Parameters optimization based on the combination of localization and auto-calibration of SWAT model in a small watershed in Chinese Loess Plateau

This study simulated the watershed flow and sediment responses based on calibration of the SWAT model in the semi-arid Chinese Loess Plateau (LP) where soil erosion intensively occurs. After the model’s initiation and manual modification, a 7-year inconsecutively observed flow and sediment data from 1984 to 1990 was used to analyze the model’s application in the selected watershed called AJW in the Chinese LP region. The model procedure included sensitivity analysis, parameter calibration and model validation. The best parameter set was finally determined based on the combination of parameter localization and auto-calibration. Then the model was assessed for its accuracy based on the NSE estimation, resulting in 0.77 and 0.67 for calibration and 0.46 and 0.32 for validation on simulations for flow and sediment, respectively, which is a moderately satisfactory accuracy among the applications of the SWAT model. Annual watershed assessment on flow and sediment with the calibrated SWAT model resulted in a multiyear averaged annual runoff coefficient of about 2.7% and an erosion modulus of 797 t/(km²·a⁻¹) in the AJW, indicating a beneficial consequence from the implementation of the historical soil and water conservations.     

Parameter estimation of SWAT and quantification of consequent confidence bands of model simulations

Soil and Water Assessment Tool (SWAT) is a river basin scale model widely used to study the impact of land management practices in large, complex watersheds. Even though model output uncertainties are generally recognized to affect watershed management decisions, those uncertainties are largely ignored in model applications. The uncertainties of SWAT simulations are quantified using various methods, but simultaneous attempt to calibrate a model so as to reduce the uncertainty are seldom done. This study aims to use an uncertainty reduction procedure that helps calibrate the SWAT model. The shuffled complex evolutionary metropolis algorithm for uncertainty analysis is employed for this purpose, and is demonstrated using the data from the St. Joseph River basin, USA. The values of the performance indices, the r² and the Nash–Sutcliffe efficiency (NSE) for the simulations during calibration period was found to be 0.81 (same for r² and NSE) and 0.79 for validation period indicating a good simulation by the model. The results also indicate that the algorithm helps reduce the uncertainty (percentage of coverage?=?62% and average width?=?19.2 m³/s), and also identifies the plausible range of parameters that simulate the processes with less uncertainty. The confidence bands of simulations are obtained that can be employed in making uncertainty-based decisions on watershed management practices.     

Use of SWAT to determine the effects of climate and land use changes on streamflow and sediment concentration in the Purna River basin,India

A semi-distributed, physically based, basin-scale Soil and Water Assessment Tool (SWAT) model was developed to determine the key factors that influence streamflow and sediment concentration in Purna river basin in India and to determine the potential impacts of future climate and land use changes on these factors. A SWAT domain with a Geographical Information System (GIS) was utilized for simulating and determining monthly streamflow and sediment concentration for the period 1980–2005 with a calibration period of 1980–1994 and validation period of 1995 to 2005. Additionally, a sequential uncertainty fitting (SUFI-2) method within SWAT-CUP was used for calibration and validation purpose. The overall performance of the SWAT model was assessed using the coefficient of determination (R²) and Nash–Sutcliffe efficiency parameter (E_NS) for both calibration and validation. For the calibration period, the R² and E_NS values were determined to be 0.91 and 0.91, respectively. For the validation period, the R² and E_NS were determined to be 0.83 and 0.82, respectively. The model performed equally well with observed sediment data in the basin, with the R² and E_NS determined to be 0.80 and 0.75 for the calibration period and 0.75 and 0.65 for the validation, respectively. The projected precipitation and temperature show an increasing trend compared to the baseline condition. The study indicates that SWAT is capable of simulating long-term hydrological processes in the Purna river basin.     

