Field experiments on quantifying the contributions of Coreopsis canopies and roots to controlling runoff and erosion on steep loess slopes

Grass recovery is often implemented in the loess area of China to control erosion.However,the effect mechanisms of grass cover on runoff erosion dynamics on steep loess hillslopes is still not clear.Taking the typical forage species(Coreopsis)in semiarid areas as subject,this study quantified the effects of canopies and roots on controlling slope runoff and erosion.A series of field experiments were conducted in a loess hilly region of China.Field plots(5 m length,2 m width,25°slope gradient)constructed with three ground covers(bare soil;Coreopsis with intact grass;only roots of Coreopsis),were applied with simultaneous simulated rainfall(60mm h^(-1))and upslope inflow(10,30,50,70,90L min^(-1)).The results showed that compared with bare soil,intact grass significantly reduced runoff and soil loss rates by 16.6% and 62.4% on average,and decreased soil erodibility parameter by 66.3%.As inflow rate increased,the reductions in runoff and soil loss rates increased from 2.93 to 14.00 L min^(-1)and 35.11 to 121.96 g m^(-2)min^(-1),respectively.Canopies relatively contributed 66.7% to lowering flow velocity,turbulence,weakening erosive force and increasing hydraulic resistance.Roots played a predominant role in reducing soil loss and enhancing soil antierodibility,with relative contributions of 78.8% and 73.8%.Furthermore,the maximum erosion depth reduced by Coreopsis was at the upper slope section which was previously eroded the most.These results demonstrated the efficiency of Coreopsis cover in controlling runoff and erosion on steep loess slopes,especially under large inflow rates and at upper slope sections.We suggest protecting Coreopsis with intact grass at upper slope sections,while the aboveground grass biomass can be used for grazing or harvesting at middle and lower slope sections,with roots reserved.

A Review on Rill Erosion Process and Its Influencing Factors

Rills are frequently observed on slope farmlands and rill erosion significantly contributes to sediment yields. This paper focuses on reviewing the various factors affecting rill erosion processes and the threshold conditions of rill initiation. Six factors, including rainfall, runoff, soil, topography, vegetation and tillage system, are discussed. Rill initiation and network are explored. Runoff erosivity and soil erodibility are recognized as two direct factors affecting rill erosion and other types of factors may have indirect influences on rill erosion through increasing or decreasing the effects of the direct factors. Certain conditions are necessary for rill initiation and the critical conditions are different with different factors. Future studies should be focused on 1) the dynamic changes of rill networks; 2) the combined effect of multiple factors; and 3) the relationships of threshold values with other related factors.

Rainfall and inflow effects on soil erosion for hillslopes dominated by sheet erosion or rill erosion in the Chinese Mollisol region

Erosion agents and patterns profoundly affect hillslope soil loss characteristics. However, few attempts have been made to analyze the effects of rainfall and inflow on soil erosion for hillslopes dominated by sheet erosion or rill erosion in the Chinese Mollisol region. The objective of this study was to discuss the erosive agent(rainfall or inflow), hillslope erosion pattern(sheet erosion or rill erosion) and slope gradient effects on runoff and soil losses. Two soil pans(2.0 m long, 0.5 m wide and 0.5 m deep) with 5° and 10° slopes were subjected to rainfall(0 and 70 mm h–1) and inflow(0 and 70 mm h–1) experiments. Three experimental combinations of rainfall intensity(RI) and inflow rate(IR) were tested using the same water supply of 70 mm by controllingthe run time. A flat soil surface and a soil bed with a straight initial rill were prepared manually, and represented hillslopes dominated by sheet erosion and rill erosion, respectively. The results showed that soil losses had greater differences among treatments than total runoff. Soil losses decreased in the order of RI70+IR70 > RI70+IR0 > RI0+IR70. Additionally, soil losses for hillslopes dominated by rill erosion were 1.7-2.2 times greater at 5° and 2.5-6.9 times greater at 10° than those for hillslopes dominated by sheet erosion. The loss of <0.25 mm soil particles and aggregates varying from 47.72%-99.60% of the total soil loss played a dominant role in the sediment. Compared with sheet erosion hillslopes, rill erosion hillslopes selectively transported more microaggregates under a relatively stable rill development stage, but rills transported increasinglymore macroaggregates under an active rill development stage. In conclusion, eliminating raindrop impact on relatively gentle hillslopes and preventing rill development on relatively steep hillslopes would be useful measures to decrease soil erosion and soil degradation in the Mollisol region of northeastern China.

