environment (pp. 463–469). Oxford: Elsevier. https://doi.org/10.1016/B0-12-348530-4/00262-9
Hengl, T., Mendes de Jesus, J., Heuvelink, G. B. M., Ruiperez Gonzalez, M., Kilibarda, M., Blagotić, A., . . . Kempen, B. (2017). Soilgrids250m: Global gridded soil information based on machine learning. PloS One, 12(2), e0169748. https://doi.org/10.1371/journal.pone.0169748
Hui, W., Yongbo, L., Junzhi, L., & A-Xing, Z. (2014). Representation of Agricultural Best Management Practices in a Fully Distributed Hydrologic Model: A Case Study in the Luoyugou Watershed. Journal of Resources and Ecology, 5(2), 179–184. https://doi.org/10.5814/j.issn.1674-764X.2014.02.011
Jung, C.‑G., Jang, W.‑J., & Kim, S.‑J. (2017a). Empirical Estimation of the Spatial Sediment Transport Capacity Coefficient Using the Rain Erosivity Factor and SWAT Model Results in the Han River Basin, South Korea. https://doi.org/10.5194/hess-2017-553
Jung, C.‑G., Jang, W.‑J., & Kim, S.‑J. (2017b). Estimation of 1 km Grid-based WATEM/SEDEM Sediment Transport Capacity Using 1 Minute Rainfall Data and SWAT Semi-distributed Sediment Transport Capacity Results for Han River Basin of South Korea. https://doi.org/10.5194/hess-2016-649
Krivoruchko, K., & Gribov, A. (2019). Evaluation of empirical Bayesian kriging. Spatial Statistics, 32, 100368. https://doi.org/10.1016/j.spasta.2019.100368
Liu, B., Xie, Y., Li, Z., Liang, Y., Zhang, W. [Wenbo], Fu, S., . . . Guo, Q. (2020). The assessment of soil loss by water erosion in China. International Soil and Water Conservation Research, 8(4), 430–439. https://doi.org/10.1016/j.iswcr.2020.07.002
Liu, B., Zhang, K., & Xie, Y. (2002). An Empirical Soil Loss Equation. Proceedings of the 12th ISCO Conference, Beijing.
Liu, S. [Shide], & Li, J. (1989). Soil Erosion and Its Physical and Chemical Properties on Slope in Luoyugou. Journal of Soil and Water Conservation, 3(1), 43–50.
Liu, X. [Xiaobing], Zhang, S., Zhang, X., Ding, G., & Cruse, R. M. (2011). Soil erosion control practices in Northeast China: A mini-review. Soil and Tillage Research, 117, 44–48. https://doi.org/10.1016/j.still.2011.08.005
Liu, X. [Xiaoyu], Xin, L., & Lu, Y. (2021). National scale assessment of the soil erosion and conservation function of terraces in China. Ecological Indicators, 129, 107940. https://doi.org/10.1016/j.ecolind.2021.107940
Liu, Y. [Yu], & Fu, B. (2016). Assessing sedimentological connectivity using WATEM/SEDEM model in a hilly and gully watershed of the Loess Plateau, China. Ecological Indicators, 66, 259–268. https://doi.org/10.1016/j.ecolind.2016.01.055
Lizaga Villuendas, I., Latorre, B., Gaspar, L., & Navas, A. (2022). Effect of historical land‐use change on soil erosion in a Mediterranean catchment by integrating 137 Cs measurements and WaTEM/SEDEM model. Hydrological Processes, 36(5). https://doi.org/10.1002/hyp.14577
McCool, D. K., Brown, L. C., Foster, G. R., Mutchler, C. K., & Meyer, L. D. (1987). Revised Slope Steepness Factor for the Universal Soil Loss Equation. Transactions of the ASAE, 30(5), 1387–1396. https://doi.org/10.13031/2013.30576
Notebaert, B., Vaes, B., Verstraeten, G. [G.], Govers, G. [G.], Van Oost, K. [K.], Van Rompaey, A., & (Keine Angabe) (2006). WaTEM/SEDEM version 2006 Manuel. Leuven.
OpenStreetMap (2022). Road network. Retrieved from https://www.openstreetmap.org
Panagos, P. [Panos], Borrelli, P. [Pasquale], & Meusburger, K. (2015). A New European Slope Length and Steepness Factor (LS-Factor) for Modeling Soil Erosion by Water. Geosciences, 5(2), 117–126. https://doi.org/10.3390/geosciences5020117
Panagos, P. [Panos], Borrelli, P. [Pasquale], Meusburger, K., Yu, B., Klik, A., Jae Lim, K., . . . Ballabio, C. (2017). Global rainfall erosivity assessment based on high-temporal resolution rainfall records. Scientific Reports, 7(1), 4175. https://doi.org/10.1038/s41598-017-04282-8
Panagos, P. [Panos], Meusburger, K., Ballabio, C., Borrelli, P. [Pasqualle], & Alewell, C. (2014). Soil erodibility in Europe: A high-resolution dataset based on LUCAS. The Science of the Total Environment, 479-480, 189–200. https://doi.org/10.1016/j.scitotenv.2014.02.010
Poggio, L., & Sousa, L. de (2020). Soilgrids250m 2.0 - WRB classes and probabilites. ISRIC - World Soil Information. https://doi.org/10.17027/ISRIC-SOILGRIDS.C4DC161C-D62D-11EA-A1A3-292680B15169
