Next Article in Journal
Numerical Analysis of the Effects of Crack Characteristics on the Stress and Deformation of Unsaturated Soil Slopes
Next Article in Special Issue
Factors Controlling of Thermal Water Hydrogeochemical Characteristics in Tatun Volcano Group, Taiwan
Previous Article in Journal
Evidence of Natural and Anthropogenic Impacts on Rainwater Trace Metal Geochemistry in Central Mexico: A Statistical Approach
Previous Article in Special Issue
Impact of Climate Change and Human Activities on Streamflow Variations Based on the Budyko Framework
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Comparison of Streamflow and Baseflow Responses to Land-Use Change and the Variation in Climate Parameters Using SWAT

1
Department of Agricultural & Biological Engineering, Purdue University, West Lafayette, IN 47906, USA
2
Geology Department, Faculty of Science, Suez Canal University, Ismailia 41511, Egypt
*
Author to whom correspondence should be addressed.
Water 2020, 12(1), 191; https://doi.org/10.3390/w12010191
Submission received: 22 November 2019 / Revised: 5 January 2020 / Accepted: 7 January 2020 / Published: 10 January 2020

Abstract

:
Alteration of land use and climate change are among the main variables affecting watershed hydrology. Characterizing the impacts of climate variation and land use alteration on water resources is essential in managing watersheds. Thus, in this research, streamflow and baseflow responses to climate and land use variation were modeled in two watersheds, the Upper West Branch DuPage River (UWBDR) watershed in Illinois and Walzem Creek watershed in Texas. The variations in streamflow and baseflow were evaluated using the Soil and Water Assessment Tool (SWAT) hydrological model. The alteration in land use between 1992 and 2011 was evaluated using transition matrix analysis. The non-parametric Mann–Kendall test was adopted to investigate changes in meteorological data for 1980–2017. Our results indicate that the baseflow accounted for almost 55.3% and 33.3% of the annual streamflow in the UWBDR and Walzem Creek watersheds, respectively. The contribution of both land use alteration and climate variability on the flow variation is higher in the UWBDR watershed. In Walzem Creek, the alteration in streamflow and baseflow appears to be driven by the effect of urbanization more than that of climate variability. The results reported herein are compared with results reported in recent work by the authors in order to provide necessary information for water resources management planning, as well as soil and water conservation, and to broaden the current understanding of hydrological components variation in different climate regions.

