Thesis advisor: Noah P. Snyder / Global studies assert that anthropogenic activity now leads to disproportionately higher rates of landscape change compared with background geomorphic processes. This study explores the relative influence of anthropogenic, glacial, and geologic processes on erosion rates (E) in the northeastern United States (NEUS) by analyzing published erosion and sedimentation data across multiple methods and timescales. I compile erosion rates and sediment yields from records of stream gauging, reservoir sedimentation, lake sedimentation, cosmogenic nuclides in stream sediment, and thermochronology. These data serve as a comparison point for quantified volumes of sediment deposited in valley bottoms as a result of European settlement in the NEUS, where glacial history may influence the availability of erodible sediment and, as a result, the relative magnitude of deposited sediment. I hypothesize that E in the formerly glaciated region will be lower than unglaciated E over last century (stream gauging and reservoir sedimentation) timescales due to the erosive power of continental glaciation and resultant thin upland soils, and that there will be an increase in E evident over the last century as a result of human influence. 499 sites with location data were compiled across the NEUS, converted to erosion rate (mm/yr) and sediment yield (Ys; t km-2 yr-1), and analyzed using statistical z-tests to determine whether the population means are significantly different. Mean E from all record types across both the glaciated and unglaciated NEUS exhibits a range smaller than one order of magnitude (0.012-0.055 mm/yr), much less variable than order-of-magnitude differences reported by other researchers comparing modern and geologic erosion, both regionally and globally. Last century timescales exhibit higher E in the unglaciated region than the glaciated region, but only reservoir sedimentation shows a significant difference in E between regions (0.012 vs. 0.055 mm/yr; glaciated and unglaciated, respectively); stream gauging E did not exhibit a significant regional difference, likely due to the large basin sizes, short measurement timescales, and disproportionate spatial distribution of the measurements. E does not increase from geologic to last century timescales: late Quaternary (lake sedimentation and cosmogenic nuclide) records consistently yield lowest E, with geologic (thermochronology) records showing the highest E in both regions, perhaps indicating the relative importance of E over timescales during which major orogenies were occurring in the NEUS. The similarities in mean E and large range of the distributions of all timescales, however, point to the relative stability of E over time in the NEUS. / Thesis (MS) — Boston College, 2018. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Earth and Environmental Sciences.

Accurately predicting sediment delivery has been a long-standing problem in the field of water resource management. Many different watershed equations and models have been developed such as the Universal Soil Loss Equation (USLE), the Geo-spatial interface for the Water Erosion Prediction Program (GeoWEPP) and many more, however, these models have not always been able to reliably predict in-stream sediment loads. In this study, two scales, watershed and site level, are used to understand where sediment transported in-stream is being produced. At the watershed scale, USLE was used to estimate sediment yield and then different factors such as connectivity topographic indices were applied as discount factors in an attempt to improve these estimates. The different parameters were then compared to turbidity to determine the level of accuracy of each method. It was found that USLE is not able to predict in-stream turbidity levels in the study area watersheds in Virginia and North Carolina. An implicit assumption of USLE is that runoff is produced on steeper slopes and that sediment production occurs on these hillslopes. However, it was found that flatter-sloped areas were highly correlated with in-stream turbidity. It was also found that in-channel and site-specific parameters such as bank height/slope and level of confinement at higher flows were more accurate predictors of in-stream sediment levels. Overall, turbidity and in-stream sediment levels are not well predicted by models that employ USLE. The distribution of runoff source areas, and channel/bank properties appear to be good predictors of sediment production at the watershed scale. These results indicate that sediment production and transport, as conceptualized by common models and equations, often associate sediment source areas with geomorphic and hydrologic processes in ways that are not consistent with the results of this study. Our results show that sediment is most