Search

The use of global climate model (GCM) precipitation simulations typically requires corrections for precipitation biases at subgrid spatial scales, typically at daily or monthly time scales. However, over many regions GCMs underestimate the magnitudes of multiyear precipitation extremes in the observed climate, resulting in a likely underestimation of the magnitudes of multiyear precipitation extremes in future scenarios. The objective of this study is to propose a method to extract from GCMs more realistic scenarios of multiyear precipitation extremes over time horizons of decades to one century. This proposed correction method is analogous to widely used bias correction methods, except that it is applied to variability at longer time scales than previous implementations (i.e., multiyear rather than daily or monthly). A case study of precipitation over a basin from the New York City water supply system demonstrates the potential magnitude of the underestimation of multiyear precipitation using uncorrected GCM scenarios, and the potential impact of the correction on multiyear hydrological extremes. Overall, it is a practical, conceptually simple approach meant for water supply system impact studies, but can be used for any impact studies that require more realistic multiyear extreme precipitation extreme scenarios. Significance Statement The purpose of this study is to present a practical method to address a particular difficulty that in some regions arises in climate change impact studies: global climate models tend to underestimate the multiyear variability of precipitation over some regions, resulting in an underestimation of the magnitudes and/or intensities of prolonged droughts as well as prolonged wet periods. The method is analogous to widely used bias correction methods, except it is applied to variability at longer time scales than previous implementations (i.e., multiyear rather than daily or monthly). It is designed to provide more realistic estimates of extreme hydrological scenarios during the twenty-first century. Our particular interest is for managers of water supply systems, but the method may be of interest to others for whom multiyear precipitation extremes are critical.

New York City (NYC)’s drinking water reservoirs supply over four billion liters each day to over nine million consumers in NYC and upstate communities. In the last 25 years, the City has invested nearly $2 billion in watershed protection programs (WPPs) to maintain high source water quality, allowing NYC to avoid filtration for 90% of the supply. This study involves the use of a model to evaluate the impact of WPPs on phosphorus (P) loading in the Cannonsville Reservoir watershed, one of the unfiltered water supply sources. The model is SWAT-Hillslope (SWAT-HS), a modified version of the Soil and Water Assessment Tool (SWAT) that can realistically predict variable source runoff processes. We applied the SWAT-HS model to this watershed to test its ability to simulate conditions observed after the implementation of watershed protection and evaluate the impact of point and non-point source WPPs on watershed export of P. When applied to a 12-year period of WPP implementation, SWAT-HS predicted streamflow very well with a daily Nash Sutcliffe Efficiency (NSE) of 0.85 at the calibrated outlet and values ranging from 0.56 to 0.78 at six other locations within the watershed. Moreover, the monthly predictions of soluble P (total dissolved P), particulate P, and sediment (total suspended solids) were good with NSE of over 0.73. Model simulations indicated that the dominant source of soluble P was pastures while particulate P originated from both from croplands and pastures. A significant quantity of P was derived from near-stream areas, particularly from pastures where cattle spent time grazing and had access to streams. SWAT-HS was also used to estimate what the P export would have been over this 12-year period without WPP implementation. Point and non-point source programs were found to be important for P control, with non-point source controls most effective during high streamflow, and point source controls more beneficial at low flow.

Climate model output is often downscaled to grids of moderately high spatial resolution (~4–6-km grid cells). Such projections have been used in numerous hydrological impact assessment studies at watershed scales. However, relatively few studies have been conducted to assess the impact of climate change on the hydrodynamics and water quality in lakes and reservoirs. A potential barrier to such assessments is the need for meteorological variables at subdaily time scales that are downscaled to in situ observations to which lake and reservoir water quality models have been calibrated and validated. In this study, we describe a generalizable procedure that utilizes gridded downscaled data; applies a secondary bias-correction procedure using equidistance quantile mapping to map projections to station-based observations; and implements temporal disaggregation models to generate point-scale hourly air and dewpoint temperature, wind speed, and solar radiation for use in water quality models. The proposed approach is demonstrated for six locations within New York State: four within watersheds of the New York City water supply system and two at nearby National Weather Service stations. Disaggregation models developed using observations reproduced hourly data well at all locations, with Nash–Sutcliffe efficiency greater than 0.9 for air temperature and dewpoint, 0.4–0.6 for wind speed, and 0.7–0.9 for solar radiation.