1. Introduction

Ecosystems and humans are fundamentally dependent on different water resources. Thus, for the general development of any country, the quality and the quantity of these water resources flowing through rivers are of vital importance to socio-economic development [1]. Issues related to changes in water resources are commonly evaluated around the globe [2,3,4]. In the United States, evaluation of streamflow and baseflow has been documented [5,6,7]. However, the quantitative change in streamflow and baseflow has yet to be evaluated across different climatic conditions.
Climate alterations and human actions both act as stressors to place severe pressure on water resources [8,9]. The variations in climate and land use directly impact total streamflow, interflow, surface runoff, and baseflow, causing events of droughts and floods that impact the sustainability of these resources and the social ecosystem [10]. Several studies have examined alterations in streamflow due to changes in temperature and precipitation [11,12,13], urbanization [14], and land use change [2,15]. Baseflow is the portion of streamflow sustained in a river by delayed pathways. Baseflow is often assumed to be equal to groundwater recharge [16]. It provides a relatively high water quality with a high clarity and stable temperature, and is considered indicative of sustained streamflow during dry periods of the season, which is important to stream biota and helps recreation-based industries [17]. These low-flow data are essential in understanding the current and future changes to watershed hydrology. Several reports have indicated that the change in baseflow over time is due to variations in agricultural management [18], climate change [8], urbanization [19], and land use alteration [20]. Therefore, to develop scenarios for water resources evaluation, land use change and climate variation are usually chosen as the main influencing factors. The impacts of climate variation and urbanization on streamflow and baseflow were reviewed by Aboelnour et al. [8] and Price [21].
Different methods have been used to evaluate the response of watershed streamflow and baseflow to human activities and climate change. These techniques include hydrologic similarities within the watersheds, paired catchments, statistical methods, and hydrological modeling [22]. Since climate and land use change need to be investigated on a local scale and can vary from place to place [23], there is a need to use comprehensive and physical tools to evaluate as much information as possible from the limited existing data [24]. Hence, hydrological models are considered the most appealing approach to carry out impact assessment studies. They provide a conceptualized framework and are suitable for use as part of scenario studies on the relationship among hydrological components, climate variability, and land use change [25,26]. Among these models is the Soil and Water Assessment Tool (SWAT) model.
The SWAT model, developed by the United States Department of Agricultural (USDA) Agriculture Research Service, is designed to model hydrology at the scale of a watershed [27]. SWAT is widely used around the world to evaluate the influences of ecological and environmental alterations and for hydrological processes at different catchment scales, even with limited data [10,28]. In addition, it offers several software tools, and was therefore selected for this research. Each watershed was divided into smaller sub-basins in the SWAT model. These sub-basins were then divided into smaller Hydrological Response Units (HRUs), which were fundamentally based on land use, soil type, and slope [29]. Within each HRU, the Soil Conservation Service (SCS) curve number and Green–Ampt infiltration were adapted to compute surface runoff using daily precipitation. In addition, SWAT subdivides the groundwater system into deep confined aquifers, which contribute to flow outside of the catchment, and shallow unconfined ones, in which the groundwater and baseflow return to the stream [30]. The SWAT model has proven to perform well in streamflow and baseflow simulations around the world and in complex catchments with extreme events [31], since it allows the interconnections of different physical processes [32,33,34]. Therefore, in this research, the SWAT model was adopted to assess the impacts of land use and climate change.
Streamflow and baseflow responses to human activities, urbanization, and climate variation are different in various basins with respect to climate regions, geographical variances, scale, and urbanization levels [21,35]. However, the need to fully understand the streamflow and baseflow responses to external stimuli is of vital importance. Many studies in the last few years have been carried out to investigate the hydrological response to urbanization and climate change [13,36]. Outputs of these studies can help in understanding the cause of shifts in water resources. However, these studies mainly focus on the single impact of either land use change or climate variation, but neglect the combined effects of climate alteration and human activities and their contributions to the change. Thus, the combined effects are still not fully understood over different climatic conditions and geographical regions. For this reason, the responses of streamflow and baseflow to urbanization and climate variation will be evaluated for varying climate conditions with different urbanization levels. Two watersheds, the Upper West Branch DuPage River (UWBDR) watershed, Illinois, and Walzem Creek watershed, Taxes, were used as examples to quantify the changes in streamflow and baseflow as a response to climate and land use change.
As evidenced by the USA Census population data, the Upper West Branch DuPage River (UWBDR) watershed, Illinois, has undergone intense urbanization in the last four decades. In addition to this dramatic urbanization, the watershed has experienced major flood events, such as the floods of 1996 and 2008 [37]. Other incidences in the watershed have been identified as impactful on the development of the UWBDR. One of the main contributors is floodplain management that addressed overbank flooding of the main stream and its tributaries [37]. Hejazi and Markus [38] investigated the impacts of urbanization and climate variability on annual flooding in 12 urban watersheds in Cook County, northeastern Illinois. They found that urbanization had a greater impact than climate on the increase in flood discharge, and, due to increasing urbanization, discharge volume may become even higher in the future. In addition to floodplain management, wetland protection, bank stabilization, stream restoration, water quality, and groundwater recharge are also concerns within the catchments. Some sections of the stream are supplied with a substantial amount of their baseflow from local groundwater discharge, while other sections release baseflow to groundwater due to the presence of a large outwash plain at the base of West Chicago Moraine that creates conditions that promote rapid flooding and groundwater movement from the border of the moraine through the outwash [37].
The second watershed is the Walzem Creek, San Antonio, Texas. The city of San Antonio, Bexar County and other partners initiated a watershed protection plan in 2006 for the Upper San Antonio basin, including the Walzem Creek watershed, to track efforts that enhance urban outreach, and to bring the basin back into compliance with water resource and water quality recreation standards. In 2015, the Environmental Protection Agency (EPA) approved this protection plan, making the state eligible for project funding within the watershed to address nonpoint source runoff. The report can be viewed at https://www.brwm-tx.org/. A combination of rocky and clay soils contributes to larger runoff than groundwater flow in this watershed. Rock, clay, and slopes create nearly impervious conditions in the northern portion of the watershed and thus reduce the effect of development and its associated impervious cover on storm water flow [39].
The main target of this study was to evaluate the impact of separate and combined impacts of land use changes and climate alteration on streamflow and baseflow in two watersheds under different land use and climatic conditions. The specific goals of this research were: (1) identify the long term trend and the abrupt changes in hydrological and meteorological data; (2) determine the change in land use maps from 1992 to 2011; (3) use the new calibrated and validated SWAT model to assess the individual and combined impacts of land use change and climate variation on streamflow and baseflow; and (4) compare the outputs of this study with the findings of Aboelnour et al. [8]. Information gleaned from this study can be used to understand the variations in hydrological flow components, and are necessary for water resources management and planning, as well as water and soil conservation in geographically different watersheds.

2. Study Areas

2.1. Upper West Branch DuPage River Watershed

The Upper West Branch DuPage River watershed (UWBDR) is located in northeast Illinois, within the six-county Chicago metropolitan region. The watershed is located approximately in the western one third of DuPage County (Figure 1a). The headwaters originate in the northwestern part of Cook County where the water flows generally to the south into and through DuPage County. The UWBDR is part of the West Branch DuPage River (WBDR) watershed that divides into upper, middle, and lower branches within the DuPage catchment and belongs to the Des Plaines River basin. The UWBDR covers approximately 91.7 km2 (USGS Gauge 05539900) with mean annual precipitation ranging from 612 to 1293 mm from 1980 to 2017, and average annual temperatures ranging from 8.4 to 12.5 °C. The minimum, maximum, and mean elevations in the area are 217, 298 and 240 m above sea level, respectively. Developed and residential areas were the dominant land use type in the UWBDR at the end of the last century (44.1%), followed by cultivated and forest land cover at 39.0% and 8.8%, respectively. Current land use varies from residential (84.2%) to forest (4.4%), vacant (4.5%), and cultivated (2.7%). The river network in the watershed receives treated effluent and wastewater from the cities of West Chicago, Illinois [40].

2.2. Walzem Creek

Walzem Creek is located in Bexar County in the state of Texas and in the San Antonio East USGS quad (Figure 1b). Currently, except for the lower most portion of the watershed, the majority of Walzem Creek is characterized by dense, urban development. The lower portion of the watershed is characterized by a mix of vegetation and forests and normally dry except during rain events. The Walzem Creek is a part of Upper San Antonio Watershed and covers approximately 109 km2 (USGS Gauge 08178800), with a mean annual precipitation ranging from 320 to 1200 mm and average annual temperatures ranging from 19.3 to 22.3 °C. Mean elevation in the area is 204 m, with a minimum and maximum of 152 and 266 m above sea level, respectively. Similar to the UWBDR watershed, recently, most of Walzem Creek is covered with developed areas (84.5%); however, other land uses include wetlands (7.2%), shrublands (4.2%), and forests (2.5%). However, it was characterized by only 64.4% of residential areas, 17.5% of planted cover, and 8.7% and 7.0% of forest and shrublands covers, respectively, at the beginning of the 1990s. This area is a large portion of the Upper San Antonio Watershed; hence, it contributes a large amount of total streamflow. According to the main Koppen–Geiger climate classes for US counties, the San Antonio, Bexar County area lies at the border between warm, humid, equatorial zone and fully hot arid and steppe zone [41]. Therefore, this watershed is representative of semi-arid regions.