likely being sourced from the channels and in stream areas. / Master of Science / Predicting how sediment moves through a watershed has been a long-standing problem in the field of water resource management. There are many equations and models that have been developed to calculated the amount of sediment that exits a watershed; such as the Universal Soil Loss Equation (USLE), the Geo-spatial interface for the Water Erosion Prediction Program (GeoWEPP) and many more. However, these models have not always been reliable or accurate in their predictions. In this study, two scales, watershed and site level, are used to understand where sediment transported within streams is being produced. At the watershed scale, USLE was used to estimate sediment leaving a system and then different factors, with different approaches to the understanding of sediment movement, were applied as discount factors in an attempt to improve these estimates. The different values that were calculated were then compared to turbidity to determine the level of accuracy of each parameter. It was found that USLE is not able to predict in-stream turbidity levels in the study area watersheds in Virginia and North Carolina. An assumption of USLE is that runoff is produced on steeper slopes and that sediment erosion occurs on these steeper sloped areas. However, it was found that flatter-sloped areas were highly correlated with turbidity. It was also found that in-channel and site-specific parameters such as bank height/slope and the level of confinement at higher flows were more accurate predictors of turbidity. Overall, USLE and models that used USLE were not able to predict turbidity. The distribution of runoff source areas and channel/bank properties appear to be good predictors of turbidity at the watershed scale. These results indicate that sediment movement, as conceptualized by common models and equations, often associate sediment source areas with watershed level morphology and hydrology in ways that are not consistent with the results of this study. Our results show that sediment is most likely being produced from the channels and in stream areas.

La producción de sedimentos en zonas de montaña es un motivo de preocupación no sólo por la pérdida de la fertilidad de la capa superficial de suelo, sino también debido a sus efectos aguas abajo, tales como la deposición de sedimentos en embalses. En este trabajo, la producción de sedimentos se estudio a dos escalas espaciales: una cuenca pequeña y una grande en el noreste de los Pirineos orientales. La cuenca pequeña es Cal Rodó (4,2 km2), y la grande es la cuenca Alta Llobregat del Río Llobregat (504 km2), donde a la salida se encuentra el embalse de La Baells, la primera cuenca está anidado dentro de la última. En Cal Rodó, el objetivo fue determinar y comparar los intervalos de confianza de la carga de sedimentos en suspensión a escala de evento y a escala anual. Los datos utilizados abarcan 10 años (1996-2005). Estos incluyen muestras discretas de agua recogidas por un muestreador automático de agua, lecturas de turbidez proporcionada por un sensor de infrarojo de turbidez, un sensor de ultrasonidos, y mediciones de la lámina de agua. Todo conectado a un registrador de datos. La incertidumbre de producción de sedimentos fueron evaluados por simulaciones de Monte Carlo usando la función de distribución de la carga de sedimentos y modelación de residuales. La producción total para el periodo estudiado y estimada a través de la calibración de sensores de turbidez fue 17,217.0 Mg (4,1 Mg/ha/año) y su intervalo de confianza al 90% está entre 16,311.4 Mg y 18,395.8 Mg. Por otra parte la producción de sedimentos a través de las curvas de transporte de sedimento a escala anual fue de 103,441.0 Mg (24,6 Mg/ha/año), y su intervalo de confianza oscila entre 49,107.0 Mg y 254,803.0 Mg. Las conclusiones generales son que la incertidumbre de la carga de sedimentos por las curvas de transporte de sedimentos son amplias (> 300%) y que la producción de sedimentos se sobreestimo en un orden de magnitud. Una buena distribución de las muestras en todo el hidrograma es necesaria, a fin de desarrollar las curvas de transporte más fiable. La segunda fase de este estudio fue realizada en la cuenca Alta del Río Llobregat donde la cubierta del suelo es principalmente bosque de coníferas y pastos, con presencia de algunas zonas intensamente erosionadas (badlands). El promedio anual de erosión de suelo se estimó cuantitativamente a través de RUSLE integrado en SIG. Los datos tratados provienen de las precipitaciones registradas en ocho estaciones meteorológicas que abarcan 14 años (1991-2004), las propiedades físicas del suelo, el inventario de cubierta vegetal, las características de gestión del suelo y un modelo digital de elevaciones. Todo ello se utilizó como