Runoff and sediment yield modeling using SWAT model: case of Wadi Hatab basin,central Tunisia

This study presents an application of the model Soil and Water Assessment Tool (SWAT) to simulate daily and monthly water flow and sediment fluxes in the Wadi Hatab watershed (2200 km²) located in central Tunisia. The study basin is characterized by a significant climatic contrast, abrupt topography, and soil fragility, resulting thereby in flash floods and important water erosion rates. This alarming situation requires urgent interventions in order to preserve water and soil resources, implying the need for a decision tool for proper integrated management of the watershed. The model was calibrated and validated based on a comparison of simulated and observed flow rates at the basin outlet (hydrometric station Khanguet Zazia), during the periods 1987–1988 and 1989–1990, respectively. The comparison was based not only on visual inspection of the agreement between observed and simulated time series, but also on statistical parameters. Indeed, for the daily time step application, the Nash—Sutcliffe efficiency (NSE) values were 0.52 and 0.61, and the coefficient of determination (R²) was 0.54 and 0.61 for calibration and validation, respectively. As for the monthly time-step application, the obtained NSE values were 0.67 and 0.89 while R² values were 0.83 and 0.87 for calibration and validation, respectively. This clearly shows the reasonably good agreement between simulated and observed flow rates. In terms of erosion, the model gave sediment yield values ??of 1.15 and 5.37 t/ha/year during the periods of calibration and validation, respectively.     

Determination of water balance equation components in irrigated agricultural watersheds using SWAT and MODFLOW models : A case study of Samalqan plain in Iran

Increasing water demands,especially in arid and semi-arid regions,continuously exacerbate groundwater as the only reliable water resources in these regions.Samalqan watershed,Iran,is a groundwater-based irrigation watershed,so that increased aquifer extraction,has caused serious groundwater depletion.So that the catchment consists of surface water,the management of these resources is essential in order to increase the groundwater recharge.Due to the existence of rivers,the low thickness of the alluvial sediments,groundwater level fluctuations and high uncertainty in the calculation of hydrodynamic coefficients in the watershed,the SWAT and MODFLOW models were used to assess the impact of irrigation return flow on groundwater recharge and the hydrological components of the basin.For this purpose,the irrigation operation tool in the SWAT model was utilized to determine the fixed amounts and time of irrigation for each HRU(Hydrological Response Unit)on the specified day.Since the study area has pressing challenges related to water deficit and sparsely gauged,therefore,this investigation looks actual for regional scale analysis.Model evaluation criteria,RMSE and NRMSE for the simulated groundwater level were 1.8 m and 1.1%respectively.Also,the simulation of surface water flow at the basin outlet,provided satisfactory prediction(R²=0.92,NSE=0.85).Results showed that,the irrigation has affected the surface and groundwater interactions in the watershed,where agriculture heavily depends on irrigation.Annually 11.64 Mm3 water entered to the aquifer by surface recharge(precipitation,irrigation),transmission loss from river and recharge wells 5.8 Mm3 and ground water boundary flow(annually 20.5 Mm³).Water output in the watershed included ground water extraction and groundwater return flow(annually 46.4 Mm³)and ground water boundary flow(annually 0.68 Mm³).Overally,the groundwater storage has decreased by 9.14 Mm3 annually in Samalqan aquifer.This method can be applied to simulate the effects of surface water fluxes to groundwater recharge and river-aquifer interaction for areas with stressed aquifers where interaction between surface and groundwater cannot be easily assessed.     

考虑径流成分的雅鲁藏布江月径流模拟

通过构建雅鲁藏布江流域分布式水文模型SWAT,并以1989~1998年为率定期,1999~2005年为验证期对3个水文站(奴各沙、羊村、奴下)的月径流进行了模拟。从模型参数组里选取径流成分模拟不同的2组参数,以Nash-Sutcliffe效率系数(NS)、相关系数(R~2)及相对误差(PBIAS)为评价指标,来分析模型的模拟结果好坏。结果表明:率定期内,在径流成分1下奴各沙、羊村站和奴下站的R~2值高达0.9以上,NS系数分别为0.9、0.92和0.72,PBIAS的值为7.1%、18.5%及34.2%,但径流成分与实际情况出入较大。而径流成分2下,模拟得到的评价指标虽然不是很高,但径流成分比较合理。验证期总体结论一致。因此,建议在模型率定验证中要充分考虑径流成分的作用,从而提高模型模拟的精度。     