3. Materials and Methods

The data described herein include spatial topography, Digital Elevation Model (DEM), land use and soil data, and hydro-meteorological data. Data analysis procedures and methods used are detailed extensively in the work by Aboelnour et al. [8]. A flow chart depicting procedures used in this study is summarized in Figure 2.

3.1. Data Development

3.1.1. Spatial Data

Two raster land use maps for the years 1992 and 2011 were obtained from the National Map Viewer (NMV). Digital Elevation Model (DEM) data were acquired from the Geospatial Data Gateway (GDG) with a resolution of 10 m. Soil Survey Geographic Data (SSURGO) data were used in this research with a resolution ranging from 1:12,000 to 1:63,630. Land use, soil type, and slope were then used to divide the delineated sub-basins into a small series of uniform HRUs that represent the smallest representative units within the watershed [42].

3.1.2. Hydro-Meteorological Data

The required datasets used include daily observed streamflow data at gauged USGS stations for the period 1980 to 2017. The streamflow data were used to separate baseflow from surface runoff, and for the SWAT model calibration and validation. In addition, long-term daily meteorological datasets for the same period (1980–2017) were collected from the National Climate Data Center (NCDC). The meteorological weather stations were 12 km and 0.8 km away from the borders of the UWBDR and Walzem Creek watersheds, respectively.

3.2. Methodology

3.2.1. Baseflow Separation

Baseflow measurements were separated from daily streamflow data acquired from USGS gauged stations using the automatic baseflow digital filter method (BFlow). The BFlow filter separates streamflow data into baseflow and surface runoff by passing the observed streamflow through the filtering equation three times [43,44]:
B F t = α × B F t 1 + 1 α 2 × ( Q t + Q t 1 )
where B F is the baseflow, α is the filter parameter (0.925), Q is the total streamflow, and t is the time step. Equation (1) is applied only when BFQt [45]. BFlow is a conservative filter that enables the user to filter streamflow data to calculate the baseflow, and also to generate a tabular dataset or graphical hydrograph interface from USGS gaging stations. Herein, BFlow filtered streamflow data three times (Equation (1)), and it is commonly observed that the 1-pass baseflow is consistent with manually estimated baseflow and thus was subsequently used in this study [45].

3.2.2. Soil and Water Assessment Tool (SWAT) Model Calibration and Validation

The monotonic trends in the historical meteorological data were evaluated using the modified Mann–Kendall (MK) test developed by Hamed and Rao [46]. Based on the abrupt change in the trends in precipitation and temperature using the MK test, the study period from 1980 to 2017 was split into two time spans, 1980–1998 and 1999–2017, with a breakpoint in 1998. The period 1980–1998 was assigned as a baseline for model calibration and validation. The model simulation time was segmented into a warm up period (1980–1983), calibration period (1984–1993), and validation period (1994–1998). The SWAT model calibration and validation were performed using the land use map of 1992 and streamflow data from 1980 to 1998 for each of the selected watersheds. Model optimization, sensitivity analysis, calibration, validation, and uncertainty analysis of parameters were carried out using the Sequential Uncertainty Fitting program algorithm (SUFI-2) approach within the SWAT-CUP interface developed by Abbaspour et al. [47]. Based on Aboelnour et al. [8], the twenty hydrologic parameters listed in Table 1 were used in this study for the UWBDR and Walzem Creek watersheds calibration of streamflow and baseflow. However, sensitivity analysis using the SUFI-2 global sensitivity analysis was carried out in the first stage due to the presence of many parameters within the SWAT model [44]. Only parameters sensitive for the watersheds were then used in the calibration process and optimized based on monthly values [48]. Both automatic and manual calibration were carried out to allow qualitative and quantitative comparisons of the values, to fine tune the values of the auto-calibrated parameters, and to decrease the differences between the observed and simulated outputs [49].

3.2.3. Model Sensitivity Analysis

The global sensitivity analysis procedures showed that the sensitive parameters obtained from the LEC by Aboelnour et al. [8] were critical in the case of the UWBDR watershed, but with a different rank order. It was also found that these rankings were impacted by the selected objective function in the model. For example, curve number (CN2), soil evaporation compensation factor (ESCO), snowfall temperature (SFTMP), melt factor for snow (SMFMN), baseflow recession constant (ALPHA_BF), and deep aquifer percolation fraction (RCHRG_DP) were the most critical parameters in UWBDR when the Kling–Gupta Efficiency (KGE) was selected to be the objective function incorporated into the model (Table 2). These parameters characterize surface runoff, soil properties, and groundwater.
In the case of Walzem Creek, the parameters in Table 2 were consistent with other SWAT parameter sensitivity analyses completed for semi-arid regions. The SWAT model is highly sensitive to surface runoff and basin parameters when the watershed is characterized by inconsistent rainfall events [50,51]. ALPHA_BF followed by CN2 were the most sensitive parameters in Walzem Creek. In contrast to the other watersheds, snowfall and snow melt parameters were not sensitive in Walzem Creek since there was no persistent snowpack. The high ALPHA_BF constant in Walzem Creek indicated a rapid response to groundwater recharge. However, the lower baseflow recession constant in the UWBDR indicated large storage discharge and slow drainage in the shallow aquifer, which might be attributed to the complex geological structure of the watershed such as the presence of folds and faults [36]. The high deep aquifer percolation parameter (RCHRG_DP) in Walzem Creek indicated the increase of water movement to the deep aquifer. SOL_AWC represented the soil moisture content and hence played a role in surface runoff and was considered to be directly proportional to the soil’s ability to hold water, affecting streamflow.