经验模式分解改进算法的比较

文章针对经验模式分解中的关键问题,运用3种不同的方法进行改良分析,比较其优劣。证实改进算法在特定的方面有其优越性。

Effects of moss-dominated biocrusts on soil detachment by overland flow in the Three Gorges Reservoir Area of China

Biological soil crusts(biocrusts)are important landscape components that exist in various climates and habitats.The roles of biocrusts in numerous soil processes have been predominantly recognized in many dryland regions worldwide.However,little is known about their effects on soil detachment process by overland flow,especially in humid climates.This study quantified the effects of moss-dominated biocrusts on soil detachment capacity(Dc)and soil erosion resistance to flowing water in the Three Gorges Reservoir Area which holds a subtropical humid climate.Potential factors driving soil detachment variation and their influencing mechanism were analyzed and elucidated.We designed five levels of coverage treatments(1%–20%,20%–40%,40%–60%,60%–80%,and 80%–100%)and a nearby bare land as control in a mossdominated site.Undisturbed soil samples were taken and subjected to water flow scouring in a hydraulic flume under six shear stresses ranging from 4.89 to 17.99 Pa.The results indicated that mean Dc of mosscovered soil varied from 0.008 to 0.081 kg m^-2 s^-1,which was 1.9 to 21.0 times lower than that of bare soil(0.160 kg m^-2 s^-1).Rill erodibility(Kr)of mosscovered soil ranged from 0.0095 to 0.0009 s m^-1,which was 2 to 20 times lower than that of bare soil(0.0187 s m^-1).Both relative soil detachment rate and Kr showed an exponential decay with increasing moss coverage,whereas the critical shear stress(τc)for different moss coverage levels did not differ significantly.Moss coverage,soil cohesion,and sand content were key factors affecting Dc,while moss coverage and soil bulk density were key factors affecting Kr.A power function of flow shear stress,soil cohesion,and moss coverage fitted well to estimate Dc(NSE=0.947).Our findings implied that biocrusts prevented soil detachment directly by their physical cover and indirectly by soil properties modification.Biocrusts could be rehabilitated as a promising soil conservation measure during ecological recovery to enhance soil erosion resistance in the Three Gorges Reservoir Area.

Evaluation of an erosion-sediment transport model for a hillslope using laboratory flume data

Climate change can escalate rainfall intensity and cause further increase in sediment transport in arid lands which in turn can adversely affect water quality. Hence, there is a strong need to predict the fate of sediments in order to provide measures for sound erosion control and water quality management. The presence of micro- topography on hillslopes influences processes of runoff generation and erosion, which should be taken into account to achieve more accurate modelling results. This study presents a physically based mathematical model for erosion and sediment transport coupled to one-dimensional overland flow equations that simulate rainfall-runoff generation on the rill and interrill areas of a bare hillslope. Modelling effort at such a fine resolution considering the flow con- nection between Jnterrill areas and rills is rarely verified. The developed model was applied on a set of data gath- ered from an experimental setup where a 650 cm×136 cm erosion flume was pre-formed with a longitudinal rill and interrJll having a plane geometry and was equipped with a rainfall simulator that reproduces natural rainfall characteristics. The flume can be given both longitudinal and lateral slope directions. For calibration and validation, the model was applied on the experimental results obtained from the setup of the flume having 5% lateral and 10% longitudinal slope directions under rainfall intensities of 105 and 45 mm/h, respectively. Calibration showed that the model was able to produce good results based on the R2 （0.84） and NSE （0.80） values. The model performance was further tested through validation which also produced good statistics （R2=0.83, NSE=0.72）. Results in terms of the sedigraphs, cumulative mass curves and performance statistics suggest that the model can be a useful and an important step towards verifying and improving mathematical models of erosion and sediment transport.