Satellite-based rainfall estimation and discharge measurement of Middle Indus River,Pakistan

The impacts of floods and droughts are intensified by climate change, lack of preparedness, and coordination. The average rainfall in study area is ranging from 200 to 400 mm per year. Rain gauge generally provides very accurate measurement of point rain rates and the amounts of rainfall but due to scarcity of the gauge locations provides very general information of the area on regional scale. Recognizing these practical limitations, it is essential to use remote sensing techniques for measuring the quantity of rainfall in the Middle Indus. In this research, Tropical Rainfall Measuring Mission (TRMM) estimation can be used as a proxy for the magnitude of rainfall estimates from classical methods (rain gauge), quantity, and its spatial distribution for Middle Indus river basin. In order to use TRMM satellite data for discharge measurement, its accuracy is determined by statistically comparing it with in situ gauged data on daily and monthly bases. The daily R ² value (0.42) is significantly lower than monthly R ² value (0.82), probably due to the time of summation of TRMM 3-hourly precipitation data into daily estimates. Daily TRMM data from 2003 to 2012 was used as input forcing in Soil and Water Assessment Tool (SWAT) hydrological model along with other input parameters. The calibration and validation results of SWAT model give R ² = 0.72 and 0.73 and Nash-Sutcliffe coefficient of efficiency = 0.69 and 0.65, respectively. Daily and monthly comparison graphs are generated on the basis of model discharge output and observed data.     

Simulation of nitrogen and phosphorus loads in the Dongjiang River basin in South China using SWAT

Population growth, urbanization, and intensified agriculture have resulted in mobilization of nitrogen and phosphorus, which is the main cause of river water quality deterioration. Environmental regulation has expedited the necessity for agricultural producers to design and implement more environmentally suitable practices. Therefore, there is a need to identify critical nutrients and their loss/transport potential. Watershed model can be used to better understand the relationship between land use activities/management and hydrologic processes/water quality changes that occur within a watershed. The objective of the study is to test the performance of the SWAT model and the feasibility of using this model as a simulator of water flow and nitrogen and phosphorus yields over the Dongjiang River basin in South China. Spatial data layers of land slope, soil type, and land use were combined with geographic information system (GIS) to aid in creating model inputs. The observed streamflow and sediment at Boluo station in the Dongjiang River basin were used to calibrate and validate the model. Time series plots and statistical measures were used to verify model predictions. Predicted values generally matched well with the observed values during calibration and validation (R ²≥0.6 and Nash-Suttcliffe Efficiency ≥0.5) except for underestimation of sediment peaks and overestimation of sediment valleys at Boluo. This study shows that SWAT is able to predict streamflow, sediment generation, and nutrients transport with satisfactory results.     

水文响应单元空间离散化及SWAT模型改进

水文响应单元(Hydrological Response Units,HRU)是SWAT模型模拟的基本单元,传统方法划分的水文响应单元在空间分布上不连续且难以确定其明确的空间位置,不能反映HRU间的相互作用和进行精确空间分析。利用GIS工具对土地利用和土壤类型数据进行概化处理,提出了HRU空间离散化的方法,实现了水文响应单元在空间上的准确定位。在此基础上,针对SWAT模型中同一子流域所有HRU采取相同延迟的弱点进行改进,并选择太湖地区西苕溪流域对改进的SWAT模型进行水文模拟验证。改进后,校正期港口站Nash效率系数E_NS(Nash-Sutcliffe Efficiency)从0.64提高到0.67,验证期E_NS系数从0.70提高到0.76。研究表明:修正后的SWAT模型更能反映流域的水文特征,可以达到非常好的效果,考虑到HRU距离因素的径流延迟更为准确地刻画径流过程。实现HRU空间离散化将为模型改进和更小尺度的空间分析提供数据基础。     