3.2.4. Statistical Criteria and Model Evaluation Performance

The performance of the SWAT model can be computed using statistical indices and graphical comparisons [52]. For the simulated streamflow and baseflow, the coefficient of determination (R2), Nash–Sutcliffe model efficiency (ENS), PBIAS, and modified KGE were adopted to evaluate the model performance [53,54]. The monthly statistical streamflow and baseflow values for the calibrated models were adopted to evaluate the model performance. The performance of the SWAT model is considered good on a monthly basis when R2 > 0.75; ENS and KGE > 0.7; and PBIAS ≤ 15 according to Moriasi et al. [55] and Thirel et al. [56].

3.2.5. Scenarios Separating the Impact of Land Use Change and Climate Change

In this research, the “change-fix” approach used by Aboelnour et al. [8] was applied to evaluate the streamflow and baseflow as a response to separate and combined impacts of urbanization and climate alteration. Land use maps of 1992 and 2011 were used to represent the two time periods. The land use map of 1992 was adopted to represent the patterns in the first period (1980–1998), herein called TS1. On the other hand, the 2011 land use map was used to represent the second time span (1999–2017), herein called TS2.
A combination of four simulations were developed to evaluate the natural and human impacts on hydrology: (1) 1992 land use and TS1 climate data of 1980–1998 (X1); (2) 2011 land use and TS1 climate data of 1980–1998 (X2); (3) 1992 land use and TS2 climate data of 1999–2017 (X3); and (4) 2011 land use and TS2 climate data of 1999–2017 (X4). The well-calibrated SWAT model, using the land use data of 1992 and first climate period, was used to run the other four scenarios (X1–X4). The simulated output values obtained from these scenarios were compared to the corresponding baseline model. X1 represents the baseline scenario with the corresponding circumstances, while the difference between X4 and X1 simulation describes the combined effects of land use change and climate variation. The comparison between X1 and X2 attempts to depict the separate impact of land use change. Finally, the differences between X3 and X1 outputs emphasize the individual impact of climate alteration.

4. Results and Discussions

4.1. Trends in Hydrologic Components

Statistical analyses were performed on climatological variables using the modified non-parametric Mann–Kendall (MK) test [46] to evaluate possible transition points, trends, and their significance in the time series from 1980 to 2017. The modified MK test statistic, τ, is standardized and can be used in comparing variables that experience differences in their magnitude [57]. A positive slope magnitude indicates an upward trend and vice versa [58]. As shown in Table 3, the slope and the τ -statistics for annual streamflow and baseflow were all positive, except for the baseflow trend in Walzem Creek, which showed a significant decrease in monotonic trend. However, the null hypothesis was accepted in the case of annual streamflow, as it showed an insignificant increasing trend (Figure 3). Results also show that the annual baseflow increased at a significance level greater than 0.1 for the UWBDR watershed, which indicates a slightly increasing trend. However, a significant increasing trend in streamflow during 1980–2017 was detected for the UWBDR watershed (Figure 3).
The increase in average precipitation played an important role in the increasing trend of streamflow for the UWBDR watershed, while the slight increase in streamflow at Walzem Creek was accompanied by decreased precipitation and an increase in temperature as well. Moreover, human activity, such as construction of urban areas on agricultural areas, played a vital role in the amount of streamflow and baseflow. A combination of temperature increase and either a reduction or increase in rainfall are likely the main reasons for climate variation affecting the global water balance. In other words, the magnitude and the directions of these changes will affect any particular change in streamflow and baseflow [20].
The relationship between baseflow and human impacts and climate change varied. The reduction of annual baseflow in the Walzem Creek watershed may be attributed to the reduction of cultivated area and implementation of imperviousness, which in turn has a negative impact on the infiltration rate by increasing the surface runoff, specifically during the wet season of the year [49]. On the other hand, the increasing trend in the annual baseflow in the UWBDR was similar to the trend of the Little Eagle Creek (LEC) watershed mentioned by Aboelnour et al. [8]. This increase is likely caused by several factors, including the influx of water from outside the watershed during the process of urban development and infrastructure and leakage from water supply pipes. Lerner (2002) reported that urbanized catchments are usually associated with leakage rates of 20–50% in sewer systems and septic tanks, causing large amounts of groundwater discharge [59]. Wastewater from the West Chicago Moraine may also provide a significant amount of water, which likely originates outside the catchment. In addition, detention basins play essential roles in increasing baseflow in urban catchments, as water is retained at the surface due to an increasing amount of surface runoff, and then slowly released into the stream as a form of baseflow [8]. Lastly, physiological features may also contribute to this observed trend, including features such as the topography, geology and soil types that result from glacial melting with high porous media, which can play a significant role in increasing infiltration [60].

