Modeling catchment sediment yield: a genetic programming approach

Hydrologic processes are complex, non-linear, and distributed within a watershed both spatially and temporally. One such complex pervasive process is soil erosion. This problem is usually approached directly by considering the sediment yield. Most of the hydrologic models developed and used earlier in sediment yield modeling were lumped and had no provision for including spatial and temporal variability of the terrain and climate attributes. This study investigates the suitability of a recent evolutionary technique, genetic programming (GP), in estimating sediment yield considering various meteorological and geographic features of a basin. The Arno River basin in Italy, which is prone to frequent floods, has been chosen as case study to demonstrate the GP approach. The results of the present study show that GP can efficiently capture the trend of sediment yield, even with a small set of data. The major advantage of the GP analysis is that it generates simple parsimonious expression offering some possible interpretations to the underlying process.     

Relating soil erosion and sediment yield to geomorphic features and erosion processes at the catchment scale in the Spanish Pre-Pyrenees

Erosion and sediment redistribution are important processes in landscape changes in the short and long term. In this study, the RMMF model of soil erosion and the SEDD model of sediment delivery were used to estimate annual soil loss and sediment yield in an ungauged catchment of the Spanish Pre-Pyrenees and results were interpreted in the context of the geomorphic features. The Estaña Catchment is divided into 15 endorheic sub-catchments and there are 17 dolines. Gullies and slopes were the main erosive geomorphic elements, whereas the colluvial, alluvial, valley floor, and doline deposits were depositional elements. Spatially distributed maps of gross soil erosion, sediment delivery ratio (SDR), and sediment yield (SY) were generated in a GIS. Severe erosion rates (>100 Mg ha^?1 year^?1) were found in gullies, whereas mean and maximum erosion rates were very high on slopes developed on Keüper Facies and high in soils on Muschelkalk Facies. Where crops are grown, the depositional-type geoforms were predicted by the models to have an erosive dynamic. Those results were consistent with the rates of erosion quantified by ¹³⁷Cs which reflects the significant role of human activities in triggering soil erosion. Catchment area was positively correlated with erosion rate, but negatively correlated with SDR and SY. The latter were negatively correlated with the proportion of the surface catchment covered with forests and scrublands. The topography of the area influenced the high SDR and SY in the dolines and valley floors near the sinks. Intra-basin stored sediment was 59.2% of the total annual eroded soil in the catchment. The combination of the RMMF and SEDD models was an appropriate means of assessing the effects of land uses on soil erosion and obtaining a better understanding of the processes that underlie the geomorphic changes occurring in mountainous environments of the Mediterranean region.     

流域产沙模型的研究

由坡面泥沙颗粒的动力平衡条件,推导出坡面上细颗粒泥沙的起动切应力,由坡面径流的剩余输沙能力,导得坡面径流的侵蚀量计算公式,与文献[4]的雨滴溅蚀量计算公式一起,构成了概念清楚、过程明确、结构完整的流域产沙模型。产沙量取决于水流挟沙能力与可供沙量的对比关系,若水流挟沙能力小于供沙量,则产沙量等于水流挟沙能力;反之,水流将进一步冲刷表土,形成径流侵蚀,产沙量就等于供沙量与径流侵蚀量之和。据此,可计算出坡面和流域产沙量。模型在岔巴沟流域上应用,结果令人满意。     

坡面土壤侵蚀公式的建立及其在流域产沙计算中的应用

分析了坡面土壤颗粒的受力情况,认为坡面土壤颗粒的运动与其所受的有效切应力有关,从而推导出坡面土壤侵蚀量计算公式,建立了流域产沙模型。公式的推导具有较强的泥沙运动力学基础,反映了影响流域产沙的主要因素。模型参数少,应用方便,既可计算一次洪水产沙量,又可计算年产沙量。在陕北裴家峁流域上应用,结果令人满意。     

GIS-based lake sediment budget estimation taking into consideration land use change in an urbanizing catchment area

The objective of this study was to assess the lake sediment budget of land use changes using the Universal Soil Loss Equation (USLE), sediment delivery ratio (SDR), and trap efficiency (TE). The geographic information system was combined with the USLE to estimate the soil erosion of the Lake Asan watershed. Spatial data for each of the USLE factors were obtained from the land use, soil, and 1/25,000 scale digital contour maps. Landsat-5 TM images were selected for analyzing soil erosion changes due to land use changes. The sediment yield to Lake Asan was estimated using the SDR and TE. The estimated sediment budget was compared with observed data from the Lake Asan watershed between 1974 and 2003. The total estimated annual mean sediment budgets from Lake Asan in 1986, 1992, and 2000 were 0.267, 0.301, and 0.339 × 10⁶ ton, respectively, with an average of 0.302 × 10⁶ ton. The average measured sediment budget was 3.15 × 10⁶ ton year^?1. The average estimated value shows reasonable agreement with the observed sediment balance. The average estimated and measured sediment budgets contain uncertainties due to both the methods and the approach used by the observers. The simulated results indicated that soil erosion in the Lake Asan watershed increased at a rate of approximately 2 % per year from 1986 to 2000 due to land use change. This study may be useful for managers to identify reservoir rehabilitation management methods for stable irrigation water supply.     