Evaluation of the SWAT model for water balance study of a mountainous snowfed river basin of Nepal

In this study, a semi-distributed hydrologic model Soil and Water Assessment Tool (SWAT) has been employed for the Karnali River basin, Nepal to test its applicability for hydrological simulation. Further, model was evaluated to carry out the water balance study of the basin and to determine the snowmelt contribution in the river flow. Snowmelt Runoff Model (SRM) was also used to compare the snowmelt runoff simulated from the SWAT model. The statistical results show that performance of the SWAT model in the Karnali River basin is quite good (p-factor = 0.88 and 0.88, for daily calibration and validation, respectively; r-factor = 0.76 and 0.71, for daily calibration and validation, respectively). Baseflow alpha factor (ALPHA_BF) was found most sensitive parameter for the flow simulation. The study revealed that the average annual runoff volume available at the basin outlet is about 47.16 billion cubic metre out of which about 12% of runoff volume is contributed by the snowmelt runoff. About 25% of annual precipitation seems to be lost as evapotranspiration. The results revealed that both the models, SWAT and SRM, can be efficiently applied in the mountainous river basins of Nepal for planning and management of water resources.     

Compilation simulation of surface water and groundwater resources using the SWAT-MODFLOW model for a karstic basin in Iran

Modeling of karstic basins can provide a better understanding of the interactions between surface water and groundwater, a more accurate estimation of infiltrated water amount, and a more reliable water balance calculation. In this study, the hydrological simulation of a karstic basin in a semiarid region in Iran was performed in three different stages. In the first stage, the original SWAT model was used to simulate surface-water flow. Then, the SWAT-MODFLOW conjunctive model was implemented according to the groundwater characteristics of the study area. Finally, due to the karstic characteristics of the region and using the CrackFlow (CF) package, the SWAT-MODFLOW-CF conjunctive model was developed to improve the simulation results. The coefficient of determination (R²) and the Nash-Sutcliffe efficiency coefficient (NSE) as error evaluation criteria were calculated for the models, and their average values were 0.63 and 0.57 for SWAT, 0.68 and 0.61 for SWAT-MODFLOW, 0.73 and 0.7 for SWAT-MODFLOW-CF, respectively. Moreover, the mean absolute error (MAE) and root mean squared error (RMSE) of the calibration for groundwater simulation using the SWAT-MODFLOW model were 1.23 and 1.77 m, respectively. These values were 1.01 and 1.33 m after the calibration of the SWAT-MODFLOW-CF model. After modifying the CF code and keeping the seams and cracks open in both dry and wet conditions, the amount of infiltrated water increased and the aquifer water level rose. Therefore, the SWAT-MODFLOW-CF conjunctive model can be proposed for use in karstic areas containing a considerable amount of both surface water and groundwater resources.

     

GIS-hydrological models for managing water resources in the Zarqa River basin

Jordan is part of the arid and semi-arid region of the Middle East, where water resources are known to be scarce. Thorough planning is inevitable and must be applied wisely, especially as it is known that drought conditions were evidenced for many years in Jordan.Surface water average accounts for 693 MCM/year out of which 359 MCM/year is base flow and 334 MCM/year flood flow. Therefore, management of surface water has been given great attention in Jordan because of the overexploitation of groundwater resources. Hydrological simulation models interfaced with Geographical Information Systems (GIS) were examined in this study. GIS interfaced hydrological models were considered as a major tool for surface water management at a watershed scale because they are capable of presenting the relationship between the spatial and hydrological features of the watershed in an efficient way. This study aimed at exploring the advantages of using GIS-based hydrological models as a water management tool to study the largest river basin in Jordan namely, the Zarqa River basin.In this study, the Spatial Water Budget Model (SWBM) and HEC-HMS / HEC-GeoHMS extension model were used. The models were calibrated and validated based on King Talal Reservoir inflow for a period of eight years. The calibration was performed for the most sensitive parameters. The calibration for the two models was performed for the years 1979, 1980, 1981 and 1982. Satisfactory results were obtained for both models with an R² of 0.90 and 0.85 for the HEC-HMS and SWBM model, respectively. Validation for the models was performed using the years 1993, 1994, 1995 and 1996, and the results were satisfactory with an R² of 0.75 and 0.80 for the SWBM and HEC-HMS model, respectively. Models after their calibration and verification can then be used to test scenarios related to climate change and/or land-use change at the watershed scale.     