4.2. Trends in Climatic Components

The MK test was furthermore employed to quantify the monotonic trends of precipitation and temperature in the selected watersheds. Compared to the first climate period (1980–1998), statistical results indicated that the mean air temperature increased by 0.7 °C (from 9.7 to 10.4 °C) and 0.6 °C (from 20.7 to 21.3 °C) during TS2 at the UWBDR and Walzem Creek watersheds, respectively. Average annual precipitation increased by 9.1% (82 mm, from 890 to 972 mm) during TS2 in the UWBDR, while decreasing by 6.5% (56 mm, from 858 to 802 mm) in Walzem Creek (Figure 4).
In the case of UWBDR, the trend of τ-test statistics and the slope of precipitation and temperature were positive and are provided in Table 4. The results show a difference in the monotonic trends of annual temperature and precipitation. For the time series from 1980 to 2017, the annual air temperature increased at a significance level greater than 0.001, which indicates that the long-term trend of temperature is statistically significant. The annual precipitation increased only at a significance level greater than 0.1, indicating a minor increase of precipitation over time and that the trend is statistically insignificant at the 95% confidence level. On the other hand, the average annual precipitation after the change point in Walzem Creek exhibited a slight decrease from the average before the change point. However, the temperature at Walzem Creek showed an increasing trend at the 0.001 significance level, which indicates that the climate at Walzem Creek became warmer and drier during the study period. While the average annual precipitation and temperatures shifted over time, these trends may not reflect the true picture as the change displayed in both may have been seasonally influenced [61]. Therefore, the MK test was further performed at a monthly scale for time series data from 1980 to 2017 (Table 5). The results show that the monotonic trends of the monthly meteorological data for the study were different. For the UWBDR watershed, the results indicate that the monthly temperature showed increasing trends in slope in every month of the year. The monotonic increasing trends of monthly temperature were only statistically significant at a confidence level of p = 0.05 in June, September, and October. Monthly precipitation trend for November decreased significantly, while it showed an insignificant reduction in August, September, and December. The remaining months showed an insignificant increase in monthly precipitation, with the highest increment recorded in June (1.36 mm/month). On the other hand, monthly precipitation in Walzem Creek Watershed showed decreasing trends in February, May, June, October, November, and December, while increasing trends in the other months with the highest increment recorded in September (1.52 mm/month) and maximum reduction recorded in June (1.68 mm/month). Change in monthly precipitation in the Walzem Creek Watershed was insignificant. Similar to the UWBDR watershed, increases in monthly temperature trends were recorded in every month, with only February showing a significant increase in the monotonic trend at a confidence level of p = 0.05.

4.3. Changes in Land Use Characteristics

Cross tabulation analysis and post classification comparison were applied to evaluate the quantity of temporal conversions and nature of changes from one land cover category to another in land use maps of 1992 and 2011 [61,62,63]. In the UWBDR, a comparison of land use maps for the years 1992 and 2011 indicated that the most significant changes occurred in three classes: developed urban, planted, and forest (Figure 5). In 1992, the main land use types were planted and developed areas, which occupied 76.1% of the total watershed area. However, owing to urban expansion, the proportional extent of developed areas increased from 44% to 77% from 1992 to 2011. Conversely, the proportional extent of planted and forest decreased from 35.8% to 2.7% and from 8.1% to 4.4%, respectively. The transition matrix of UWBDR land use in Table 6 explains these changes in detail. Overall, 43.6% or 39.9 km2 of the developed area in 1992 remained unchanged, whereas 27.7 km2 (30.2%) and 5.38 km2 (5.9%) of the planted and forest areas, respectively, were primarily converted to developed urban areas from 1992 to 2011. In hydrological modeling, uncertainties in land use data is determined by the sensitivity of the model output to different land use data inputs. Some uncertainties might be associated with different classification algorithms used in both 1992 and 2011 NLCD land use data. Therefore, the presence of low percentages of land use changes between 1992 and 2011 is omitted. Uncertainties and accuracies in NLCD data are also dependent on the interpretation of the person(s) collecting the information and therefore may be assessed differently depending on how it was analyzed. Some uncertainties, therefore, might be applicable to the intended application, while others may have no effects [64].
The Walzem Creek watershed also underwent some land use changes over the past few decades (Figure 6). During the 20-year period, developed and planted areas were the two largest land use types, and they accounted for approximately 64% and 17% of the total area, respectively. The planted areas shrunk from 1992 to 2011 by 18.3 km2. Developed and wetland areas had the greatest increase from 64% to 92% and from approximately 0% to 7.8%, respectively. These increases were due to a large scale, continuous decrease in planted areas (17.5% to 0.8% of the watershed area) and a gradual decrease in forests (9.5% to 2.5%). The increase in wetland areas mostly occurred after 2006 (from 0.06 to 7.82 km2) due to the ecological restoration program for watershed protection that enhanced the urban reaches, bringing the basin back into compliance with water resources and water quality recreation standards. On the other hand, developed areas increased to the detriment of planted and cultivated areas due to the rapid urban development and expansion in the city of San Antonio (Table 7).

