Distributed watershed hydrologic/water quality (H/WQ) models are ubiquitous tools for watershed management. Despite advancements, there remain impediments for end-users. This study presents a practical framework for use of the soil and water assessment tool (SWAT). Results show variable accuracy across scales and evaluation methods using 20 model configurations based on two watershed subdivisions, two soil datasets, and five climate datasets. Nine goodness-of-fit indicators were tested, including four new indices (R-RMSE, R-MAE, R-NSE, and R-NSE1) designed to quantify model fit with flow distribution. Sixteen of 20 configurations achieved satisfactory monthly streamflow fit (NSE [greater than] 0.5, PBIAS [lesser than] 25%) without calibration. Watershed and soil resolution had negligible impact; climate input had considerable impact. Single climate station input is best used for applications requiring monthly predictions; distributed climate station input is needed for daily predictions. SWAT multi-objective auto-calibration better predicted monthly flow (PBIAS=1%, NSE=0.8) than single-objective calibration (PBIAS=16%, NSE=0.5). SWAT performs well in Central U.S. urbanizing watersheds. Accuracy can improve with auto-calibration as presented and continued model development.
Anthropogenic and natural disturbances to freshwater quantity and quality are a greater issue for society than ever before. To successfully restore water resources requires understanding the interactions between hydrology, climate, land use, water quality, ecology, and social and economic pressures. This Special Issue of Water includes cutting edge research broadly addressing investigative areas related to experimental study designs and modeling, freshwater pollutants of concern, and human dimensions of water use and management. Results demonstrate the immense, globally transferable value of the experimental watershed approach, the relevance and critical importance of current integrated studies of pollutants of concern, and the imperative to include human sociological and economic processes in water resources investigations. In spite of the latest progress, as demonstrated in this Special Issue, managers remain insufficiently informed to make the best water resource decisions amidst combined influences of land use change, rapid ongoing human population growth, and changing environmental conditions. There is, thus, a persistent need for further advancements in integrated and interdisciplinary research to improve the scientific understanding, management, and future sustainability of water resources.
Hydrological models have been increasingly used for the effect of land cover change and forest management operations on hydrological processes. In the Sierra Nevada, where timber harvest and prescribed fire are commonly employed for forest management, hydrological models have rarely been used, especially in small watersheds. In this research, the SWAT model (Soil Water and Assessment Tool) was used to simulate streamflow on a daily time-step in P301, a small headwater mountain watershed located in the southern Sierra Nevada. The watershed is 1 km2, where about 72% of the land is covered by a dense mixed-conifer forest. SWAT performs satisfactorily with a coefficient of determination (R2) of 0.59 and a Nash-Sutcliffe efficiency value (NSE) of 0.59. This is important to know given the complexity arising from model uncertainty and the intricacies of Sierra Nevada hydrology. Although SWAT performed "satisfactory", the model still missed two key hydrological processes: the timing of snowmelt and isolated peak flow events. In addition, simulating streamflow on the daily time-step is good for understanding watershed processing and functioning but is not as useful for forest and land management. SWAT will need further model adjustments as well as monthly and yearly water yield estimates in order to be considered for the evaluation of forest management operations in P301.
New York City's municipal water supply system provides about 1 billion gallons of drinking water a day to over 8.5 million people in New York City and about 1 million people living in nearby Westchester, Putnam, Ulster, and Orange counties. The combined water supply system includes 19 reservoirs and three controlled lakes with a total storage capacity of approximately 580 billion gallons. The city's Watershed Protection Program is intended to maintain and enhance the high quality of these surface water sources. Review of the New York City Watershed Protection Program assesses the efficacy and future of New York City's watershed management activities. The report identifies program areas that may require future change or action, including continued efforts to address turbidity and responding to changes in reservoir water quality as a result of climate change.