岩溶地下河在SWAT中的概化方法——以毕节倒天河流域为例

岩溶地下河的存在严重影响了土壤和水评估工具（SWAT）在岩溶地区的普遍适用性。针对岩溶地下河在SWAT中的概化问题,提出基于数字高程模型（DEM）预处理结合SWAT流域自动识别功能的方法,将岩溶地下河暴露于地表,把岩溶区复杂的地表－地下二元结构简化为地表一元结构,并以贵州毕节地下河发育的倒天河流域为例进行应用。结果表明:① 对比未经概化地下河建立的模型,概化地下河建立的模型识别流域面积增大30.73%,子流域个数增加29.27%,水文响应单元（HRU）个数增加43.82%;② 参数取最大物理意义范围时,未经概化地下河建立的模型p因子=0.64,不满足建模条件,物理模型本身存在问题;③ 概化地下河建立的模型月步长模拟结果:校准期R2=0.96,NS=0.96,验证期R2=0.94,NS=0.93,月尺度模拟效果非常好。岩溶地下河概化方法使SWAT在流域划分方面更加合理,搭建的模型模拟更加合理。该研究拓展了SWAT模型在岩溶区的应用。      

Accessing the capability of TRMM 3B42 V7 to simulate streamflow during extreme rain events: Case study for a Himalayan River Basin

The paper examines the quality of Tropical Rainfall Monitoring Mission (TRMM) 3B42 V7 precipitation product to simulate the streamflow using Soil Water Assessment Tool (SWAT) model for various rainfall intensities over the Himalayan region. The SWAT model has been set up for Gandak River Basin with 41 sub-basins and 420 HRUs. Five stream gauge locations are used to simulate the streamflow for a time span of 10 years (2000–2010). Daily streamflow for the simulation period is collected from Central Water Commission (CWC), India and Department of Hydrology and Meteorology (DHM), Nepal. The simulation results are found good in terms of Nash–Sutcliffe efficiency \((\hbox {NSE}) {>}0.65\), coefficient of determination \((R^{2}) {>}0.67\) and Percentage Bias \(\hbox {(PBIAS)}{<}15\%\), at each stream gauge sites. Thereafter, we have calculated the PBIAS and RMSE-observations standard deviation ratio (RSR) statistics between TRMM simulated and observed streamflow for various rainfall intensity classes, viz., light (\({<}7.5 \, \hbox {mm}/\hbox {d}\)), moderate (7.5 to 35.4 mm/d), heavy (35.5 to 124.4 mm/d) and extremely heavy (\({>}124.4 \, \hbox {mm}/\hbox {d}\)). The PBIAS and RSR show that TRMM simulated streamflow is suitable for moderate to heavy rainfall intensities. However, it does not perform well for light- and extremely-heavy rainfall intensities. The finding of the present work is useful for the problems related to water resources management, irrigation planning and hazard analysis over the Himalayan regions.     