Spatial distribution of soil erosion and suspended sediment transport rate for Chou-Shui river basin

In this study, a Physiographic Soil Erosion–Deposition Model (PSED) is applied for better management of a watershed. The PSED model can effectively evaluate the key parameters of watershed management: surface runoff discharge, suspended sediment transport rate, quantity of soil erosion, and spatial distribution of soil erosion and deposition. A basin usually contains multiple watersheds. These watersheds may have complex topography and heterogeneous physiographic properties. The PSED model, containing a physiographic rainfall-runoff model and a basin scale erosion–deposition model, can simulate the physical mechanism of the entire erosion process based on a detailed calculation of bed-load transportation, surface soil entrainment, and the deposition mechanism. With the assistance of Geographic Information Systems (GIS), the PSED model can handle and analyze extremely large hydrologic and physiographic datasets and simulate the physical erosion process without the need for simplification. We verified the PSED model using three typhoon events and 40 rainfall events. The application of PSED to Chou-Shui River basin shows that the PSED model can accurately estimate discharge hydrographs, suspended sediment transport rates, and sediment yield. Additionally, we obtained reasonable quantities of soil erosion as well as the spatial distribution of soil erosion and deposition. The results show that the PSED model is capable of calculating spatially distributed soil erosion and suspended sediment transport rates for a basin with multiple watersheds even if these watersheds have complex topography and heterogeneous physiographic properties.     

Application of physiographic soil erosion–deposition model in estimating sediment flushing efficiency of empty storage

In this study, we developed a physiographic soil erosion–deposition model to simulate sediment yield from a watershed into Agongdian reservoir and sediment flushing to estimate the efficiency of empty flushing. The model was verified using data related to Typhoons Morakot and Fanapi. Thereafter, we calculated the sediment flushing efficiency of empty storage under the conditions of 1- and 2-day storms with seven return periods. The simulated results revealed that the amount of sediment yield from Joushui River watershed was approximately 70% on average, whereas that from Wanglai River watershed was approximately 30%. These results are consistent with those of a government research report, which suggested that the sediment yield figures from Joushui and Wanglai River watersheds were 72 and 28%, respectively. Furthermore, the efficiency of empty flushing was more than 55% when using the shaft spillway pipe, suggesting that the model can be applied to estimate sediment yield and flushing efficiency.     

由降雨侵蚀引起的山坡面水土流失的研究进展

20世纪末，坡面降雨水土流失的研究日益深入。经验性模型是根据实际资料应用统计相关的方法建立起来的，其结构简单，使用方便。物理过程模拟模型试图对流域内发生的侵蚀产沙过程进行概化和近似，并用数学方程描述流域上侵蚀产沙的主要物理过程，再用比较严格的数值解法计算水沙运动过程，同时人们对侵蚀机理的研究也得到了深入发展。     

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.     

Identification and prioritization of sub-watersheds for land and water management using InVEST SDR model: Rmelriver basin,Tunisia

Tunisia presents many favorable conditions for the outbreak of water erosion because of its climatic and physical characteristics. This phenomenon represents a serious threat to the natural resources of soil and water. The aim of the present study is to identify the most vulnerable areas in order to help managers implement an effective management program. Thematic layers and parameters were integrated in the InVEST (Integrated Valuation of Environmental Services and Trade-offs) SDR (Sediment Delivery Ratio). Soil loss and sediment yield were calculated by the model and compared to observed data. The Rmel river basin was divided into 17 sub-watersheds using the dam axis as the main outlet. Results reveal that approximately 60% of the basin presents soil loss more than 5 ton/ha/year. Soil erosion map demonstrates that soil erosion risk increases with increased slope gradient, especially in agricultural lands. Sub-catchment prioritizations have been fixed based on soil erosion risk. Results show that sub-catchment 16 presents the highest soil loss with a value of 65 ton/ha/year. Sub-catchment presenting high soil erosion risk needs to give a high priority in conservation planning.     