The characteristics of rill development and their effects on runoff and sediment yield under different slope gradients

Rill formation is the predominant erosion process in slope land in the Loess Plateau, China. This study was conducted to investigate rill erosion characteristics and their effects on runoff and sediment yielding processes under different slope gradients at a rate of 10°, 15°, 20° and 25° with rainfall intensity of 1.5 mm min-1 in a laboratory setting. Results revealed that mean rill depth and rill density has a positive interrelation to the slope gradient. To the contrary, width-depth ratio and distance of the longest rill to the top of the slope negatively related to slope gradient. All these suggested that increasing slope steepness could enhance rill headward erosion, vertical erosion and the fragmentation of the slope surface. Furthermore,total erosion tended to approach a stable maximum value with increasing slope, which implied that there is probably a threshold slope gradient where soil erosion begins to weaken. At the same time, the correlation analysis showed that there was a close connection between slope gradient and the variousindices of soil erosion: the correlation coefficients of slope gradient with maximal rill depth, number of rills and the distance of the longest rill from the top of the slope were 0.98, 0.97 and-0.98, respectively,indicating that slope gradient is the major factor of affecting the development of rills. Furthermore,runoff was not sensitive to slope gradient and rill formation in this study. Sediment concentration,however, is positively related to slope gradient and rill formation, the sediment concentrations increased rapidly after rill initiation, especially. These results may be essential for soil loss prediction.

A Preliminary Study of Rill Marks in the Yellow River Delta

In recent years, the flow of the Yellow River has often been interrupted, which has resulted in exposure of channel bars and point bars, and even extensive exposure of the riverbed. Consequently, a large number of rill marks have developed. They are diverse in morphology. According to the hydrodynamic types of their formation, they can be grouped into 6 categories, i. e. the wave eroded, backwash, seepage, rain eroded, water drainage and runoff rill marks. Morphologically, they can be divided into more than ten types: the linear, tooth shaped, comb shaped, fence like, ear like, braided, branched, leaf like, flower like, root like, dendritic, net like, radial etc. Their cross sections include the broad u type (the width/depth ratio is over 2, and may reach 10—20), U type (width/depth ratio from 1 to 2), V type, Ω type and ( type. Their occurrences may be attributed to the variations in composition, grain size, color, fabric and morphology. They have 5 scales: the micro scale (length and width within 1 cm), small scale (length and width within 10 cm), medium scale (length and width ranging from 10—100cm), large scale (length and width 1—5 m) and giant-scale (length or width over 5 m).

Rill flow velocity affected by the subsurface water flow depth of purple soil in Southwest China

Subsurface water flow above the weakly permeable soil layer commonly occurs on purple soil slopes.However,it remains difficult to quantify the effect of subsurface water flow on the surface flow velocity.Laboratory experiments were performed to measure the rill flow velocity on purple soil slopes containing a subsurface water flow layer with the electrolyte tracer method considering 3 subsurface water flow depths(SWFDs:5,10,and 15 cm),3 flow rates(FRs:2,4,and 8 L min^(-1)),and 4 slope gradients(SGs:5°,10°,15°,and 20°).As a result,the pulse boundary model fit the electrolyte transport processes very well under the different SWFDs.The measured rill flow velocities were 0.202 to 0.610 m s^(-1) under the various SWFDs.Stepwise regression results indicated a positive dependence of the flow velocity on the FR and SG but a negative dependence on the SWFD.The SWFD had notable effects on the rill flow velocity.Decreasing the SWFD from 15 to 5 cm increased the flow velocity.Moreover,the flow velocities under the 10-and 15-cm SWFDs were 89%and 86%,respectively,of that under the 5-cm SWFD.The flow velocity under the 5-,10-and 15-cm SWFDs was decreased to 89%,80%,and 77%,respectively,of that on saturated soil slopes.The results will enhance the understanding of rill flow hydrological processes under SWFD impact.

Assessment of Rill Erosion Development during Erosive Storms at Angereb Watershed, Lake Tana Sub-basin in Ethiopia

Application of simple and locally based erosion assessment methods that fit to the local condition is necessary to improve the performance and efficiency of soil conservation practices. In this study, rill erosion formation and development was investigated on the topo-sequence of three catchments （300-500 m slope length）; and on agricultural fields （6 m and 24 m slope lengths） with different crop-tillage surfaces during erosive storms. Rill density and rill erosion rates were measured using rill cross section survey and close range digital photogrammetry. Rill formation and development was commonly observed on conditions where there is wider terrace spacing, concave slope shapes and unstable stone terraces on steep slopes. At field plot level, rill development was controlled by the distribution and abrupt change in the soil surface roughness and extent of slope length. At catchment scale, however, rill formation and development was controlled by landscape structures, and concavity and convexity of the slope. Greater rill cross sections and many small local rills were associated to the rougher soil surfaces. For instance, relative comparison of crop tillage practices have showed that faba-bean tillage management was more susceptible to seasonal rill erosion followed by Teff and wheat tillage surfaces under no cover condition. Surface roughness and landscape structures played a net decreasing effect on the parallel rill network development. This implies that spatial and temporal variability of the rill prone areas was strongly associated with the nature and initial size of surface micro-topography or tillage roughness. Thus, it is necessary to account land management practices, detail micro-topographic surfaces and landscape structures for improved prediction of rill prone areas under complex topographic conditions. Application of both direct rill cross section survey and close range digital photogrammetric techniques could enhance field erosion assessment for practical soil conservation improvement.