Study region: The New York City water supply watersheds Study focus: This study is a modeling analysis on climate change impact on streamflow using a stochastic weather generator (SWG), a hydrologic model, and downscaled future climate scenarios. Streamflow generated using synthetic time series of precipitation and air temperature from a SWG were compared to those simulated from observed historical and projected future weather. New hydrologic insights for the region: Synthetic weather was able to mimic the observed annual streamflow cycle for the six watersheds studied, including the seasonal pattern as well as magnitude and occurrence of extreme hydrologic events. Streamflow simulations using projected climate from 20 global climate models (GCM) for one of the New York City water supply watersheds indicate the potential for changes in the hydrologic regime in this region. The models indicate a shift in the timing of spring melt runoff from a distinct peak in late March and April under historical (1950–2009) conditions towards earlier in the year for mid-century (2041–2060) period. Results indicate that the region may experience an overall increase in mean streamflow in the future due to the combined effect of decreasing spring runoff peak and increasing streamflow during other seasons. More importantly, the magnitude and frequency of extreme hydrological events are projected to increase under future scenarios. These results have implications for future operation and management of the water supply. Keywords: Climate change, Extreme events, NYC water supply, Scenarios, Water resources

Uncertainty in hydrological modeling is of significant concern due to its effects on prediction and subsequent application in watershed management. Similar to other distributed hydrological models, model uncertainty is an issue in applying the Soil and Water Assessment Tool (SWAT). Previous research has shown how SWAT predictions are affected by uncertainty in parameter estimation and input data resolution. Nevertheless, little information is available on how parameter uncertainty and output uncertainty are affected by input data of varying complexity. In this study, SWAT-Hillslope (SWAT-HS), a modified version of SWAT capable of predicting saturation-excess runoff, was applied to assess the effects of input data with varying degrees of complexity on parameter uncertainty and output uncertainty. Four digital elevation model (DEM) resolutions (1, 3, 10 and 30 m) were tested for their ability to predict streamflow and saturated areas. In a second analysis, three soil maps and three land use maps were used to build nine SWAT-HS setups from simple to complex (fewer to more soil types/land use classes), which were then compared to study the effect of input data complexity on model prediction/output uncertainty. The case study was the Town Brook watershed in the upper reaches of the West Branch Delaware River in the Catskill region, New York, USA. Results show that DEM resolution did not impact parameter uncertainty or affect the simulation of streamflow at the watershed outlet but significantly affected the spatial pattern of saturated areas, with 10m being the most appropriate grid size to use for our application. The comparison of nine model setups revealed that input data complexity did not affect parameter uncertainty. Model setups using intermediate soil/land use specifications were slightly better than the ones using simple information, while the most complex setup did not show any improvement from the intermediate ones. We conclude that improving input resolution and complexity may not necessarily improve model performance or reduce parameter and output uncertainty, but using multiple temporal and spatial observations can aid in finding the appropriate parameter sets and in reducing prediction/output uncertainty.

Streamflow-based rating curves are widely used to estimate turbidity or suspended sediment concentrations in streams. However, such estimates are often inaccurate at the event scale due to inter- and intra-event variability in sediment-streamflow relationships. In this study, we use a quantile regression approach to derive a probabilistic distribution of turbidity predictions for Esopus Creek, a major stream in one of the watersheds that supply drinking water to New York City, using measured daily mean streamflow-turbidity data pairs for 2003 to 2016. Although a single regression curve can underpredict or overpredict the actual observation, quantile regression can estimate a range of possible turbidity values for a given value of streamflow. Regression relationships for various quantiles were applied to streamflows simulated by a watershed model to predict stream turbidity under: (i) the observed historical climate, and (ii) a future climate derived from 20 global climate model (GCM) scenarios. Future scenarios using quantile regression in combination with these GCMs and a stochastic weather generator indicated an increase in the frequency and magnitude of hydrological events that may generate high stream turbidity and cause potential water quality challenges to the water supply. The methods outlined in this study can be used for probabilistic estimation of stream turbidity for operational decisions and can be part of a vulnerability-based method to explore climate impacts on water resources.

Saturation-excess runoff is the major runoff mechanism in humid well-vegetated areas where infiltration rates often exceed rainfall intensity. While the Soil and Water Assessment Tool (SWAT) is one of the most widely used models, it predicts runoff based mainly on soil and land use characteristics, and is implicitly an infiltration-excess runoff type of model. Previous attempts to incorporate the saturation-excess runoff mechanism in SWAT fell short due to the inability to distribute water from one Hydrological Response Unit (HRU) to another. This paper introduces a modified version of SWAT, referred to as SWAT-Hillslope (SWAT-HS). This modification improves the simulation of saturation-excess runoff by redefining HRUs based on wetness classes, and by introducing a surface aquifer with the ability to route interflow from “drier” to “wetter” wetness classes. Mathematically, the surface aquifer is a non-linear reservoir that generates rapid subsurface stormflow as the water table in the surface aquifer rises. The SWAT-HS model was tested in the Town Brook watershed in the upper reaches of the West Branch Delaware River in the Catskill region of New York, USA. SWAT-HS predicted discharge well with a Nash-Sutcliffe Efficiency of 0.68 and 0.87 for daily and monthly time steps. Compared to the original SWAT model, SWAT-HS predicted less surface runoff and groundwater flow and more lateral flow. The saturated areas predicted by SWAT-HS were concentrated in locations with a high topographic index and were in agreement with field observations. With the incorporation of topographic characteristics and the addition of the surface aquifer, SWAT-HS improved streamflow simulation and gave a good representation of saturated areas on the date that measurements were available. SWAT-HS is expected to improve water quality model predictions where the location of the surface runoff matters.