Qin, F. C. (2006). Forest Pattern on Regulation of Erosion and Sediment Processes of Watershed in Loess Area (Dissertation). Beijing, Beijing Forestry University.
Qin, R. J. [R. J.], Lu, G., & An, L. (2019). Influence of Soil and Water Conservation Measures to Runoff and Sediment Variations of Luoyugou Watershed. Soil and Water Conservation in China, 1, 19–24.
Qin, R. J. [Rui Jie], Li, P., Xiao, P. Q., & An, L. P. (2015). Precipitation Features of the Luoyugou Basin in Recent 26 Years. In Advances in Intelligent Systems Research, Proceedings of the 2015 International Conference on Automation, Mechanical Control and Computational Engineering. Atlantis PressParis, France. https://doi.org/10.2991/amcce-15.2015.197
R Foundation for Statistical Computing (2022). R: A language and environment for statistical [Computer software]. Vienna, Austria: R Core Team: R Core Team. Retrieved from https://www.R-project.org/
Ralston, D. K., Warner, J. C., Geyer, W. R., & Wall, G. R. (2013). Sediment transport due to extreme events: The Hudson River estuary after tropical storms Irene and Lee. Geophysical Research Letters, 40(20), 5451–5455. https://doi.org/10.1002/2013GL057906
Ran, L., Yang, X., Tian, M., Shi, H., Liu, S. [Shaoda], Yu, R., & Zhou, Y. (2020). Riverine export of water, sediment and carbon during flood events in the arid to semi‐arid Wuding River on the Chinese Loess Plateau. Earth Surface Processes and Landforms, 45(8), 1777–1788. https://doi.org/10.1002/esp.4845
Renard, K., Foster, G., Weesies, G., McCool, D., & Yoder, D. (Eds.) (1997). Agriculture Handbook: Vol. 703. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). Washington D.C.
Renard, K., & Freimund, J. (1994). Using monthly precipitation data to estimate the R-factor in the revised USLE. Journal of Hydrology, 157(1-4), 287–306. https://doi.org/10.1016/0022-1694(94)90110-4
Römkens, M., Young, R., Poesen, J. [J.W.A.], McCool, D., El-Swaify, S., & Bradford, J. (1997). Soil erodibility factor (K). In K. Renard, G. Foster, G. Weesies, D. McCool, & D. Yoder (Eds.), Agriculture Handbook: Vol. 703. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE) (Vol. 703). Washington D.C.
RStudio: Integrated Development for R [Computer software] (2020). Boston, MA: RStudio Team: RStudio Team. Retrieved from http://www.rstudio.com/
Sang-Arun, J., Mihara, M., Horaguchi, Y., & Yamaji, E. (2006). Soil erosion and participatory remediation strategy for bench terraces in northern Thailand. CATENA, 65(3), 258–264. https://doi.org/10.1016/j.catena.2005.11.010
Schmidt, S., Tresch, S., & Meusburger, K. (2019). Modification of the RUSLE slope length and steepness factor (LS-factor) based on rainfall experiments at steep alpine grasslands. MethodsX, 6, 219–229. https://doi.org/10.1016/j.mex.2019.01.004
Shen, Q. (2011). Assessing soil erosion risk in the Hilly-gullied area: A case study of Luoyugou watershed in Tianshui, Loess Plateau, North China. In X. Li (Ed.), 2011 19th International Conference on Geoinformatics: Shanghai, China, 24 - 26 June 2011 (pp. 1–5). Piscataway, NJ: IEEE. https://doi.org/10.1109/GeoInformatics.2011.5981073
Shirazi, M., & Boersma, L. (1984). Unifying Quantitative Analysis of Soil Texture. Soil Science Society of America Journal, 48, 142.
Tu, T., Ercan, A., Carr, K. J., Kavvas, M. L., Trinh, T., Ishida, K., . . . Brown, K. (2020). Coupling hydroclimate-hydraulic-sedimentation models to estimate flood inundation and sediment transport during extreme flood events under a changing climate. The Science of the Total Environment, 740, 140117. https://doi.org/10.1016/j.scitotenv.2020.140117
Van Oost, K. [Kristof], Govers, G. [Gerard], & Desmet, P. [Phillipe] (2000). Evaluating the effects of changes in landscape structure on soil erosion by water and tillage. Landscape Ecology, 15(6), 577–589. https://doi.org/10.1023/A:1008198215674
Van Rompaey, A. J. J., Verstraeten, G. [Gert], Van Oost, K. [Kristof], Govers, G. [Gerard], & Poesen, J. [Jean] (2001). Modelling mean annual sediment yield using a distributed approach. Earth Surface Processes and Landforms, 26(11), 1221–1236. https://doi.org/10.1002/esp.275