A Preliminary Study on Traditional Level-trench Method to Prevent Rill Initiation in the Three Gorges Region,China

The level-trench method is a traditional tillage operation to develop a drainage system on slope land in the Three Gorges region. Before crops are planted, farmers prepare the land with level trenches spaced 3-1o m apart, depending on the slope gradient： steeper slopes require shorter distances. Little scientific analysis has been done on the use of traditional level trenches for soil conservation. We conducted a field investigation and simulation experiments, and present a comparison between the spacing of level trenches and the slope length required for rill initiation. The results indicate that the spacing of level trenches in farmland is close to the slope length required for rill initiation in the experimental plots, and regression models of trench spacing with the gradient and the slope length required for rill initiation with the gradient have similar formats; the coefficients of determination （R2） for the two equations are 0.99 and 0.88, respectively. There were some differences between the field survey and experiments, but we conclude that the traditional level-trench method has an important role in preventing rill initiation.

Apportioning above-and below-ground effects of moss biocrusts on soil detachment by overland flow in a subtropical climate

Biocrusts affect soil detachment through above-ground(top crust’s surface covering)and below-ground(sub-crust’s binding and bonding,B&B)effects,which might vary with biocrust development or coverage.However,these effects in humid climates are still unclear.This study was conducted to apportion and quantify the surface covering and B&B effects of moss biocrusts with five coverage levels(1%–20%,20%–40%,40%–60%,60%–80%,and 80%–100%)on soil detachment by overland flow in a subtropical humid climate.Two treatments with one being intact moss crusts and one removing the aboveground moss tissues were designed for each coverage level,and bare soil was used as the baseline.The results indicated that soil detachment capacity(Dc)and rill erodibility(Kr)decreased with biocrust coverage.After removing the above-ground moss tissues,the impeding effect of biocrusts on soil detachment weakened,but still increasing soil erosion resistance relative to bare soil.For intact crust,Dc was reduced by 50%–95%compared with bare soil,wherein 36%–55%and 14%–40%were attributed to the surface covering and B&B,respectively.The top crust contributed more than sub-crust to the soil detachment reduction,which were related to but not linear with biocrust coverage.When biocrust coverage reached mid-to-higher level(40%–100%),both top crust and sub-crust steadily contributed to soil detachment reduction with 60%and 40%,respectively.The findings advance a better understanding of the influencing mechanism of biocrusts on soil erosion in humid climates and highlight the importance of saving biocrusts as ecosystem functions.

Developments of Rill Networks: An Experimental Plot Scale Study

Enumerating the relative proportions of soil losses due to rill erosion processes during monsoon and post-monsoon season is a significant factor in predicting total soil losses and sediment transport and deposition. Present study evaluated the rill network with simulated experiment of treatments on varying slope and rainfall intensity to find out the rill erosion processes and sediment discharge in relation to slope and rainfall intensity. Results showed a significant relationship between the rainfall intensity and sediment yield (r = 0.75). Our results illustrated that due to increase in rainfall intensity represent the development of efficient rill network while, no rill was found with a slope of 20° and a rainfall intensity of 60 mm·h-1. The highest rill length was observed in plot E with 20° slope and 120 mm·h-1 rainfall intensity at 360 minutes. Positive and strong correlation (R2 = 0.734, P 0.001) was observed between the cumulative rainfall intensity and sediment discharge. A longitudinal profile was delineated and showed that the depth and numbers of depressions amplified with time and were more prominent for escalating rainfall intensity for its steeper slopes. Information derived from the study could be applied to estimate longer-term erosion stirring over larger areas possessing parallel landforms.