entradas (inputs) para generar cada uno de los factores de RUSLE, las imágenes tienen una resolución de 20 m. La producción de sedimentos se calculó mediante la aplicación de una tasa de entrega de sedimentos (sediment delivery ratio SDR) a los resultados obtenidos por RUSLE. También se llevaron a cabo análisis de sensibilidad para cada uno de los factores de RUSLE y una evaluación de la incertidumbre global de erosión de suelo. Adicionalmente diversos escenarios de erosión de suelos debido al Cambio Global fueron desarrollados. La producción de sedimentos media anual fue de 3,35 Mg/ha/año y el intervalo de confianza al 90% fue entre 0,95 y 13,7 Mg/ha/año. Estos resultados son similares a los resultados de un estudio previo de batimetría del embalse el cual fue de 4.54 Mg/ha/año, y su intervalo de confianza al 90% fue entre 4,29 Mg/ha/año y 4,79 Mg/ha/año. El mapa de riesgos de erosión muestra que pequeñas áreas altamente erosionables (5%) producen el 50% de sedimento. / A better understanding of soil erosion and sediment yield in mountainous areas is needed, for management purposes and to envisage uncertainties when developing soil erosion models. In this thesis, sediment yield estimates and associated uncertainties were studied at two spatial scales: a small catchment (Cal Rodó 4.2 km2) and a relatively large basin (Upper Llobregat Basin 504 km2), the former is nested within the latter, in the SE Pyrenees, Spain. In Cal Rodó stream which is characterized by an event based regime, is assessed the uncertainty of suspended sediment load (SSL) at event and annual scales. SSL in streams are currently estimated using diverse types of sediment rating curves (SRC) that allow the interpolation and extrapolation of the discrete observations of sediment concentrations to the whole range of recorded discharges. However, the accuracy of outcomes using SRCs in mountain event based streams is still unknown. The parameters of the non-linear relationship are obtained by a log transformation and then a bias correction factor is applied, based on the residual analysis. The uncertainty of SSL can be assessed by using Monte-Carlo approaches based on the simulation of residuals. Nevertheless, when these methods are going to be applied in streams with a regime characterized by events, it is necessary to verify if different events have different discharge-concentration relationships; if this occurs, the role of events must be taken into account in the load analyses, as they modify the structure of the residuals. Results showed that when SRCs are used for interpolation between samples, paying attention to the role of events allows to avoid nonlinear bias overcorrection and to reduce the uncertainty associated with the sediment load estimates. The sediment yield from annual and seasonal rating curves were 6.0 and 5.1 times larger than the estimated by sensor calibrations. Therefore a good distribution of samples throughout the hydrograph needs to be ascertained in order to develop reliable SRCs. The second part of this research deals with soil erosion and sediment yield assessment and its associated uncertainties in the Upper Llobregat River Basin. Land cover is mainly coniferous forest and pastures with some intensely eroded areas (badlands). Annual average soil erosion was quantitatively estimated with an integration of GIS and the Revised Universal Soil Loss Equation (RUSLE). Maps for each of the factors involved where obtained with a 20 m resolution. Rainfall dataset from eight weather stations and spanning 14 years, soil properties, land cover inventory, land management features and digital elevation model were used as resource datasets to generate each of the RUSLE factor maps. Annual average sediment yield was computed by applying a sediment delivery ratio to the results obtained by RUSLE, and this result was compared with an existing bathymetric survey. Uncertainty and sensitivity analyses were undertaken for each of the RUSLE factors, in order to assess its magnitude and determine which of the factors influences the most the soil loss estimate. Results show that the annual average sediment yield was 3.35 Mg km-2 yr-1 and its confidence interval lied between 0.95 and 13.7 Mg km-2 yr-1 with 90% confidence. These results are in agreement with results from reservoir bathymetric survey. Additional comparisons of estimated sediment yield were done with empirical methods such as PSIAC and FSM. Erosion risk maps showed that about 5% of the area generated 50% of soil loss. These areas are located in steep slopes and coincide in many cases with badlands which are near the streams. Finally, sediment yield changes under climate conditions for the late 21st Century where not relevant, whilst spreading agricultural activity as in the 1950's had a relevant increase in sediment yield.