Applications of the SWAT model typically involve delineation of a watershed into subwatersheds/subbasins that are then further subdivided into hydrologic response units (HRUs) which are homogeneous areas of aggregated soil, landuse, and slope and are the smallest modeling units used within the tool. In a standard SWAT application, multiple potential HRUs (farm fields) in a subbasin are usually aggregated into a single HRU feature. In other words, the standard version of the model combines multiple potential HRUs (farm fields) with the same landuse/landcover (LULC), soil, and slope, but located in different places within a subbasin (spatially non-unique), and considers them as one HRU. In this study, ArcGIS pre-processing procedures were developed to spatially define a one-to-one match between farm fields and HRUs (spatially unique HRUs) within a subbasin prior to SWAT simulations to facilitate input processing, input/output mapping, and further analysis at the individual farm field level. Model input data such as LULC, soil, crop rotation and other management data were processed through these HRUs. The SWAT model was then calibrated/validated for the Raccoon River watershed in Iowa for 2002 to 2010 and the Big Creek River watershed in Illinois for 2000 to 2003. SWAT was able to replicate annual, monthly and daily streamflow, as well as sediment, nitrate and mineral phosphorous within recommended accuracy in most cases. The one-to-one match between farm fields and HRUs created and used in this study is a first step in performing LULC change, climate change impact, and other analyses in a more spatially explicit manner. The calibrated and validated SWAT model was then used to assess agricultural scenario and climate change impacts on watershed water quantity, quality, and crop yields. Modeling impacts of agricultural scenarios and climate change on surface water quantity and quality provides useful information for planning effective water, environmental, and land use policies. Despite the significant impacts of agriculture on water quantity and quality, limited literature exists modeling the combined impacts of agricultural scenarios and climate change on crop yields and watershed hydrology. Here, SWAT, was used to model the combined impacts of five agricultural scenarios and three climate scenarios downscaled using eight climate models. These scenarios were implemented in a well calibrated SWAT model for the Raccoon River watershed (RRW), IA. We run the scenarios for the historical baseline, early-century, mid-century, and late-century periods. Results indicate that historical and more corn intensive agricultural scenarios with higher CO2 emissions consistently result in more water in the streams and greater water quality problems, especially late in the 21st century. Planting more switchgrass, on the other hand, results in less water in the streams and water quality improvements relative to the baseline. For all given agricultural landscapes simulated, all flow, sediment and nutrient outputs increase from early-to-late century periods for the RCP4.5 and RCP8.5 climate scenarios. We also find that corn and switchgrass yields are negatively impacted under RCP4.5 and RCP8.5 scenarios in the mid and late 21st century. Finally, various agricultural best management practice (BMP) scenarios were evaluated for their efficiency in alleviating watershed water quality problems. The vast majority of the literature on efficiency assessment of BMPs in alleviating water quality problems base their scenarios analysis on identifying subbasin level simulation results. In the this study, we used spatially explicit HRUs, defined using ArcGIS-based pre-processing methodology, to identify Nitrate (NO3) and Total Suspended Solids (TSS) hotspots at the HRU/field level, and evaluate the efficiency of selected BMPs in a large watershed, RRW, using the SWAT model. Accordingly, analysis of fourteen management scenarios were performed based on systematic combinations of five agricultural BMPs (fertilizer/manure management, changing cropland to perennial grass, vegetative filter strips, cover crops and shallower tile drainage systems) aimed to reduce NO3 and TSS yields from targeted hotspot areas in the watershed at field level. Moreover, implications of climate change on management practices, and impacts of management practices on water availability and crop yield and total production were assessed. Results indicated that either implementation of multiple BMPs or conversion of an extensive area into perennial grass may be required to sufficiently reduce nitrate loads to meet the drinking water standard. Moreover, climate change may undermine the effectiveness of management practices, especially late in the 21 st century. The targeted approach used in this study resulted in slight decreases in watershed average crop yields, hence the reduction in total crop production is mainly due to conversion of croplands to perennial grass.
This dissertation describes the modeling efforts on the Upper Mississippi River Basin (UMRB) using the Soil and Water Assessment Tool (SWAT) model. The main goal of this study is to apply the SWAT model to the UMRB to evaluate the model as a tool for agricultural policy analysis and climate change impact analysis. A sensitivity analysis using influence coefficient method was conducted for eight selected hydrologic input parameters to identify the most to the least sensitive parameters. Calibration and validation of SWAT were performed for the Maquoketa River Watershed for streamflow on annual and monthly basis. The model was then validated for the entire UMRB streamflow and evaluated for a climate change impact analysis. The results indicate that the UMRB hydrology is very sensitve to potential future climate changes. The impact of future climate change was then explored for the streamflow by using two 10-year scenario periods (1990 and 2040s) generated by introducing a regional climate model (RegCM2) to dynamically downscale global model (HadCM2) results. The combined GCM-RCM-SWAT model system produced an increase in future scenario climate precipitation of 21% with a resulting 50% increase in total water yield in the UMRB. Furthermore, evaluation of model-introduced uncertainties due to use of SWAT, GCM, and RCM models yielded the highest percentage bias (18%) for the GCM downscaling error. Building upon the above SWAT validation, a SWAT modeling framework was constructed for the entire UMRB, which incorporates more detailed input data and is designed to assess the effects of land use, climate, and soil conditions on streamflow and water quality. An application of SWAT is presented for the Iowa and Des Moines River watersheds within the modeling framework constructed for the UMRB. A scenario run where conservation tillage adoption increased to 100% found a small sediment reduction of 5.8% for Iowa River Watershed and 5.7% for Des Moines River Watershed. On per-acre basis, sediment reduction for Iowa and Des Moines River Watersheds was found to be 1.86 and 1.18 metric tons respectively. Furthermore an attempt to validate the model for the entire UMRB yielded strong annual results.