The Soil Water Assessment Tool (SWAT) is employed throughout the world to simulate watershed processes. A limitation of this model is that locations of saturation excess overland flow in hilly and mountainous regions with an impermeable layer at shallow depth cannot be simulated realistically. The objective of this research is to overcome this limitation with minor changes in the original SWAT code. The new approach is called SWAT-with-impervious-layers (SWAT-wil). Adaptations consisted of redefining the hillslope length, restricting downward percolation from the root zone, and redefining hydrologic response units (HRUs) such that they are associated with the landscape position. Finally, input parameters were chosen such that overland flow from variable saturated areas (VSAs) corresponds to the variable source interpretation of the Soil Conservation Service (SCS) curve number runoff equation. We tested the model for the Town Brook watershed in the Catskill Mountains. The results showed that the discharge calculated with SWAT-wil agreed with observed outflow and results simulated with the original SWAT and SWAT-hillslope (SWAT-HS) models that had a surface aquifer that transferred water between groups of HRUs. The locations of the periodically saturated runoff areas were predicted by SWAT-wil at the right locations. Current users can utilize the SWAT-wil approach for catchments where VSA hydrology predominates.

In this study we investigated the impact of climate change on nutrient loading in the Cannonsville Reservoir watershed of the New York City (NYC) water supply system where management practices have reduced nutrient inputs in the last 25 years. A modified version of the SWAT hydrological and water quality model (SWAT-HS) that had been previously tested and verified for streamflow and phosphorus for this watershed was calibrated for nitrate to estimate contributions from point and nonpoint sources. Model simulations show that forests that occupy 64% of the watershed area contribute the greatest proportion of nitrate at 39%, while pastures that are in close proximity to streams and runoff generating areas contribute the greatest proportion of total nitrogen at 34%. Point sources contribute less than 5% of the annual nitrate load. Stream export accounts for only about 23% of the annual total N input to the watershed from anthropogenic sources, suggesting significant storage or loss from the landscape. We assumed stationary land use and management practices to assess the change in nutrient loading from baseline (2001–2010) to middle of the century (2051–2060) period due to a change in climate. Results indicated no change to moderate increase in the annual loading of dissolved forms of nutrients (N and P) whereas particulate forms of nutrients and sediment loadings are projected to increase due to an increase in the frequency and magnitude of large storm events. A seasonal shift in streamflow due to warmer winter temperatures, greater amounts of precipitation falling as rain, and earlier melting of snowpack may play an important role in controlling the seasonal pattern of nutrient loading. The methodology used in this study can be adapted in other watersheds to estimate the relative importance and partition contributions from various nonpoint sources to water quality, and to investigate the impacts of climate change.

Dissolved organic carbon (DOC) is not only a critical component of global and regional carbon budgets, but also an important precursor for carcinogenic disinfection byproducts (DBP) generated during drinking water disinfection process. The lack of process based watershed scale model for carbon cycling has been a limiting factor impeding effective watershed management to control DOC fluxes to source waters. Here, we integrated terrestrial and aquatic carbon processes into the widely tested Soil and Water Assessment Tool (SWAT) watershed model to enable watershed-scale DOC modeling (referred to as SWAT-DOC hereafter). The modifications to SWAT mainly fall into two groups: (1) DOC production in soils and its transport to aquatic environment by different hydrologic processes, and (2) riverine transformation of DOC and their interactions with particular organic carbon (POC), inorganic carbon and algae (floating and bottom). We tested the new SWAT-DOC model in the Cannonsville watershed, which is part of the New York City (NYC) water supply system, using long-term DOC load data (from 1998 to 2012) derived from 1399 DOC samplings. The calibration and verification results indicate that SWAT-DOC achieved satisfactory performance for both streamflow and DOC at daily and monthly temporal scales. The parameter sensitivity analysis indicates that DOC loads in the Cannonsville watershed are controlled by the DOC production in soils and its transport in both terrestrial and aquatic environments. Further model uncertainty analysis indicates high uncertainties associated with peak DOC loads, which are attributed to underestimation of high streamflows. Therefore, future efforts to enhance SWAT-DOC to better represent runoff generation processes hold promise to further improve DOC load simulation. Overall, the wide use of SWAT and the satisfactory performance of SWAT-DOC make it a useful tool for DOC modeling and mitigation at the watershed scale.