4.4. SWAT Model Calibration and Validation Results

At the UWBDR, the total observed and simulated streamflow during the calibration period were 7.52 and 7.60 m3/s, respectively. The resulting hydrograph from SWAT streamflow in the UWBDR also showed agreement in trends between the two (Figure 7). The best calibration achieved was an R2 of 0.69, PBIAS of 4.86, ENS of 0.67, and KGE of 0.82. Note that KGE was used as an objective function type in the SUFI-2 calibration and validation because it could be decomposed into three terms that represented the correlation, bias, and relative variability between the measured and simulated values [65]. Hence, it allowed the simultaneous use of baseflow and streamflow in calibration and enabled comparison between different strategies. The summed observed and simulated streamflow during the validation period were 9.03 and 8.27 m3/s, respectively. Streamflow validation showed a higher performance than the calibration with an R2 of 0.84, PBIAS of 23.1, ENS of 0.68, and KGE of 0.67 (Table 8).
On the other hand, the total annual baseflow during the calibration and validation periods for both measured and simulated data were 8.02 and 7.82 m3/s, respectively. Goodness-of-fit measures were evaluated to test the performance of baseflow predictions. The R2 for the calibration period was 0.67, with a PBIAS of −1.08, ENS of 0.60, and KGE of 0.80 (Table 8). Figure 8 shows the results of model calibration and validation of baseflow at the UWBDR. Overall, there was reasonably good agreement between computed and simulated baseflow. Further, the model performance was validated using data for the subsequent time period. It was observed that the computed baseflow from the USGS streamflow values were reasonably close to the simulated ones. The evaluation indices R2, PBIAS, ENS, and KGE were 0.79, 8.43, 0.58, and 0.79 for the baseflow of the validation period, respectively.
In general, the results suggested that the SWAT model performed satisfactorily in the UWBDR watershed according to the criteria set by Moriasi et al. [52]. However, the model underestimated the simulated streamflow for the validation period at a monthly time step during low streamflow, which indicates that there may be uncertainty in the results of SWAT simulations for urban watersheds. The lower performance of the SWAT model in the UWBDR may be attributed to the fact that the climate data obtained from the main weather station were located outside the basin, and the distribution of the climate stations with a complete record was sparse. In addition, the overestimation of some peaks in baseflow could be related to the existence of the West Chicago Moraine outwash plain, creating circumstances that promote fast groundwater movement from the moraine through the outwash. Ratios of baseflow to the total annual streamflow were 55.3% and 60.8% for both measured and simulated streamflow, respectively. This discrepancy is acceptable because all of the separation methods of baseflow using different filters are subject to uncertainties [36].
Unlike the UWBDR and the LEC watersheds reported by Aboelnour et al. [8], the baseflow proportion of the observed and simulated streamflow at the Walzem Creek watershed were 33.3% and 26.8%, respectively, which indicated that surface runoff was a major supply component for the stream. Figure 9 shows the comparison between the simulated and observed monthly streamflow for the calibration and validation periods. USGS records show that the total monthly streamflow for Walzem Creek was 18.7 m3/s, while the simulated one was 19.5 m3/s. However, streamflow was overestimated for most of the light rainfall events (dry climate periods) and showed very good agreement with the large rainfall events (wet periods). Previous studies have shown that SWAT performed better under more humid climatic conditions [65,66]. In addition, SWAT has some problems with precisely accounting for water loss through infiltration and evapotranspiration, especially during dry climate seasons, and evaluating the soil moisture storage [67,68,69].
During the streamflow calibration period, the R2, ENS, PBIAS, and KGE were 0.87, 0.87, −4.31, and 0.91, respectively, while they were 0.83, −3.83, 0.70, and 0.52 during the validation period (Table 8). The SWAT performance for the monthly streamflow during both the calibration and validation periods was very good [52]. Moreover, the high values of R2 and ENS in the calibration and validation periods indicated that, with calibrated parameters, the SWAT model was useful to simulate streamflow in semi-arid regions and to further quantify the hydrological impacts of climate variation and land use change over water balance components. Although the SWAT performance for the streamflow validation period was not as good as the calibration period, the results show that its performance was still good, implying that SWAT is applicable to Walzem Creek. The reason that SWAT validation performance was less than the calibration performance is most likely due to the occurrence of an extreme flooding event in October 1998, in which a strong flood killed at least 25 people and caused hundreds of millions of dollars in damages across counties in the southern and eastern regions of San Antonio. The SWAT model poorly matched the peak flow of this large event.
The results also indicate that the simulated values of baseflow were slightly lower than those of the computed ones from observed USGS records. The computed monthly baseflow from USGS records and the simulated one were 6.23 and 5.22 m3/s, respectively, during the whole calibration and validation periods. Figure 10 shows the comparison between the simulated and the computed monthly baseflow values at the Walzem Creek watershed in the calibration and validation periods. In the calibration period, the baseflow of the computed and simulated results had a similar trend. Meanwhile, the values of R2, ENS, PBIAS, and KGE were 0.85, 21.65, 0.76, and 0.70, respectively, with a P-factor of 70% and R-factor of 0.62. In the validation period, these measures were 0.70, −5.12, 0.68, and 0.79, respectively. The statistical measure results indicate a “very good” to “good” match between the simulated baseflow in the calibration and validation periods and the computed records [52]. However, SWAT overestimated the computed baseflow during the validation period, which was exemplified in the negative values of PBIAS. The statistical indicator and the similar trend between the computed and simulated results showed that the SWAT model was adequate in the semi-arid region of Walzem Creek, and confirmed that the optimized and calibrated model can be applied to evaluate the responses of the basin’s hydrology to land use and climate change.
However, considering the study area was in a semi-arid region and only one meteorological station within the catchment was used, it was difficult to detect whether the climatic conditions in the entire watershed were precisely captured. In addition, the design of the SWAT model may not fully capture the groundwater flow characteristics. However, the outputs are expected to be accurate and reliable since the model was calibrated and validated using observed streamflow.