The goal of this study is to develop a model using GIS to estimate the source and quantity of accumulated sediment in the Emory & Henry College (EHC) duck pond.  Located in the Highlands of Southwest Virginia, the 1,194 acre duck pond watershed consists primarily of agricultural, forested, and low density urban land uses. The Revised Universal Soil Loss Equation (RUSLE) and the Sediment Distributed Delivery (SEDD) prediction models were used to determine the quantity of eroded sediment and the sediment yield at the duck pond, respectively.  These models require numerous computations, which were performed at the watershed scale with the aid of ArcGIS software.  In ArcGIS the watershed was broken into a raster grid of approximately 5,200 discrete 100 foot by 100 foot grid cells. The resulting watershed erosion model identified two main sources of sediment: a cluster of farms relatively close to and east of the duck pond, and a harvested timber site north of the duck pond.  The model predicted that 1,076 tons of sediment are delivered into the duck pond annually. The estimated sediment yield was then compared to the estimated amount deposited between October 2011 and September 2012, as measured by a topographic survey.  The model prediction was found to be within a factor of 6.3x of the measured value.  The predicted and measured sediment yields as well as identified erosion sources can be used to develop a water quality improvement plan and to help alleviate the need for periodic dredging. / Master of Science

Erosion is a natural process that occurs when soil particles are detached from one site and transported to another by water or wind, and can occur naturally or be accelerated by humans. Sediment can cause direct mortality or reduce growth of fish and other aquatic resources, particularly larval fish and eggs. Three treatments consisting of compost and jute netting, crimped straw with native seeds, and jute netting and vegetation filter strip were used to evaluate loss of runoff water and sediment on steep slopes. Erosion plots were built on slopes of 27 percent and filled with soil. The treatments were applied in triplicate and irrigated at 14-day intervals. Each erosion plot was irrigated with approximately 79 gallons of water for 10 hours and runoff was collected in pre-weighed containers during each period. After each collection, runoff from each plot was measured and recorded by subtracting the weight of each empty container from the combined weight of water and sediment. Turbidity, pH, and electrical conductivity were measured, and the separate weights of runoff and dry sediment were determined. The combination of jute netting over a layer of commercially available compost was over 99 percent effective at reducing runoff and sediment loss as compared to the untreated control erosion plots. Jute netting combined with a vegetative filter strip of creeping wild rye was over 94 percent and 99 percent effective at reducing runoff and sediment loss when compared to the untreated control erosion plots, respectively. Runoff and sediment loss from plots treated with crimped straw and native seeds was significantly greater than that of the other treatments, yet this treatment was over still 80 percent effective at reducing runoff and nearly 97 percent effective at reducing sediment loss. The cost-effectiveness of each treatment was evaluated based on the cost of the total treatment compared to the amount of water each treatment prevented from leaving the site, using the control plots as a baseline. The combination of jute and vegetation filter strip cost approximately $0.47 per liter of water prevented from leaving the site and was almost as effective at reducing runoff and sediment loss as was jute combined with compost, which cost $1.04 per liter. While each of the three treatments significantly reduced runoff and sediment loss when compared to the barren control plots, the jute and filter strip treatment was the most cost-effective of the three treatments. All treatments were effective at decreasing runoff and sediment loss when compared to the control, though no significant difference in runoff was observed between the control and any of the treatments after ten weeks. Thus, erosion control BMPs should be implemented well before the first storm causing runoff in order to be most effective.