Uncertainty in hydrological and water quality modelling is of significant concern due to its effects on prediction and subsequent application in watershed management. Similar to other distributed hydrological models, model uncertainty is an issue in applying the Soil and Water Assessment Tool (SWAT). Previous research has shown how SWAT predictions are affected by uncertainty in parameter estimation and input data resolution. Nevertheless, little information is available on how parameter uncertainty and output uncertainty are affected by input data of varying complexity. In this study, SWAT-Hillslope (SWAT-HS), a modified version of SWAT capable of predicting saturation excess runoff was applied to assess the effects of input data with varying degrees of complexity on parameter uncertainty and output uncertainty. Four digital elevation model (DEM) resolutions (1, 3, 10 and 30 m) were tested for their ability to predict streamflow and saturated areas. In a second analysis, three soil maps and three land use maps were used to build nine SWAT-HS setups from simple to complex (fewer to more soil types/ land use classes), which were then compared to study the effect of input data complexity on model prediction/output uncertainty. The case study was the Town Brook watershed in the upper reaches of the West Branch Delaware River in the Catskill Region, New York, USA. Results show that DEM resolution did not impact parameter uncertainty or affect the simulation of streamflow at the watershed outlet but significantly affected the spatial pattern of saturated areas, with 10 m being the most appropriate grid size to use for our application. The comparison of nine model setups revealed that input data complexity did not affect parameter uncertainty. Model setups using intermediate soil/land use specifications were slightly better than the ones using simple information, while the most complex setup did not show any improvement from the intermediate ones. We conclude that increasing spatial input details may not necessarily improve model performance or reduce parameter and output uncertainty, but using multiple temporal and spatial observations can aid in finding the appropriate parameter sets and in reducing prediction/output uncertainty.

In-stream nutrient release and retention control the timing and quantity of export at the watershed outlet by mobilization and transport of phosphorus (P) sources from land to the channel, and remobilization of transient stores of P from stream beds. We investigated the significance of stream processes in regulating P loading to the Cannonsville watershed, NY, USA. A mass balance of estimated P inputs to the stream with observed P export at the watershed outlet was used to quantify P delivery and explore the behavior of P. Stream channel transport of both dissolved and particulate P is found to be non-conservative, with dissolved P being retained during low flows and particulate P released during high flows. The results suggest that differences in the magnitude and relative importance of in-stream biogeochemical processes under different flow regimes regulate P delivery in ways that may influence ecological impacts to downstream river reaches and reservoirs.

Stream turbidity levels tend to increase during high stream discharge events, and it is important to quantify the suspended sediment flux during these events that could potentially lead to water quality problems. Here, a case study for estimating suspended sediment loads (as a product of turbidity and stream discharge) in streams that are part of the New York City (NYC) water supply in the Catskill region of New York State is presented. Over the 8 year study period 80% of the suspended sediment load was transported during less than 4% of the time, indicating the importance of estimating storm event suspended sediment loads. The objective of this study was to understand the underlying factors controlling the uncertainty in the discharge vs turbidity relationship at the outlet of the watershed draining into the NYC Ashokan Reservoir. High frequency (15-min) automated monitoring of stream turbidity was combined with stream discharge measurements of a similar frequency to provide an estimate of the true suspended sediment load that could be used for model testing and verification at two time scales; daily and events. Multivariate statistical analyses indicate that average daily stream turbidity during storm events can be influenced by the spatial variability in runoff, antecedent conditions, and season. A predictive relationship of event mean stream turbidity based on stream discharge alone led to a strong predictive relationship (r2 = 0.81), but also a 10% underestimation of the cumulative measured event mean suspended sediment load. Inclusion of information on the time between events improved the regression equation (r2 = 0.89), and reduced the cumulative difference between estimated and measured event mean suspended sediment loads to 7% underestimation.

Recent works have indicated that climate change in the northeastern United States is already being observed in the form of shorter winters, higher annual average air temperature, and more frequent extreme heat and precipitation events. These changes could have profound effects on aquatic ecosystems, and the implications of such changes are less understood. The objective of this study was to examine how future changes in precipitation and temperature translate into changes in streamflow using a physically based semidistributed model, and subsequently how changes in streamflow could potentially impact stream ecology. Streamflow parameters were examined in a New York City water supply watershed for changes from model-simulated baseline conditions to future climate scenarios (2081-2100) for ecologically relevant factors of streamflow using the Indicators of Hydrologic Alterations tool. Results indicate that earlier snowmelt and reduced snowpack advance the timing and increase the magnitude of discharge in the winter and early spring (November-March) and greatly decrease monthly streamflow later in the spring in April. Both the rise and fall rates of the hydrograph will increase resulting in increased flashiness and flow reversals primarily due to increased pulses during winter seasons. These shifts in timing of peak flows, changes in seasonal flow regimes, and changes in the magnitudes of low flow can all influence aquatic organisms and have the potential to impact stream ecology.