Vanmaercke, M., Panagos, P. [Panos], Vanwalleghem, T., Hayas, A., Foerster, S., Borrelli, P. [Pasquale], . . . Poesen, J. [Jean] (2021). Measuring, modelling and managing gully erosion at large scales: A state of the art. Earth-Science Reviews, 218, 103637. https://doi.org/10.1016/j.earscirev.2021.103637
Verstraeten, G. [G.], Oost, K., Rompaey, A., Poesen, J. [J.], & Govers, G. [G.] (2002). Evaluating an integrated approach to catchment management to reduce soil loss and sediment pollution through modelling. Soil Use and Management, 18(4), 386–394. https://doi.org/10.1111/j.1475-2743.2002.tb00257.x
Wang, B., Zheng, F., & Guan, Y. (2016). Improved USLE-K factor prediction: A case study on water erosion areas in China. International Soil and Water Conservation Research, 4(3), 168–176. https://doi.org/10.1016/j.iswcr.2016.08.003
Wang, R., Qin, B., Huang, C., Zhang, H. [Hulin], Chang, D., Dong, Y., & Zhao, L. (2008). Characteristics of typical rainstorm,flood and sediment yield in Luoyugou Watershed. Science of Soil and Water Conservation, 6(4), 12–17.
Wen, X., & Zhen, L. (2020). Soil erosion control practices in the Chinese Loess Plateau: A systematic review. Environmental Development, 34, 100493. https://doi.org/10.1016/j.envdev.2019.100493
Wu, D., Liu, Q., Xia, R., & Li, T. (2022). Study on the changes in vegetation structural coverage and its response mechanism to hydrology. Open Geosciences, 14(1), 79–88. https://doi.org/10.1515/geo-2020-0322
Xin, Z., Qin, Y., & Yu, X. (2016). Spatial variability in soil organic carbon and its influencing factors in a hilly watershed of the Loess Plateau, China. CATENA, 137, 660–669. https://doi.org/10.1016/j.catena.2015.01.028
Xu, X., Zhu, T., Zhang, H. [Hongwu], & Gao, L. (2020). A Close Look of the Gravity Erosion on the Loess Plateau of China. In X. Xu, T. Zhu, H. Zhang, & L. Gao (Eds.), Springer eBook Collection. Experimental Erosion: Theory and Practice of Soil Conservation Experiments (1st ed., pp. 75–93). Singapore: Springer Singapore; Imprint Springer. https://doi.org/10.1007/978-981-15-3801-8_6
Yang, H. (2013). Statement on the Tianshui Experimental Site of Soil and Water Conservation in 1940s. Asian Agricultural Research, 75(8), 68–71.
Yue, D. P., & Dong, M. Y. (2010). Investigation and estimation of soil Erosion of rural settlements in abundant and coarse sediment area of north Shaanxi - An example in Nianzhuanggou river basin. J. Catastrophol., 25(1), 59–63. Retrieved from https://www.scopus.com/inward/record.uri?eid=2-s2.0-84958175357&partnerID=40&md5=ddb485066df4ab7bb4961e78501207c2
Yue, T., Yin, S., Xie, Y., Yu, B., & Liu, B. (2022). Rainfall erosivity mapping over mainland China based on high-density hourly rainfall records. Earth System Science Data, 14(2), 665–682. https://doi.org/10.5194/essd-14-665-2022
Zhang, S. J., & Jiao, J. (2011). Soil loss tolerance in the Loess Plateau based on the healthy function of the lower reaches of the Yellow River. Science of Soil and Water Conservation, 9(1), 9–15.
Zhang, W. [Wenyin], & Liu, Y. [Yinfeng] (2019). Research on Visual Interpretation and Spatial Distribution Pattern of the Erosion Gully in Luoyugou Watershed of China. Environment and Natural Resources Research, 9(3), 23. https://doi.org/10.5539/enrr.v9n3p23
Zhao, G., Klik, A., Mu, X., Wang, F., Gao, P., & Sun, W. (2015). Sediment yield estimation in a small watershed on the northern Loess Plateau, China. Geomorphology, 241, 343–352. https://doi.org/10.1016/j.geomorph.2015.04.020

VideoA1.mp4
Video A1: DJI Mavic Pro 2 record overflying a part of the Luoyugou catchment.

Download PDF

Journal Publication

published 03 Dec, 2023

Read the published version in International Soil and Water Conservation Research →

Version 1

posted

You are reading this latest preprint version

Soil Loss and Sedimentation Rates in a Sub-Catchment of the Yellow River Basin in China

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1. Study area

2.2. Datasets and Software

2.3. Model structure of WaTEM/SEDEM

2.4. Specific calculation of the RUSLE parameters

3. Results & Discussions

3.1. Soil loss and deposition rates in the catchment for the most recent year 2020

3.3. Changes in modeled erosion and deposition rates from 1986 to 2020

3.4. RUSLE parameters in the Luoyugou catchment

4. Conclusions

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1