Global climate change is a natural process that currently appears to be strongly influenced by human activities, which increase atmospheric concentrations of greenhouse gases (GHG). Agriculture contributes about 20% of the world's global radiation forcing from carbon dioxide, methane and nitrous oxide, and produces 50% of the methane and 70% of the nitrous oxide of the human-induced emission. Managing Agricultural Greenhouse Gases synthesizes the wealth of information generated from the GRACEnet (Greenhouse gas Reduction through Agricultural Carbon Enhancement network) effort with contributors from a variety of backgrounds, and reports findings with important international applications. - Frames responses to challenges associated with climate change within the geographical domain of the U.S., while providing a useful model for researchers in the many parts of the world that possess similar ecoregions - Covers not only soil C dynamics but also nitrous oxide and methane flux, filling a void in the existing literature - Educates scientists and technical service providers conducting greenhouse gas research, industry, and regulators in their agricultural research by addressing the issues of GHG emissions and ways to reduce these emissions - Synthesizes the data from top experts in the world into clear recommendations and expectations for improvements in the agricultural management of global warming potential as an aggregate of GHG emissions
Excess nitrate and phosphorus export to surface water bodies may yield negative environmental impacts, including the eutrophication of downstream areas. To address this issue, a modeling technique was deployed to quantify and assess these processes under various seasonal scenarios at the Strawberry Creek Watershed. The Soil and Water Assessment Tool (SWAT) model was modified (SWAT tile) and parameters defined to simulate the effects of tile drainage on flow and nitrate (NO3−) export from small watersheds during the four seasons characteristic of southern Ontario. This study compares differences and similarities between observed watershed processes against model output by: (1) utilizing the SWATtile model for comparison of simulated to measured discharge from a watershed, and a tiled field, from several years of data, (2) utilizing the SWATtile model for comparison of simulated to measured NO3− from a watershed and tiled field, and (3) several scenarios are presented on how modifications to tile spacing (density) can be manipulated to achieve a balance between improving soil drainage while minimizing NO3− export. The effects of tile density changes were evaluated to determine the impact of moisture availability (for tile flow) as precipitation cycled from rain to snow and back to rain. In the first part of this study, comparison of detailed simulations of seasonal flow patterns from both the gauged watershed and a gauged tiled field for winter 2007 to winter 2008, reveals similarities and contrasts in flow patterns for daily time scales. Due to its distributed nature, the SWAT model is subdivided into fundamental units of analysis designated as Hydrologic Response Units (HRUs). Each HRU consists of a unique soil and landuse type and is capable of autonomous analysis and result generation. The gauged subwatershed area drained by the Below Middle Road (BMR) tile has been continuously monitored for more than six years. This subwatershed was defined in the SWAT model setup as an independent HRU so that results generated from simulations can be directly compared to measured values from the same area. In terms of landuse, soil type and tile spacing, the BMR - HRU is representative of other tiled fields in the watershed. Simulated stream and tile flow for each season were comparable to that of the observed. Linear trends between measured main channel flow and that of measured tile flow was statistically significant. However, trend agreement between simulated main channel discharge and BMR tile was not statistically significant, although it demonstrated a general linear pattern. For the second part of the study, comparison of observed/measured watershed NO3− concentration against results generated by SWATtile model were quantified across all seasons with the contrast being greatest for the spring season. The general trend in modeled NO3− is for more of it to be exported during low flows. NO3− then increases with volume of flow. The tile outlet yields a higher NO3− load per unit area, as this contribution originates from a much smaller area (0.43 km2) compared to main stream outlet with contribution from the entire watershed which is a much large area (2.86 km2). For both the first and second scenarios, the tile drainage component was also disabled to enable observation of dominance of overland flow as a result of an elevated water table. Consequently, there was an observable reduction in crop NO3- uptake which is largely due to an increased rate of denitrification under anaerobic conditions. The third part of the study introduces variability in density between feeder tiles and thus altered the drainage intensity. The drainage intensity is the rate at which water is removed from a field and is thus proportional to tile density. As the intensity is increased, drainage and NO3− export also increases proportionally. On the other hand, as the lateral distance is increased above 50 ft. (15.24 m), tile drainage and NO3− export from the field are reduced. Crop NO3− uptake was also reduced (decreased productivity) with an increase in tile density (from 50ft. to 35ft.). This was also characterised by increased NO3− export. The anoxic conditions might also favour denitrification which may lead to further NO3− loss. For the watershed simulation, although decreasing tile density helped reduce NO3− mass export (density reduced from 50ft [90 kg/ha] to 65ft [82 kg/ha]), it was still not enough to attain the required drinking water standard of 10 mg/L (and the limit of 12.8 mg/L for aquatic species). However, when the tile density was reduced to 85ft. (30m), the concentration of NO3− decreased to 25 kg/ha.
The effectiveness of conservation practices depends on their placement on the fields within the watershed. Cost-effective placement of these practices for maximum water quality benefits on each field requires comparing a very large number of possible land-use scenarios. To address this problem, we combine the tools of evolutionary algorithm with the Soil and Water Assessment Tool (SWAT) model and cost data to develop a trade-off frontier of least cost of achieving nutrient reductions and the corresponding locations of conservation practices. This approach was applied to the Raccoon River Watershed, which drains about 9,400 km2 of an intensive agriculture region in west-central Iowa. Applying genetic algorithm to the calibrated SWAT modeling setup produced multitudes of optimal solutions of achieving nutrient reductions in relation to the total cost of placing these practices. For example, a 30% reduction in nitrate (and a corresponding 53% reduction in phosphorus) at the watershed outlet can be achieved with a cost of $80 million per year. This solution frontier allows policymakers and stakeholders to explicitly see the trade-offs between cost and nutrient reductions.