Recent studies have indicated that potential future climate change may lead to changes in the timing and quantity of snowpack accumulation and winter (November to April) streamflow patterns including increased streamflow and turbidity in early winter and a slight reduction in turbidity loads at the time of traditional early spring run-off. In this study, we examine the potential effects of these predicted changes on reservoir turbidity levels. Our analysis focuses on Ashokan Reservoir, NY, which can at times receive significant watershed turbidity inputs mainly from stream channel erosion from the Esopus Creek. Both measured and simulated turbidity loads and climatology are input to CE-QUAL-W2 (W2), a reservoir turbidity transport model that can simulate reservoir turbidity and other water quality parameters. The W2 model is applied to estimate the effects of hydroclimatology on effective settling rates and turbidity transport as a result of differences in reservoir thermal structure during summer and winter events. Simulations suggest that the effective settling velocity is substantially lower at low temperatures during winter time. Winter average stream flow is simulated to increase by 12% and 20%, which leads to increases in reservoir turbidity by 11% and 17% for the future period 2046–2065 and 2081–2100, respectively. From a seasonal perspective, a change in timing of peak streamflow with increased flows during the winter and slightly reduced flows during early spring leads to increased average reservoir turbidity during winter and slightly decreased in-reservoir turbidity during early spring and summer. Copyright © 2013 John Wiley & Sons, Ltd.

Snow is a substantial component of historical annual precipitation in New York City (NYC) water supply watersheds in the Catskill Mountains, and the pattern of snow accumulation and snowmelt has important implications for the management of the reservoirs and watersheds that are part of the NYC water supply. NYC currently estimates reservoir basin-scale snowpack throughout the snow season by extrapolation from biweekly snow survey data. These estimates are complemented by the NOAA Snow Data Assimilation System (SNODAS) product. Snowpack models are used in short-term projections to support reservoir operations and long-term simulations to evaluate the potential effects of climate change, land use change, and watershed management on the water supply. We tested three snowpack estimation approaches compared with snow survey data: the lumped-parameter temperature index approach from the Generalized Watershed Loading Function (GWLF) watershed model, a spatially distributed temperature index (SDTI) model, and the spatially distributed NOAA SNODAS product. Of the spatially distributed approaches, SNODAS estimated the spatial variability of snow water equivalent (SWE) among snow survey sites within a basin better than the SDTI model. All three snowpack estimation approaches, including the lumped-parameter GWLF model, performed well in estimating basin-wide SWE for most of the basins studied. Copyright © 2013 John Wiley & Sons, Ltd.

This study focuses on the impact of changes in winter streamflow on in-stream turbidity in the Esopus Creek watershed, one of the New York City (NYC) water supply watersheds. Projected changes in daily precipitation and air temperature from a suite of five global climate models (GCMs) and three emission scenarios for future periods 2046–2065 and 2081–2100 were downscaled for the study region. The simulated climate scenarios were used to project future streamflows using the Generalized Watershed Loading Functions—Variable Source Area (GWLFVSA) watershed model. Seasonal turbidity rating curves based on measured historical streamflow and stream turbidity were used in combination with the simulated streamflow for generating future stream turbidity scenarios. Results indicate increase in future ambient stream turbidity from November to March and a decrease during April. These results are the effects of increased winter rainfall, reduced snowfall, and a shift to early timing of spring snowmelt runoff causing an increase in streamflow during early winter. It also suggests a reduction in the traditional peak streamflow around April that is expected to occur in this region. As a result, our models simulate a consistent increase in the low to medium percentile range of turbidity values associated with low to medium range of streamflows and no apparent change in high percentile turbidity values associated with high streamflows.

High suspended sediment loads and the resulting turbidity can impact the use of surface waters for water supply and other designated uses. Changes in fluvial sediment loads influence material fluxes, aquatic geochemistry, water quality, channel morphology, and aquatic habitats. Therefore, quantifying spatial and temporal patterns in sediment loads is important both for understanding and predicting soil erosion and sediment transport processes as well as watershed-scale management of sediment and associated pollutants. A case study from the 891 km 2 Cannonsville watershed, one of the major watersheds in the New York City water supply system is presented. The objective of this study was to apply Soil and Water Assessment Tool-Water Balance (SWAT-WB), a physically based semi-distributed model to identify suspended sediment generating source areas under current conditions and to simulate potential climate change impacts on soil erosion and suspended sediment yield in the study watershed for a set of future climate scenarios representative of the period 2081–2100. Future scenarios developed using nine global climate model (GCM) simulations indicate a sharp increase in the annual rates of soil erosion although a similar result in sediment yield at the watershed outlet was not evident. Future climate related changes in soil erosion and sediment yield appeared more significant in the winter due to a shift in the timing of snowmelt and also due to a decrease in the proportion of precipitation received as snow. Although an increase in future summer precipitation was predicted, soil erosion and sediment yield appeared to decrease owing to an increase in soil moisture deficit and a decrease in water yield due to increased evapotranspiration.