Assessment of soil erosion risk using SWAT model

Soil erosion is one of the most serious land degradation problems and the primary environmental issue in Mediterranean regions. Estimation of soil erosion loss in these regions is often difficult due to the complex interplay of many factors such as climate, land uses, topography, and human activities. The purpose of this study is to apply the Soil and Water Assessment Tool (SWAT) model to predict surface runoff generation patterns and soil erosion hazard and to prioritize most degraded sub-catchment in order to adopt the appropriate management intervention. The study area is the Sarrath river catchment (1,491 km²), north of Tunisia. Based on the estimated soil loss rates, the catchment was divided into four priority categories for conservation intervention. Results showed that a larger part of the watershed (90 %) fell under low and moderate soil erosion risk and only 10 % of the watershed was vulnerable to soil erosion with an estimated sediment loss exceeding 10 t?ha^?1?year^?1. Results indicated that spatial differences in erosion rates within the Sarrath catchment are mainly caused by differences in land cover type and gradient slope. Application of the SWAT model demonstrated that the model provides a useful tool to predict surface runoff and soil erosion hazard and can successfully be used for prioritization of vulnerable areas over semi-arid catchments.     

Evaluating the impact of management scenarios and land use changes on annual surface runoff and sediment yield using the GeoWEPP: a case study from the Lighvanchai watershed,Iran

This study was undertaken to evaluate land use change impact and management scenarios on annual average surface runoff (SR) and sediment yield (SY) using the GeoWEPP tool in the Lighvanchai watershed (located in northwestern Iran). Following a sensitivity analysis, the WEPP model was calibrated (2005–2007) and validated (2008–2010) against monthly observed SY and SR. The coefficient of determination (R ²), Nash–Sutcliffe efficiency (NSE), mean bias error (MBE), and root-mean-square error (RMSE) were applied to quantitatively evaluate the WEPP model. The results indicate a satisfactory model performance with R ² > 0.80 and NSE > 0.60. Therefore, the model for current land use (scenario 1) was run for a 30-year time period (1982–2011). The annual average of SR and sediment load were predicted as 93,584 m³/year and 4340 ton/year, respectively. To reduce the annual average surface runoff and sediment yield at the watershed scale, the second scenario (alfalfa cultivation with suitable tillage) and the third scenario (grassland development) as two management scenarios of land use changes were defined by identifying the critical hillslopes. The rate of SR and sediment load in the second scenario were 42,096 m³/year and 429 ton/year, respectively. For the third scenario, the model predictions were 30,239 m³/year and 226 ton/year, respectively. Compared to the first scenario, the reduction rates in annual average of sediment load were about 90 and 94%, respectively. Moreover, for the second and third management scenarios, the reduction rates in annual average of SR were about 55 and 67%, respectively.     

基于SUFI-2算法和SWAT模型的妫水河流域水文模拟及参数不确定性分析

敏感性分析和不确定性分析是分布式水文模型参数校准和模型应用的基础。本文以妫水河流域为例,基于SWAT模型和SUFI-2算法进行模型参数敏感性分析,结合手动调参和自动率定,通过SUFI-2的P因子和R因子进行模型不确定性分析,构建妫水河流域分布式水文模型。本次妫水河流域月尺度模拟中:率定期,确定系数R2=0.59,效率系数NSE=0.56;验证期,确定系数R2=0.82,效率系数NSE=0.80;P因子均大于0.5,R因子均小于0.5。结果表明,妫水河流域SWAT模型水文模拟效果较好。     

江苏固城湖流域1951-2000年农业非点源氮、磷输移的数值模拟研究

选择位于长江下游的固城湖流域作为研究区域,基于分布式流域水文模型SWAT,采用数值模拟的手段,反演了1951～2000年流域农业非点源氮、磷的输移规律。模拟结果与实测值的一致性反映了模型的良好模拟能力。比较两个时段的模拟结果发现,1981～2000年,流域农业非点源氮、磷年平均浓度和输移量分别为:总氮0.82mg/L和411.88×10³kg/a,总磷0.084mg/L和43.04×10³kg/a;远高于1951～1960年的氮、磷年平均浓度和输移量分别为:总氮0.22mg/L和49.55×10³kg/a,总磷0.036mg/L和7.67×10³kg/a。模拟主要反映了流域下垫面条件和农作物耕作模式对农业非点源氮、磷浓度及输移量的影响。     