Erosion, a surface process, can be quantified over long-term (assumed to be the natural erosion rate of the landscape) and contemporary (modern) timeframes. My research used the rare cosmogenic isotope 10Be in sand and cobbles collected from rivers in southeastern Brazil (Santa Catarina and Rio de Janeiro states) and southwestern China (Yunnan province) to quantify long-term, background rates of erosion and sediment supply. These measurements will also increase number of such measurements in tropical and subtropical climates. I assessed the relationship between landscape parameters (topographic and climatic) and background erosion rates in order to understand factors related to erosion. My data from so far unsampled states in Brazil shows that background erosion rates range between 13 and 90 m/Myr. I found that mean basin slope (R2=0.73) and mean annual precipitation (R2=0.57) are strongly correlated to erosion rates. Steep, escarpment-draining basins in Brazil erode faster than lower gradient basins draining the highlands. Comparing the isotopic concentration of river sand and cobbles, my data show that these grain sizes are sourced from different parts of the landscape. I compiled all published Brazilian cosmogenic 10Be data, and compared them to erosion rates from similar tectonic settings. While the erosion rates in Brazil are relatively low, they are similar to those in southeastern North America, but faster than rates measured on escarpments in southern Africa. In China, I tested the human effects on denudation by comparing long-term erosion rates derived from in-situ 10Be concentration and the modern sediment yield of 22 watersheds in Yunnan. Background erosion rates range between 17 and 386 m/Myr; long term sediment yields based on these erosion rates range from 79 to 893 tons km-2 yr-1. Modern sediment yields range from 90 to 2,879 tons km-2 yr-1 (data from Schmidt et al., 2011). In most watersheds, the modern sediment yield is 2-3X higher than long-term rates, likely the effect of a long history of land use in Yunnan. I found a statistically significant, positive relationship between erosion rates and both area (R2 = 0.60) and mean basin slope (R2 = 0.42). There is a negative but strong relationship between erosion rates and precipitation in my dataset (R2 = 0.60). I sampled some places where 10Be samples had been collected before to test the methodological assumption of time-invariant 10Be concentration. Concentrations generally agree on samples taken 6 months apart and in samples from the active channel and from floodplains, but not in samples collected a decade and centuries apart.

abstract: Human endeavors move 7x more volume of earth than the world’s rivers accelerating the removal of Earth’s soil surface. Measuring anthropogenic acceleration of soil erosion requires knowledge of natural rates through the study of 10Be, but same-watershed comparisons between anthropogenically-accelerated and natural erosion rates do not exist for urbanizing watersheds. Here I show that urban sprawl from 1989 to 2013 accelerated soil erosion between 1.3x and 15x above natural rates for different urbanizing watersheds in the metropolitan Phoenix region, Sonoran Desert, USA, and that statistical modeling a century of urban sprawl indicates an acceleration of only 2.7x for the Phoenix region. Based on studies of urbanization’s erosive effects, and studies comparing other land-use changes to natural erosion rates, we expected a greater degree of urban acceleration. Given that continued urban expansion will add a new city of a million every five days until 2050, given the potential importance of urban soils for absorbing anthropogenically-released carbon, and given the role of urban-sourced pollution, quantifying urbanization’s acceleration of natural erosion in other urban settings could reveal important regional patterns. For example, a comparison of urban watersheds to nearby non-urban watersheds suggests that the Phoenix case study is on the low-end of the urban acceleration factor. This new insight into the urban acceleration of soil erosion in metropolitan Phoenix can help reduce the acute risk of flooding for many rapidly urbanizing desert cities around the globe. To reduce this risk, properly engineered Flood Control Structures must account for sediment accumulation as well as flood waters. While the Phoenix area used regional data from non-urban, non-desert watersheds to generate sediment yield rates, this research presents a new analysis of empirical data for the Phoenix metropolitan region, where two regression models provide estimates of a more realistic sediment accumulation for arid regions and also urbanization of a desert cities. The new model can be used to predict the realistic sediment accumulation for helping provide data where few data exists in parts of arid Africa, southwest Asia, and India. / Dissertation/Thesis / Doctoral Dissertation Geography 2019