Information on the nature and relative contribution of different watershed sediment sources is recognized as a key requirement in the design and implementation of targeted management strategies for sediment control. A direct method of assessing sediment sources in a watershed that has attracted attention in recent years is sediment fingerprinting. The aim of this article is to describe the development of sediment fingerprinting as a research tool and to consider how the method might be transformed from a research tool to a management tool within a regulatory framework, with special reference to the United States total maximum daily load (TMDL) program. When compared with the current source assessment tools in developing sediment TMDLs, sediment fingerprinting offers considerable improvement as a tool for quantifying sources of sediment in terms of source type (e.g., channel vs. hillslope) as well as spatial location (subwatershed). While developing a conceptual framework for sediment TMDLs, we recognize sediment fingerprinting along with sediment budgeting and modeling as valuable tools in the TMDL process for developing justifiable sediment TMDLs. The discussions presented in this article may be considered as a first step toward streamlining the sediment fingerprinting approach for its wider application in a regulatory framework.

Snow is an important component of the water resources of New York State and the watersheds and reservoirs of New York City (NYC) water supply. In many of the NYC water supply watersheds the hydrologic regimes of high-elevation headwaters are linked to streamflow and channel processes in low-elevation stream reaches that serve as inputs to water supply reservoirs. To better simulate this linkage there is a need to understand spatial variations in snowpack and snowmelt. Snowmelt hydrology is an important component of the Soil and Water Assessment Tool (SWAT) model in watersheds where spring runoff is strongly affected by melting snow. This study compares model simulated snowpack and snowmelt at different elevation bands with snow survey data available for the Cannonsville reservoir watershed. Simulations examine the effects of parameterising the SWAT snowmelt sub-model using 1, 3, and 5 elevation bands by comparison with measured snow and streamflow. Comparison between measured and simulated snowpack produced correlation coefficients ranging from 0·35 to 0·85. Simulations of both daily and seasonal streamflow, improved when using 3 elevation bands with r2 of 0·73 and ENS of 0·72. Streamflow simulations showed slightly lower model performance when basin elevation was assumed to be equal to snow survey site elevation, due to the snow survey sites being somewhat biased toward lower elevations. The effect of climate change was also evaluated and showed that under higher air temperatures in future climate change scenarios, SWAT indicated more precipitation falling as rain, increased and earlier snowmelt, and a reduced snowpack leading to a change in the pattern of streamflow, particularly during winter and early spring. Copyright © 2011 John Wiley & Sons, Ltd.

This study focuses on the effect of projected changes in rainfall, snow accumulation and snowmelt, and consequent changes in the timing of runoff on NYC water supply system storage and operation as simulated by the NYC reservoir system Operational Analysis and Simulation of Integrated Systems (OASIS) model. The Generalized Watershed Loading Functions—Variable Source Area (GWLF-VSA)—watershed model is used with future climate scenarios derived from different General Circulation Models (GCMs) to simulate future inflows to reservoirs that are part of the New York City Water Supply System (NYCWSS). Future scenarios that use current system operation rules and demands, but changed reservoir inflows, suggest that changes in precipitation and snowmelt will affect regional water availability on a seasonal basis. The combined effect of projected increases in winter air temperatures, increased winter rain, and earlier snowmelt may result in more runoff during winter. This will cause reservoir storage levels, water releases and spills to increase during the winter and earlier reservoir refill in the spring. An overall increase in precipitation will result in a reduction in number of days the system is under drought conditions, despite increased evapotranspiration later in the year. Copyright © 2011 John Wiley & Sons, Ltd.

Streams can be classified as stable or unstable, depending on the stage of channel evolution. Many streams of the southern Piedmont in United States have high sediment loads and are listed as impaired under the total maximum daily load (TMDL) program and may be unstable. It is not clear as to what the target (reference) load or remediation measures should be for unstable streams. The objective of this study was to determine the relative channel stability for a typical southern Piedmont stream using rapid geomorphic assessments (RGAs) and sediment yield analysis. The results were supported through a sediment fingerprinting analysis. RGAs were performed along 52 reaches on the North Fork Broad River (NFBR) main stem and two tributaries. Annual sediment yields were calculated and compared with yields in the southern Piedmont for stable streams that are resilient to degradation or aggradation and unstable streams that are susceptible to such disturbances. Majority of the NFBR main stem was found to be unstable with signs of geomorphic instability in the form of degradation and aggradation. The estimated average annual sediment yield was 0·78 T ha−1 year−1. By comparison, the median annual yield is 0·20 T ha−1 year−1 for stable streams and 0·48 T ha−1 year−1 for unstable streams in the Piedmont ecoregion with comparable drainage basin size. We conclude that the NFBR is in an unstable stage of channel evolution. Sediment fingerprinting proved that majority of the stream-suspended sediment emanated from eroding stream channels. The methods outlined in this study have implications for the reference condition and remediation efforts related to stream turbidity and stream channel restoration. Copyright © 2010 John Wiley & Sons, Ltd.