4.5. Impacts of Land Use Change

One of the vital parameters assessed for sustainable management of water resources is water yield. Total water yield is the aggregate amount of water entering the main channel after leaving the HRUs during a time step and can be computed using the following equation [29]:
W Y L D = S U R Q + L A T Q + G W Q T L O S S
where W Y L D is the total water yield (mm); S U R Q is the surface runoff (mm); L A T Q and G W Q   are the contributions of lateral flow and groundwater to streamflow (mm), respectively; and T L O S S is the transmission loss through the bed from the tributary channels in the HRU (mm).
The SWAT simulation suggested that the conversion of the existing planted land cover to urban areas in the UWBDR watershed caused a minor increase in the annual mean water yield by 0.5% (Table 9). The variation could be explained by the reduction in the extent of forests and planted areas and implementation of imperviousness, leading to the reduction of evapotranspiration and infiltration, and increase in surface runoff. However, the reduction of evapotranspiration and the increase in surface runoff were considered not significant in the UWBDR watershed. This could explain the minor increase in the annual mean water yield at the area. Other than the total water yield, the SWAT simulation also suggested a considerable change in baseflow due to the effect of urbanization. It was observed that baseflow increased by 2 mm (accounting for 3.0%) when only the effect of land use dynamics between the two different periods was considered.
Figure 11 and Figure 12 show the distribution of the monthly average water yield and baseflow simulated by SWAT, respectively, for the four scenarios for the UWBDR. We observed that the average monthly water yield was concentrated in the late fall/spring seasons and accounted for 29% in the land use change scenario (X2). The change in water yield tended to be positive under the X2 scenario except for the winter season. On the other hand, land use change had minimal effect on baseflow, with no obvious change between X1 and X2. Baseflow variation showed increasing trends in warm months from May to September, and then decreased from October to April. Such increase may be attributed to leakage from an outwash plain at the base of West Chicago Moraine and the increased precipitation during the wet season.
The results in Table 10 show that the average annual water yield increased by 8.0% due to the urbanization effect in Walzem Creek (X2-X1). Meanwhile, urbanization caused the baseflow to experience a reduction by 26.0%. Based on the proposed approach, the average annual evapotranspiration and surface runoff variability during the three scenarios were further analyzed to provide deeper insight into how climate and land use dynamics interacted with hydrologic systems in Walzem Creek watershed. In semi-arid regions, hydrologic systems could be very sensitive to climate variability. Evapotranspiration was an important component of the hydrologic process, often nearly equaling precipitation in the catchment water balance, and, under given climate conditions, it was mainly affected by vegetation cover [70]. Under the same precipitation conditions, decreased evapotranspiration brought an increase in baseflow and streamflow, while increased evapotranspiration led to the reduction in both [71]. This is illustrated in our findings shown in Table 10, in which evapotranspiration experienced a minor reduction by 1.5% due to the land use alteration. However, the reduction of groundwater discharge reported at Walzem Creek Watershed is mainly due to urbanization, agriculture loss, and deforestation. Of note, the evapotranspiration rate is higher in the UWBDR Watershed than in the Walzem Creek Watershed, despite having a higher potential evapotranspiration (PET), as a result of several key differences. The PET recorded in Walzem Creek Watershed was higher than that of the UWBDR Watershed for all scenarios. The PET ranged from 1016 to 1091 mm for the UWBDR Watershed, but was between 1800 and 2084 mm for the Walzem Creek Watershed. However, the amount of precipitation in the Walzem Creek Watershed is lower than that of the UWBDR Watershed, resulting in dryer soils that limit ET. Moreover, the areas of vegetation and forest cover in the UWBDR Watershed are higher for both 1992 and 2011 (Table 6 and Table 7). In addition, the UWBDR watershed experienced higher increases in temperature in TS2 compared to TS1 (0.7 °C). Finally, the UWBDR Watershed average daily wind speed was 4.5 and 4.3 ms−1 in TS1 and TS2, respectively. In comparison, the average for the Walzem Creek watershed was 3.9 and 3.8 ms−1 in TS1 and TS2, respectively.
Figure 13 illustrates the monthly impacts of land use change, climate, and their joint effect on Walzem Creek’s water yield. Land use change had a more pronounced effect for all months in conjunction with a higher monthly average of rainfall in the first period of time (TS1). For example, the monthly average precipitation in June was 120.2 mm in TS1, and decreased to 75.0 mm in TS2. The contributions of land use impacts on monthly water yield were the highest in May, June, October, and December. Conversely, deforestation and urban expansion resulted in a reduction in monthly baseflow in all months from with the highest reduction recorded in the summer season from May to July with a total of 33.9% (Figure 14).
The streamflow changes in the UWBDR watershed appeared to occur in the same manner as changes for the LEC watershed discussed by Aboelnour et al. [8], with some minor differences (Figure 15). For instance, under X2, streamflow was reduced in winter months by 5.4% to 35.8%. In addition, the average annual water year experienced an increase of 6.7% in the LEC watershed, while it was simulated to be only 0.5% and 8.0% in the UWBDR watershed and Walzem Creek watershed, respectively. In contrast to the LEC watershed, which showed a reduction in average annual baseflow as a result of reducing infiltration rate due to urbanization by 28.8%, the UWBDR watershed experienced an average increase in baseflow regardless of the urbanization trend. The reduction in baseflow in the LEC watershed could be caused by over-exploitation and excessive pumping of groundwater used in industry and production [7], while the minor increase in average annual baseflow in the UWBDR might be attributed to flooding of underground structure and the leakage of the groundwater into wastewater systems. The significant decrease in average annual baseflow at Walzem Creek might be due to clearing vegetation, deforestation, and increasing imperviousness, which in turn led to the reduction of evapotranspiration and groundwater discharge while increasing surface runoff. Urbanization is usually associated with measures that play a vital role in accelerating the removal of water from the catchment and stream system, especially during heavy rainfall events. Compacted soil, channelization, and imperviousness allow water to flow rapidly as a result of lower hydraulic resistance of channels and land surfaces of urbanized catchments, which might be an explanation for decreasing baseflow in the Walzem Creek Watershed.