Changes in landscape pattern caused by changes in forest management, namely the Sustainable Forestry Initiative (SFI), and the implications of these structural changes on landscape processes were analyzed. Landscape structure was studied based upon the comparison of landscapes with different management histories. Ecological processes were analyzed based upon simulation of stand and landscape attributes of habitats for several vertebrate species and upon simulation of hydrological processes such as water and sediment yield. A methodology to integrate landscape and stand pattern and dynamics with landscape processes was developed for this work. It integrates a forest landscape structure model, several stand level growth and yield models, vertebrate habitat models, and a hydrological model. The comparisons among landscapes revealed that forest management has a strong influence on landscape structure. The SFI program increases fragmentation of the landscape indicated by the presence of more and smaller patches, more edges, more complex shapes, and less and smaller core areas. Traditional intensive and extensive management show comparable patterns characterized by high aggregation and connectivity. Landscapes managed according to the SFI program show higher Habitat Suitability Index (HSI) values for American woodcock, American beaver, wild turkey, fox squirrel, and gray squirrel. HSI is higher for pine warbler in the landscape not managed according to the SFI program. Downy woodpecker and barred owl present very reduced HSI values in either landscape. The SFI program induced fragmentation of the habitat of pine warbler and the establishment of narrow and elongated habitats in a network structure for the remaining species. Both patterns are determined by SMZs. The scenario representing management according to the SFI program presents higher sediment yield at the watershed level than the scenario representing management not according to the SFI program due to higher channel erosion related to the absence of buffer strips in the non-SFI scenario. In general, management according to the SFI program increases landscape diversity and evenness, habitat suitability for most species, potential vertebrate diversity, and provides habitat structure suitable for most species. This management also decreases sediment loss at the watershed level.

High rates of soil loss and high sediment loads in rivers necessitate efficient monitoring and quantification methodologies so that effective land management strategies can be designed. Constructing a sediment budget is a useful approach to address these issues. Quantifying a sediment budget using classical field-based techniques, however, is labour intensive and expensive for poorly gauged, large drainage basins. The availability of global environmental datasets in combination with GIS techniques provides an opportunity for studying large basins. Following this approach, a framework is presented for constructing sediment budgets for large, data-sparse drainage basins, which is applied to the mountainous upper Indus River basin in northern Pakistan. The methodological framework consists of five steps: (1) analyzing hydro-climatological data for dividing the drainage basin into characteristic regions, and calculating sediment yields; (2) investigation of major controls on sediment yields; (3) identification and mapping of sediment source areas by spatially distributed modelling of erosional processes; (4) spatially distributed modelling of sediment yields; and (5) carrying out the sediment budget balance calculation at the basin outlet. Further analysis carried out on the Indus data has enabled a better understanding of sediment dynamics in the basin.<p> Analysis of the available hydro-climatological data indicates that the basin can be subdivided into three characteristic regions based on whether runoff production and subsequent sediment generation is controlled by temperature (Region 1, upper, glacierized sub-basins), precipitation caused by the monsoon and western disturbances (Region 3, lower sub-basins), or a combination of the two (Region 2, middle reach sub-basins). It is also demonstrated that contrary to the conventional model, the specific sediment yield increases markedly with drainage area along the Indus River. An investigation of major controls on specific sediment yield in the basin indicates that percent snow/ice cover is a major land cover control for specific sediment yield. Spatially distributed erosion modelling predictions indicate that 87% of the annual gross erosion takes place in the three summer months with greatest erosion potential concentrated in sub-basins with high relief and a substantial proportion of glacierized area. Lower erosion rates can be explained by the arid climate and low relief on the Tibetan Plateau, and by the dense vegetation and lower relief in the lower monsoon sub-region. The model predicts an average annual erosion rate of 3.2 mm/a or 868 Mt/a. Spatially distributed sediment yield predictions made with coupled models of erosion and sediment delivery indicate that the Indus sub-basins generally show an increase of sediment delivery ratio with basin area. The predicted annual basin sediment yield is 244 Mt/a and the overall sediment delivery ratio in the basin is calculated as 0.28. The long-term mean annual sediment budget, based on mass balance, is characterized by a gross erosion of 762.9, 96.7 and 8.4 Mt, and a gross storage of 551.4, 66.1, and 6.5 Mt in the upper, middle, and lower regions of the basin, respectively. The sediment budget indicates that the major sources of eroded sediment are located in the Karakoram, in particular in the Hunza basin. Substantial sediment storage occurs on the relatively flat Tibetan Plateau and the Indus River valley reach between Partab Bridge and Shatial. The presented framework for sediment budget construction requires relatively few data, mostly derived from global datasets. It therefore can be utilized for other ungauged or poorly gauged drainage basins of the world.