Thousands of stream miles in the southern Piedmont region are impaired because of high levels of suspended sediment. It is unclear if the source is upland erosion from agricultural sources or bank erosion of historic sediment deposited in the flood plains between 1830 and 1930 when cotton farming was extensive. The objective of this study was to determine the source of high stream suspended sediment concentrations in a typical southern Piedmont watershed using sediment fingerprinting techniques. Twenty-one potential tracers were tested for their ability to discriminate between sources, conservative behavior, and lack of redundancy. Tracer concentrations were determined in potential sediment sources (forests, pastures, row crop fields, stream banks, and unpaved roads and construction sites), and suspended sediment samples collected from the stream and analyzed using mixing models. Results indicated that 137Cs and 15N were the best tracers to discriminate potential sediment sources in this watershed. The delta15N values showed distinct signatures in all the potential suspended sediment sources, and delta15N was a unique tracer to differentiate stream bank soil from upland subsurface soils, such as soil from construction sites, unpaved roads, ditches, and field gullies. Mixing models showed that about 60% of the stream suspended sediment originated from eroding stream banks, 23 to 30% from upland subsoil sources (e.g., construction sites and unpaved roads), and about 10 to 15% from pastures. The results may be applicable to other watersheds in the Piedmont depending on the extent of urbanization occurring in these watersheds. Better understanding of the sources of fine sediment has practical implications on the type of sediment control measures to be adopted. Investment of resources in improving water quality should consider the factors causing stream bank erosion and erosion from unpaved roads and construction sites to water quality impairment.

Water quality modeling efforts for developing total maximum daily loads often use geographical information system data of varying quality in watershed-scale models and have shown varying impacts on model results. Several streams in the southern Piedmont are listed for sediment total maximum daily loads. The objective of this study was to test the effect of spatial resolution of soil data on the SWAT (Soil and Water Assessment Tool) model predic- tions of flow and sediment and to calibrate the SWAT model for a watershed dominated by channel erosion. The state soil geographic (STATSGO) database mapped at 1:250,000 scale was compared with the soil survey geographic (SSURGO) database mapped at 1:12,000 scale in an ArcSWAT model of the North Fork Broad River in Georgia. Model outputs were compared for the effect of soil data before calibration using default model parameters as calibration can mask the effect of soil data. The model predictions of flow and sediment by the two models were similar, and the differences were statistically insignificant (α = 0.05). These results were attributed to the similarity in key soil property values in the two databases that govern stream flow and sediment transport. The two models after calibration had com- parable model efficiency in simulating stream flow and sediment loads. The calibrated models indicated that channel erosion contributed most of the suspended sediment in this watershed. These findings indicate that less detailed soil data can be used because more time, effort, and computational resources are required to set up and calibrate a model with more detailed soil data, especially in a larger watershed.

The use of watershed models as cost-effective tools to quantify the impact of conservation practices on water quality is often constrained by lack of data for model parameterization. This study uses short-term (3 years) detailed monitoring data to guide spatially distributed model parameterization and modeling analysis for suspended sediment in the Upper Esopus Creek Watershed (UECW) that is part of the New York City water supply. The calibrated Soil and Water Assessment Tool (SWAT) model simulated suspended sediment loading from tributary sub-basins and at the watershed outlet that were comparable to field measurements. Model simulations estimated that stream channels contributed the majority (85%) of stream sediment in the study watershed followed by upland erosion (11%) and point sources (4%), consistent with previous estimates and field observations. Long-term (12 years) simulation of the calibrated model was used to apportion the average annual sediment yields from tributary sub-basins which ranged between 12 and 161 t km(-2) year(-1). Model simulations were also used to understand the inter-annual variability and seasonality in suspended sediment loading in the study watershed. We demonstrate the wider applicability of short-term detailed monitoring for model parameterization and calibration, and long-term simulation of water quality using the SWAT model.