Using Stable Isotopes to Assess the Contribution of Terrestrial and Riverine Organic Matter to Diets of Nearshore Marine Consumers in a Glacially Influenced Estuary

Terrestrial and marine ecosystems in Southeast Alaska are linked by the flow of freshwater from precipitation and glacial runoff, which transports nutrients and organic matter (OM) downstream to estuaries. We examined the contribution of terrestrial-riverine and marine OM to diets of fishes (N = 257, four species) and invertebrates (N = 90, six species) collected from glacially influenced estuaries in Southeast Alaska using multiple stable isotopes (δ¹³C, δ¹⁵N, and δ³⁴S). Multivariate analysis of similarity (ANOSIM) was used to quantify variation in stable isotope composition of consumers across 6 months and three sites with watersheds that differed in their glacier and forest composition. Fishes showed weak differences (ANOSIM R = 0.141) in stable isotope composition among sampling months, moderate differences (ANOSIM R = 0.375) among sites, and strong differences (ANOSIM R = 0.583) among species. Invertebrates showed moderate differences (ANOSIM R = 0.352) in stable isotope composition among sampling months and strong differences among sites (ANOSIM R = 0.710) and species (ANOSIM R = 0.858). We found the greatest differences in stable isotope composition between the two estuary sites with watersheds containing the highest and lowest glacial coverage, indicating that the contribution of allochthonous OM to consumer diets varies across watershed types. Invertebrates collected from the site with the lowest glacial coverage in the watershed were more depleted in δ¹³C and δ³⁴S, indicating higher use of terrestrial-riverine OM, than those at sites with higher watershed glacial coverage. High variation in stable isotope composition among species, months, and sites underscores the complexity of estuary food web responses to future glacier loss.     

The morphological characteristics of gully systems and watersheds in Dry-Hot Valley,SW China

Gully systems and watersheds are geomorphic units with clear boundaries that are relatively independent of basin landscapes and play an important role in natural geography. In order to explore the morphological characteristics of gully systems and watersheds in the Dry-Hot Valley [South West (SW) China], gullies are interpreted from online Google images with high resolution and watersheds are extracted from digital elevation model at a scale of 1:50,000. The results show that: (1) There are 17,382 gullies (with a total area of 1141.66 km²) and 42 watersheds in the study area. (2) The average gully density of the study area (D) is 4.29 km/km², gully frequency (F) is 14.39 gullies/km², the branching ratio (B) is 5.13, the length ratio (L) is 3.12, and the coefficient of the main and tributary gullies (M) is 0.06. The degree of gully erosion is strong to extremely strong, the main development intensity of gully erosion ranges from intense to moderate, and the type of gully system is tributary. (3) The watershed areas (A) are between 0.39 and 96.43 km², the relief ratio (R) is from 0.10 to 0.19, the circularity ratio (C) is from 0.30 to 0.83, the texture ratio (T) is from 0.82 to 39.35, and the dominant geomorphological texture type is fine. (4) There is a quantitative relationship between F and D:F?=?0.624D² (R?=0.84) and T is closely related to D, F, M (R²?>?0.7). A, R and C are related to M (R²?>?0.5). The development of gully systems is the result of coupling effects between multiple factors. In this area, the degree of erosion and the condition of the main and tributary gullies can be controlled by the degree of topographic breakage in the watershed, which provides some theoretical basis for the evaluation of gully erosion by the latter. In addition, the scale, relief, and shape have a significant impact on the locations of the main and tributary gullies. For tributary gullies, attention should be paid to the interception and control of runoff and sediment in the small confluence branches in order to prevent gully expansion and head advance. These features can inform the development of targeted measures for the control of soil erosion.