Sediment Yield Index mapping and prioritization of Madia subwatershed,Sagar District of Madhya Pradesh (India)

Assessment of soil loss through Sediment Yield Index (SYI) is important for watershed planning, prioritization, and development. In the absence of measured sediment data, SYI expressing the relative sediment yield from different basins work as a basis for grading another basin to adopt erosion control measures. An attempt was made to evaluate SYI in wider scale by using cost-effective tools like remote sensing and geographical information system (GIS). SYI was calculated for Madia subwatershed, which consists of 29 microwatersheds and located in Sagar District, Madhya Pradesh (M.P.) The IRS LISS III data and Shuttle Radar Topography Mission (SRTM) digital elevation models (DEM) of 90-m resolution were used to identify land use characteristics and geomorphometric analysis. Major land use was observed as agricultural land (24.7 %), water bodies (16.7 %), forest area (10.2 %), and settlement (21.3 %). In categorization, similar overall accuracy was observed for dense forest, barren land, settlement, and water bodies. The highest SYI with a value more than 20 was observed in microwatershed Mw6, Mw7, and Mw24, which comprises 33 % of the total watershed area. It gives the information about the watershed area that requires very high priority.     

An improved correlation model for sediment delivery ratio assessment

In the present work, a new simplified model approach for sediment delivery ratio (SDR) assessment is proposed. Modelling and assessing SDR is still an open question. Difficulties rise from the lack of sufficient data availability, on one side, and, on the other, from the inherent uncertainties including spatial variability and temporal discontinuity of the land cover, climatic, hydrological and geomorphological variables involved. The proposed SDR_SIM model tries to skip over the limitations observed in other models generally adopted. A comparison with two different selected models amongst the most widespread simplified models, i.e. the area-model and the slope-model, is showed in application to a wide range of catchments extensions across different landscapes of Italy. The SDR_SIM estimates were also evaluated against observed SDR over a validation dataset of 11 basins sparse over different regions of the world. The results showed the effectiveness of the proposed model approach based on easily available catchment parameters.     

Identifying sediment sources from the inter-gully area and gully area in a small watershed in the Loess Hilly Region of China

Understanding the sediment source is very significant for erosion control in small watersheds. On the Loess Plateau of China, over 110 thousands of check dams were constructed in the past 60 years, which played an important role in controlling soil loss and also kept much information of erosion and sediment yield in the past. The objective of this investigation is to identify the sediment source by the ¹³⁷Cs tracing method in a small watershed in the Loess Hilly Region of China. Fifty-five sampling sites were selected from the watershed (44 from the inter-gully area, 7 from the gully sides and 4 from the reference sites), and a total of 114 soil samples were collected from three sediment profile cores at the Sidizui watershed check dam constructed in 1959. Based on the erosion rate from the inter-gully area by the ¹³⁷Cs models, and the total erosion amount deposited in the check dam since 1963, the relative contributions of sediment from the inter-gully area and gully area were estimated during 1963–2013. By comparing the difference of ¹³⁷Cs activities of surface soils from the check dam (deposited in the 2011–2013 flood events), the inter-gully and gully areas, the relative contributions of sediment derived from the two source areas to the flood sediment during recent years (2011–2013) were estimated by a simple mixing model. Results showed that the erosion rate from the inter-gully area was about 3054 t/km² a during 1963–2013, and the relative contributions of sediment from the inter-gully area and gully area were estimated to be 27 and 73%, respectively. The sediment from the inter-gully area was about 20% of the total sediment yield amount to the 2011–2013 floods, and from the gully area it was about 80%. The relative contribution of sediment from the inter-gully or gully area was not a fixed value in the watershed. Both of the inter-gully erosion and gully erosion should be simultaneously controlled, and more erosion control measures on the gully area should be taken in the subsequent watershed management in order to reduce the erosion amount in this region.     

Effects of landscape patterns on soil erosion processes in a mountain–basin system in the North China

This study was taken up to investigate the effects of landscape patterns on the soil erosion processes in a mountain–basin watershed. The revised universal soil loss equation and sediment delivery distribution models were used to estimate the soil erosion processes. The landscape patterns include the landscape metrics at the landscape level, landscape composition and configuration indicators on the basis of source–sink landscape theory. In the study area, the grassland, bare land, farmland and construction land were the sediment-source landscape; the forest and shrub were the sediment-sink landscape. The correlation analysis results showed that the soil erosion processes were significantly associated with the landscape patterns of the study area. At the landscape level, fragmentation metric was positively correlated with soil erosion; diversity metric was negatively related to soil erosion and sediment yield at the sub-basin scale. Among the source–sink landscape composition and configuration indicators, the composition indicator was positively correlated with soil erosion rate and sediment yield. In the configuration landscape indices, the shape index was negatively correlated with soil erosion rate and sediment yield; the fragmentation index was positively correlated with soil erosion rate and negatively correlated with sediment delivery rate. These results indicated that the optimization measures, such as increase in the area, connectivity and regularity of sediment-sink landscape, or decrease in the proportion, connectivity and regularity of sediment-source landscape, were favorable for soil conservation. Furthermore, the landscape indicators based on the source–sink theory could provide more information for landscape pattern optimization to reduce soil erosion.     