Many of us who have watched a stream turn turbid during a rainstorm have often wondered—where is all that sediment coming from? Geomorphologists, engineers, and environmental scientists have thought about this question for decades and have placed understanding sediment sources under the general framework of sediment budgets. This framework which examines not only sediment sources but the storage, transport, and delivery of sediment has been used to understand surficial processes, as well as address management concerns as they relate to stream turbidity and surface-water quality. Traditional tools used in sediment budgets have included field measurements of erosion, storage and transport rates, photogrammetric analysis, and modeling. The relationship between soil erosion and sediment yield at the watershed outlet has been long identified as a major research need (Walling 1983) and is still a poorly understood or misunderstood component of fluvial sediment transport (Kinnell 2004). In recent decades, sediment fingerprinting approaches using the geochemical and physical properties of sediment to determine sediment sources have become increasingly popular. This special issue on Watershed Sediment Source Identification, for the Journal of Soils and Sediments was produced as a result of a special session convened at the Association of American Geographers (AAG) 2012 annual meeting in New York City, NY, USA. The session was entitled “Watershed Sediment Source Identification” and brought together leading experts in the field of sediment sourcing. As a result of the AAG session, the Journal of Soils and Sediments offered to have the session and a few other papers on the same topic published as a special issue in the journal. This special issue reflects contributions from 29 researchers discussing sediment sourcing results from studies in Africa, Asia, Europe, and North America where settings range from urban to agricultural to forested. Sediment source assessment is not only important to our understanding of sediment dynamics in fluvial systems but is increasingly becoming an important management tool. The papers in this special issue reflect a range of studies that address both topics. Studies emphasize traditional field approaches to determine sediment yields and sources (Davis and Sims 2013; Zhu 2013) as well as the use of the sediment fingerprinting approach to determine significant sediment sources (Dutton et al. 2013; Gellis and Noe 2013; Huisman et al. 2013; Koiter et al. 2013; Mckinley et al. 2013; Voli et al. 2013). Several papers highlight site-specific methods and approaches to streamline the sediment fingerprinting approach into a practical management tool. Other papers relate findings from sediment fingerprinting to hydrologic conditions, and caution users on over-interpreting the sediment fingerprinting results. Walling (2013), in the opening paper, provides a history on the sediment fingerprinting approach and illustrates how publications on the subject have increased exponentially since the 1970s. A review of these publications indicate that many studies have used sediment fingerprinting to determine sediment sources as a means to improve our understanding of erosion and sediment delivery processes. Other papers use sediment fingerprinting as a management tool. Walling (2013) lists key advances in the sediment fingerprinting approach over the last 30 years that include the use of multiple or composite fingerprints, statistical tests and models, size and organic correction factors, increased source and target assessments, and improved estimates of uncertainty. Given these advances in the sediment A. C. Gellis (*) U.S. Geological Survey, 5522 Research Park Drive, Baltimore, MD 21228, USA e-mail: agellis@usgs.gov

Chlorine, a commonly used disinfectant in most water supply systems, can combine with organic carbon to form disinfectant byproducts, including carcinogenic trihalomethanes. We used water quality data from 24 monitoring sites within the New York City water supply distribution system, measured between January 2009 and April 2012, to develop an empirical model for predicting total trihalomethane (TTHM) levels. Terms in the model included the following water quality parameters: total organic carbon, pH, water age (reaction time), and water temperature. Reasonable estimates of TTHM levels were achieved with overall R² of about 0.75, and predicted values on average were within 6 μg L⁻¹ of measured values. A sensitivity analysis indicated that total organic carbon and water age are the most important factors for TTHM formation, followed by water temperature; pH was the least important factor within the boundary conditions of observed water quality. Although never out of compliance in 2011, the TTHM levels in the water supply increased after tropical storms Irene and Lee, with 45% of the samples exceeding the 80 μg L⁻¹ maximum contaminant level in October and November. This increase was explained by changes in water quality parameters, particularly by the increase in total organic carbon concentration during this period. This study demonstrates the use of an empirical model to understand TTHM formative factors and their relative importance in a drinking water supply. This has implications for simulating management scenarios and real-time estimation of TTHMs in water supply systems under changing environmental conditions.

Scaling-up sediment transport has been problematic because most sediment loss models (e.g., the Universal Soil Loss Equation) are developed using data from small plots where runoff is generated by infiltration excess. However, in most watersheds, runoff is produced by saturation excess processes. In this article, we improve an earlier saturation excess erosion model that was only tested on a limited basis, in which runoff and erosion originated from periodically saturated and severely degraded areas, and apply it to five watersheds over a wider geographical area. The erosion model is based on a semi-distributed hydrology model that calculates saturation excess runoff, interflow, and baseflow. In the development of the erosion model, a linear relationship between sediment concentration and velocity in surface runoff is assumed. Baseflow and interflow are sediment free. Initially during the rainy season in Ethiopia, when the fields are being plowed, the sediment concentration in the river is limited by the ability of the surface runoff to move sediment. Later in the season, the sediment concentration becomes limited by the availability of sediment. To show the general applicability of the Saturation Excess Erosion Model (SEEModel), the model was tested for watersheds located 10,000 km apart, in the U. S. and in Ethiopia. In the Ethiopian highlands, we simulated the 1.1 km(2) Anjeni watershed, the 4.8 km(2) Andit Tid watershed, the 4.0 km(2) Enkulal watershed, and the 174,000 km(2) Blue Nile basin. In the Catskill Mountains in New York State, the sediment concentrations were simulated in the 493 km(2) upper Esopus Creek watershed. Discharge and sediment concentration averaged over 1 to 10 days were well simulated over the range of scales with comparable parameter sets. The Nash-Sutcliffe efficiency (NSE) values for the validation runs for the stream discharge were between 0.77 and 0.92. Sediment concentrations had NSE values ranging from 0.56 to 0.86 using only four calibrated sediment parameters together with the subsurface and surface runoff discharges calculated by the hydrology model. The model results suggest that correctly predicting both surface runoff and subsurface flow is an important step in simulating sediment concentrations. International Foundation for Science (IFS) Higher Education for Development (HED) USDA International Science and Education (ISE) program