Monitoring of Soil Loss from Erosion Using Geoinformatics and Geotechnical Engineering Methods

In this study, the position of all major rill and gully erosion sites were located using hand held GPS （Global Positioning System） receiver during reconnaissance surveys. Based on severity rating and geopolitical considerations, six of the erosion gully sites were selected for monitoring. Control points were established around each of the gully sites using three Leica 500 dual frequency GPS receivers by method of DGPS （differential GPS） surveys. Detailed topographical survey of the gully sites was carried out using total stations. With the aid of SPOT satellite imageries in combination with total station data and GIS （geographic information system） location maps, contoured maps along with DEM （digital elevation model） were generated using ARCGIS 9.2 software. The morphological parameters of the gullies including depth, width, length and area of the gullies were determined. Volumetric estimate of the amount of soil loss from gully erosion was also carried out. Soil samples were recovered from the gully sites to determine their erodibility and other parameters to be used for soil loss modeling. The result of the studies was used as an indicator for determining the gully initiation point. Slope-area relationship and threshold of gully initiation was established. The minimum volume of soil loss occurred in gully No. 2 （Queen Ede）. The minimum AS^2 value was 345 while the maximum was 3,267.

Gorilla Thriller More Chinese tourists keen to encounter mountain gorillas

The majestic mountain gorilla has a very secluded home,living exclusively in the tropical forests of the Democratic Republic of the Congo(DRC),Rwanda and Uganda.According to the World Wildlife Fund for Nature(WWF),there are more than 1,000 mountain goril las living in the world,less than a half of which live in the Virunga Mountains on the border of the DRC and Rwanda,with the remainder calling tlie Bwindi Impeneti able National Park in Uganda home.

Effects of physical crust on soil detachment by overland flow in the Loess Plateau region of China

Physical soil crust(PSC),a key component of surface soil structure,exists extensively in loess areas.PSC is considered to have a significant effect on soil detachment processes.However,the long-term effects and the corresponding mechanisms of PSC on soil detachment by overland flow are still not well understood,especially in natural environments.To investigate temporal variation in soil erosion resistance and the underlying factors during PSC formation,an 8×8-m soil plot was exposed to natural conditions in the Loess Plateau over a 524-day period spanning two rainy seasons and a winter between them.A flume test was conducted to determine soil detachment capacity(Dc)under six designed flow shear stress levels(5.66-22.11 Pa)using crusted(SC)and non-crusted(NSC)soil samples at different PSC development stages.Subsequently,two soil erosion resistance parameters,rill erodibility(K_(r))and critical shear stress(τ_(c)),were calculated.Over time,in the SC and NSC treatments,K_(r)decreased from 0.516 to 0.120 s m^(-1)and 0.521 to 0.223 s m^(-1),respectively,whileτ_(c)increased from 0.49 to 4.42 Pa and 0.26-2.46 Pa,respectively.Variation in soil erosion resistance was rapid in the first one to two months,and then slowed down,with slight fluctuations afterwards.In the SC treatment,K_(r)was 42%lower andτ_(c)was 67%greater than those in the NSC treatment.Soil properties changed greatly for both treatments.SCT increased from 0 to 7.09 mm in the SC treatment.Coh increased from 2.91 to 9.04 kPa and 3.01-4.78 kPa in SC and NSC treatments,respectively.Both soil erosion resistance parameters could be well predicted by SCT and Coh in the SC treatment(R^(2)≥0.82),while their best predictor was Coh in the NSC treatment(R^(2)≥0.90).The results demonstrate that PSC formation enhances soil erosion resistance in the soil detachment process in the loess region under natural conditions.Our study revealed the important role and complexity of PSC in the process of soil erosion,and provided theoretical and data support for accurate understanding and prediction of soil erosion.

Plot investigation on rill flow resistance due to path tortuosity

The path tortuosity t is an indicator of rill morphology accounting for the deviation of the thalweg from a straight alignment.The effect of t on flow resistance has been little investigated for rills.This paper reports the results of a plot investigation aimed to establish the suitable accuracy of the rill thalweg measurement to determine the tortuosity parameter and to test the reliability of a theoretical flow resistance law.Four rills were incised in clay soil(CS)and clay-loam soil(LS)and shaped by a clear flow discharge.The three-dimensional Digital Terrain Models were created by the Structure from Motion technique.For rills on LS,an approximate thalweg was tracked by photo-interpretation,and a specific calculation routine was applied to identify the cross sections with a constant spacing d.The actual rill thalweg was obtained as the line joining the lowest points of these cross-sections.Among the different tested d values,d=0.075 m was chosen to determine t.For both CS and LS,the Darcy-Weisbach friction factorffeatured a non-monotonic relation with t,which was explained as the result of three additive components due to bed roughness,sediment transport,and localized energy losses due to curves.The effect of the former two components onff contrasts that of the third,resulting in a linearly decreasing f-t relationship and constant flow velocity for the three lowest tortuosity values,and an increased friction factor and reduced flow velocity for the highest tortuosity value.The flow resistance law was positively tested,and the predicted friction factor was dependent on t.