Evaluation of sediment yield and soil loss by the MPSIAC model using GIS at Golestan watershed, northeast of Iran

Watershed degradation due to soil erosion and sedimentation is considered to be one of the major environmental problems in Iran. In order to address the critical conditions of watershed degradation in arid and semiarid regions, a study based on the Modified Pacific Southwest Inter-Agency Committee (MPSIAC) model was carried out at Golestan watershed, northeast of Iran. The model information layers comprising nine effective factors in erosion and sedimentation at the watershed site were obtained by digitalization and spatial interpolation of the basic information data in a GIS program. These factors are geology, soil, climate, runoff, topography, land cover, land use, channel, and upland erosion. The source data for the model were obtained from available records on rainfall and river discharge and sediment, topography, land use, geology, and soil maps as well as field surveys and laboratory analysis. The results of the MPSIAC model indicated that 60.75 % (194.4 km²) and 54.97 % (175.9 km²) of the total watershed area were classified in the heavy sedimentation and erosion classes, and the total basin sediment yield and erosion were calculated as 4,171.1 and 17,813.4 m³ km^?2 year^?1, respectively. In the sensitivity analysis, it was found that the most sensitive parameters of the model in order of importance were topography (slope), land cover and use, runoff, and channel erosion (R ²?=?0.92–0.94), while geology, climate (rainfall), soil, and upland erosion factors were found to have moderate effect to the model output (R ²?=?0.74–0.59).     

西南喀斯特区土壤侵蚀研究进展

西南喀斯特区土层浅薄、成土速率低等特点决定了其允许土壤流失量小,土壤一旦流失,极难恢复,土壤侵蚀及其造成的石漠化现象已成为制约该区可持续发展最严重的生态环境问题。文章首先明晰西南喀斯特区土壤侵蚀特征,从坡面、小流域和区域三个尺度上系统概括西南喀斯特区土壤侵蚀的相关研究进展。针对当前喀斯特区土壤侵蚀研究野外径流小区、小流域及区域空间尺度数据缺少和相关研究模型限制性强等不足,建议从不同尺度深入研究喀斯特区土壤侵蚀发生发展规律及时空演化格局,并结合高新遥感、地球物理探测技术及模型,同步监测坡面—小流域—区域土壤流失,对土壤侵蚀进行定量评估,结合不同空间尺度土壤侵蚀特征构建系统性水土保持生态恢复治理模式和监测系统评价体系。      

Soil erosion modeling of a Himalayan watershed using RS and GIS

Employing the remote sensing (RS) and geographical information system (GIS), an assessment of sediment yield from Dikrong river basin of Arunachal Pradesh (India) has been presented in this paper. For prediction of soil erosion, the Morgan-Morgan and Finney (MMF) model and the universal soil loss equation (USLE) have been utilized at a spatial grid scale of 100 m × 100 m, an operational unit. The average annual soil loss from the Dikrong river basin is estimated as 75.66 and 57.06 t ha⁻¹ year⁻¹ using MMF and USLE models, respectively. The watershed area falling under the identified very high, severe, and very severe zones of soil erosion need immediate attention for soil conservation.     

Soil erosion and sediment yield modeling using RS and GIS techniques: a case study, Iran

Water erosion is a serious and continuous environmental problem in many parts of the world. The need to quantify the amount of erosion, sediment delivery, and sediment yield in a spatially distributed form has become essential at the watershed scale and in the implementation of conservation efforts. In this study, an effort to predict potential annual soil loss and sediment yield is conducted by using the Revised Universal Soil Loss Equation (RUSLE) model with adaptation in a geographic information system (GIS). The rainfall erosivity, soil erosivity, slope length, steepness, plant cover, and management practice and conservation support practice factors are among the basic factors that are obtained from monthly and annual rainfall data, soil map of the region, 50-m digital elevation model, remote sensing (RS) techniques (with use of Normalized Difference Vegetation Index), and GIS, respectively. The Ilam dam watershed which is located southeast part of Ilam province in western Iran is considered as study area. The study indicates that the slope length and steepness of the RUSLE model are the most effective factors controlling soil erosion in the region. The mean annual soil loss and sediment yield are also predicted. Moreover, the results indicated that 45.25%, 12.18%, 12.44%, 10.79%, and 19.34% of the study area are under minimal, low, moderate, high, and extreme actual erosion risks, respectively. Since 30.13% of the region is under high and extreme erosion risk, adoption of suitable conservation measures seems to be inevitable. So, the RUSLE model integrated with RS and GIS techniques has a great potential for producing accurate and inexpensive erosion and sediment yield risk maps in Iran.     