High rates of soil loss and high sediment loads in rivers necessitate efficient monitoring and quantification methodologies so that effective land management strategies can be designed. Constructing a sediment budget is a useful approach to address these issues. Quantifying a sediment budget using classical field-based techniques, however, is labour intensive and expensive for poorly gauged, large drainage basins. The availability of global environmental datasets in combination with GIS techniques provides an opportunity for studying large basins. Following this approach, a framework is presented for constructing sediment budgets for large, data-sparse drainage basins, which is applied to the mountainous upper Indus River basin in northern Pakistan. The methodological framework consists of five steps: (1) analyzing hydro-climatological data for dividing the drainage basin into characteristic regions, and calculating sediment yields; (2) investigation of major controls on sediment yields; (3) identification and mapping of sediment source areas by spatially distributed modelling of erosional processes; (4) spatially distributed modelling of sediment yields; and (5) carrying out the sediment budget balance calculation at the basin outlet. Further analysis carried out on the Indus data has enabled a better understanding of sediment dynamics in the basin.<p> Analysis of the available hydro-climatological data indicates that the basin can be subdivided into three characteristic regions based on whether runoff production and subsequent sediment generation is controlled by temperature (Region 1, upper, glacierized sub-basins), precipitation caused by the monsoon and western disturbances (Region 3, lower sub-basins), or a combination of the two (Region 2, middle reach sub-basins). It is also demonstrated that contrary to the conventional model, the specific sediment yield increases markedly with drainage area along the Indus River. An investigation of major controls on specific sediment yield in the basin indicates that percent snow/ice cover is a major land cover control for specific sediment yield. Spatially distributed erosion modelling predictions indicate that 87% of the annual gross erosion takes place in the three summer months with greatest erosion potential concentrated in sub-basins with high relief and a substantial proportion of glacierized area. Lower erosion rates can be explained by the arid climate and low relief on the Tibetan Plateau, and by the dense vegetation and lower relief in the lower monsoon sub-region. The model predicts an average annual erosion rate of 3.2 mm/a or 868 Mt/a. Spatially distributed sediment yield predictions made with coupled models of erosion and sediment delivery indicate that the Indus sub-basins generally show an increase of sediment delivery ratio with basin area. The predicted annual basin sediment yield is 244 Mt/a and the overall sediment delivery ratio in the basin is calculated as 0.28. The long-term mean annual sediment budget, based on mass balance, is characterized by a gross erosion of 762.9, 96.7 and 8.4 Mt, and a gross storage of 551.4, 66.1, and 6.5 Mt in the upper, middle, and lower regions of the basin, respectively. The sediment budget indicates that the major sources of eroded sediment are located in the Karakoram, in particular in the Hunza basin. Substantial sediment storage occurs on the relatively flat Tibetan Plateau and the Indus River valley reach between Partab Bridge and Shatial. The presented framework for sediment budget construction requires relatively few data, mostly derived from global datasets. It therefore can be utilized for other ungauged or poorly gauged drainage basins of the world.