Structure-from-Motion Photogrammetry and Rare Earth Oxides can quantify diffuse and convergent soil loss and source apportionment

Accurately quantifying rates of soil erosion requires capturing both the volumetric nature of the visible,convergent fluvial pathways(also known as rills)and the subtle nature of the less-visible,diffuse pathways(interrill areas).The aim of this study was to use Rare Earth Oxide(REO)tracers and Structure-from-Motion(SfM)photogrammetry to elucidate retrospective information about soil erosion rates and sediment sources during different soil erosion conditions,within a controlled laboratory environment.The experimental conditions created erosion events consistent with diffuse and convergent erosion processes.REO tracers allowed the sediment transport distances of over 2 m to be described,and helped resolved the relative contribution of diffuse and convergent soil erosion;interrill areas were also iden-tified as a significant sediment sources soil loss under convergent erosion conditions.While the potential for SfM photogrammetry to resolve sub-millimetre elevations changes was demonstrated,under some conditions non-erosional changes in surface elevation,such as compaction,exceeded volumes of soil loss via diffuse erosion.The discrepancies between SfM Photogrammetry calculations and REO tagged sediment export were beneficial,identifying that during soil erosion events sediment in both aggregate and particle form is deposited within the convergent features,even when the rill extended the full length of the soil surface.The combination of SfM photogrammetry and REO tracers has provided a novel platform for building a spatial understanding of patterns of soil loss and source apportionment between rill and interrill erosion.

Experimental determination of sediment transport capacity of rill flow over sandified loess slope

Rill erosion is affected by the sand particle content in soil,especially in the wind and water erosion transition region of the Loess Plateau.The sediment transport capacity(STC)is a key parameter in rill erosion research,assessing the impact of aeolian sand intrusion on the STC of rill flow is of importance for a better understanding of rill erosion.This study aimed to assess the effect of aeolian sand intrusion on the STC on sandified loess slopes,with typical slopes and flow discharges,using a flume system which consisting of a sediment-feeding and a sediment-supply/settlement flume.The sediment feeding flume was jointed by 10°higher than that of the sediment measurement flume section.Three flow discharges(2,4,and 8 L min^(-1))and four slope gradients(5°,10°,15°,and 25°)were used to represent the natural hydrological conditions under three intrusion rates(SIR)of aeolian sands(10%,20%,and 50%).The results show that STC increased with slope gradient and flow discharge,and the relationship between the STC and the SIR was significantly affected by the slope gradient;the STCs decreased with the SIR on a slope of 5°but increased with the SIR on steep slopes of 15°-25°,implying a significant impact of slope gradient on the relationship between SIR and STC.The SIR of 50%resulted in the highest sediment concentration nearly 1200 kg m^(-3)on slopes of 25°.On sandified loess slopes of 10%,20%,and 50%SIR,the STC were about 30%,46%,and 57%higher than on loess slopes,indicating an increased erosion rate by sand particle intrusion into loess soil.These results highlight the impact of sand intrusion on STC of rill flow and provide deeper insights into the soil loss process on the sandified loess slope.

Sediment load change with erosion processes under simulated rainfall events

It is of great significance to quantify sediment load changing with erosion processes for improving the precision of soil loss prediction. Indoor rainfall experiments were conducted in 2 rainfall intensities(90 mm·h^(-1) and 120 mm·h^(-1)), four slope gradients(17.60%, 26.80%, 36.40%, 46.60%) and 2 slope lengths(5 m, 10 m). Erosion processes are divided into five stages. Results show that sediment yield is mainly sourced from rill erosion, contributing from 54.60% to 95.70% and the duration of which is extended by slope gradients. Sediment load and sediment concentration are significantly different along erosion stages, with the highest values in rill development stage(SIV). Surface flow velocities(interrill and rill) demonstrate less significant differences along erosion stages. Rainfall intensity increases sediment load in all stages, with up to 12.0 times higher when changing from 90 to 120 mm·h^(-1). There is an increasing trend for sediment load and sediment concentration with the rising slope gradient, however, fluctuations existed with the lowest values on 26.80% and 36.40%, respectively, among different treatments. The slope gradient effects are enhanced by rainfall intensity and slope length. Results from this study are important for validating and improving hillslope erosion modelling at each erosion stage.