Modeling impacts of sediment delivery ratio and land management on adsorbed non-point source nitrogen and phosphorus load in a mountainous basin of the Three Gorges reservoir area, China

Agricultural nonpoint source (NPS) pollution at the Three Gorges reservoir area in China has been increasingly recognized as a threat to aquatic environment in recent years due to the serious eutrophication problem. Adsorbed NPS pollution is one of the major forms of NPS pollution in mountainous regions, the essential of the adsorbed NPS pollution is soil loss. Thus, simple, highly sensitive and continuous methods are required to simulate and quantify sediments yield at watershed scales. It is imperative to construct an integrated model to estimate the sediment yield and adsorbed NPS pollution load. According to the characteristics of climate, hydrology, topography, geology, geomorphology and land use types in Three Gorges reservoir area, a GIS-based dynamic-integrated-distributed model of annual adsorbed NPS load was presented in view of impacts of the rainfall intensity, sediment delivery ratio (SDR) and land management, where the temporally dynamic-continuous model of annual sediment yield was established by modifying the revised Universal Soil Loss Equation (RUSLE), and the spatially integrated-distributed model of annual adsorbed NPS load was then developed via the correlation between sediment yield and adsorbed NPS load. Furthermore, a case study of the Jialing River basin in China was applied to validate the integrated model, the dynamic-distributed coupling among GIS technology, sediment yield model, and adsorbed NPS load model was achieved successfully. The simulation results demonstrate the following: (1) runoff and sediment are influenced greatly by rainfall intensity, SDR and vegetation cover; rainfall and land management show high sensitivities to the integrated model; the average annual adsorbed TN and TP pollution loads from 2006 to 2010 decreased by 76 and 74 % compared with the previous treatment (1990), respectively. (2) Spatio-temporal variations of adsorbed NPS nitrogen and phosphorus load are mainly related to different land use types and the background level of nutriments in topsoil; different land use types have different contribution rates; the largest contribution rates of adsorbed total nitrogen (TN, 58.9 %) and total phosphorus (TP, 53 %) loads are both from the dryland cropland. (3) The identification of critical source areas can help to implement the prevention and control measures aiming at the reduction of water environmental pollution. These results will provide useful and valuable information for decision makers and planners to take sustainable land use management and soil conservation measures for the control of sediment pollution in the Three Gorges reservoir area. The application of this model in the catchment shows that the integrated model may be used as a major tool to assess sediment yield risks and adsorbed NPS pollution load at mountainous watersheds.     

Investigation of RS and GIS techniques on MPSIAC model to estimate soil erosion

Soil erosion due to surface water is a standout among the serious threat land degradation problem and an hazard environmental destruction. The first stage for every kind of soil conservation planning is recognition of soil erosion status. In this research, the usability of two new techniques remote sensing and geographical information system was assessed to estimate the average annual specific sediments production and the intensity erosion map at two sub-basins of DEZ watershed, southwest of Lorestan Province, Iran, namely Absorkh and Keshvar sub-basins with 19,920 ha, using Modified Pacific Southwest Inter-Agency Committee (MPSIAC) soil erosion model. At the stage of imagery data processing of IRS-P6 satellite, the result showed that an overall accuracy and kappa coefficient were 90.3% and 0.901, respectively, which were considered acceptable or good for imagery data. According to our investigation, the study area can be categorized into three level of severity of erosion: moderate, high, and very high erosion zones. The amount of specific sediments and soil erosion predicted by MPSIAC model was 1374.656 and 2396.574 m³ km^?2 year^?1, respectively. The areas situated at the center and south parts of the watershed were subjected to significant erosion because of the geology formation and ground cover, while the area at the north parts was relatively less eroded due to intensive land cover. Based on effective of nine factors, the driving factors from high to low impact included: Topography > Land use > Upland erosion > Channel erosion > Climate > Ground cover > Soil > Runoff > Surface geology. The measured sediment yield of the watershed in the hydrometric station (Keshvar station) was approximately 2223.178 m³ km^?2 year^?1 and comparison of the amount of total sediment yield predicted by model with the measured sediment yield indicated that the MPSIAC model 38% underestimated the observed value of the watershed.