4.6. Impacts of Climate Variation

In comparison to the land use change scenario, the climate variation scenario caused the average annual water yield to increase by 9.9% as a result of a prominent increase in precipitation at the UWBDR watershed. Baseflow also showed an increase when only climate variation was considered (X3); however, this increase was much more pronounced than the change in water yield, with an amount of 53.8 mm (81.6%) (Table 9). These results indicate that both land use change and climate variability played a role in increasing baseflow. However, climate change played a more pronounced role than land use change in impacting the hydrologic regime of the UWBDR during the recent past, due mainly to the increase in precipitation. This can also be seen in Table 9, in which the surface runoff decreased by 9.1% and evapotranspiration increased by 7.0%. Together, these results indicate that the climate alteration contributes more substantially to the effects observed on hydrological components compared to urbanization.
Similar to the land use change scenario, the average monthly water yield was predominantly observed in late fall/spring at the UWBDR watershed. Of note, the highest change in monthly water yield is observed in July (39%) due to the X3 (climate change) scenario. The change in water yield tended to be positive in months that experienced a significant increase in precipitation in the second period (TS2) compared to the first one (TS1) (Figure 11). On the other hand, the results show an increase in average monthly baseflow under the effect of climate change only, impacts of X3 in all months, although the highest growth was detected in the warmest months of the year (May to September) (Figure 12).
The climate change scenario had the maximum impact on the average annual water yield, causing it to decrease by 11.9% for Walzem Creek Watershed, while it caused the average annual baseflow to decrease by 42.9% (51.7 mm) compared to the baseline scenario (X1) (Table 10). This may be attributed to the significant reduction in the precipitation pattern and the increase in temperature in TS2 as compared to TS1, where the climate became warmer and drier. Therefore, these likely played an important role in the contribution to the total streamflow for Walzem Creek. Climatic variables, specifically precipitation, largely determined the runoff hydrograph. Precipitation reduction in the second climatic period (TS2) resulted in the significant decline of surface runoff by 20.8 mm (8.2%), and a reduction in evapotranspiration by 2.5%, within the X3 scenario (Table 10). These results indicate that impact of climate variability on baseflow and evapotranspiration was larger than the land use alteration scenario; however, both scenarios had opposite impacts on average annual water yield and surface runoff. Overall, the impacts of climate variation were greater than those of land use change.
On a monthly basis, the highest negative impacts of climate change over the monthly average water yield were detected in June, October, and December, with amounts of 19.2, 12.5, and 12.3 mm, respectively, where the average monthly precipitation was significantly higher in TS2 as compared to TS1. On the other hand, monthly water yield increased at the end of summer and beginning of fall seasons, especially in September; in which it increased by 17.5 mm. It could be inferred that climate variation had a lasting negative effect on water yield (Figure 13). Similarly, the climate change scenario (X3) caused a reduction in monthly baseflow in all months except August, September, October, and December, with the highest difference recorded in September (3.1 mm). The increase of baseflow in these months was mainly due to changes in precipitation and temperature patterns from TS1 to TS2. For example, TS2 experienced less precipitation as compared to TS1, while the temperature was higher in TS2 compared to TS1. Hence, baseflow played a role in water contribution to total streamflow when the weather got warmer and drier in the semi-arid watershed (Figure 14). The probable climate alteration impacting most of the globe is mainly determined by the combination of temperature increase and either decrease or increase in rainfall intensity, and any particular baseflow response will depend on the direction and magnitude of both precipitation and temperature. For instance, the change in average monthly precipitation was positive in September, while it was negative in October. Meanwhile, average monthly baseflow showed an increase in both September and October, which might also be explained by the changes in monthly temperature between the two scenarios as it was higher in October than in September.
The streamflow changes under the X3 scenario, which was considered the climate change scenario, were remarkably similar at the UWBDR watersheds to those at the LEC watershed, in which all months exhibited an increase of 12.2–34.5% (Figure 15). In addition, the relative change in streamflow percentage in the UWBDR watershed was higher than the change in the LEC watershed, suggesting that streamflow change is more sensitive to climate change than to land use dynamics. Climate change had a similar impact on the average annual water yield in both the UWBDR and the LEC watersheds, in which it increased by 17.9% and 9.9% in the LEC and UWBDR watersheds, respectively. However, negative impacts due to the X3 scenario occurred in Walzem Creek, indicating that urbanization and climate change had opposite impacts in the semi-arid region of the Walzem Creek Watershed. On the other hand, the climate change caused the average annual baseflow to increase by 15.2% and 81.6% at the LEC and UWBDR watersheds, respectively. However, it declined by 42.9% at the Walzem Creek Watershed. In addition, the average annual surface runoff exhibited an increase in the LEC Watershed due to the impact of climate change by an amount of 22.7%, but decreased in Walzem Creek by 8.2%. These findings imply that the runoff hydrographs of a catchment are largely impacted by climatic variables, especially precipitation, which in turn affects the percolation of soil water to the groundwater.

4.7. Combined Impacts of Both Land Use Change and Climate Variations

To evaluate the combined impacts of land use and climate change, the simulated results under the X4 scenario were compared to the calibrated baseline scenario. The annual mean water yield increased by 11.1% as a response to the X4 scenario at the UWBDR watershed (Table 9). These changes, compared to X2 and X3 scenarios, emphasize that the joint effects of land use change and climate variability led to consistent growth in water yield in the UWBDR watershed. Furthermore, the effect of climate variation was larger than that of the land use dynamic on the total water yield. This can be clearly seen by the X3 and X4 scenarios, in which the mean annual precipitation showed an increase of 81.4 mm, resulting in an increase in the mean annual water yield. These changes are similar to the changes reported in the LEC watershed discussed by Aboelnour et al. [8], resulting from the combined impacts of land use and climate change. In contrast to the LEC watershed, where baseflow decreased, the X4 scenario for the UWBDR watershed led to an increase in the average annual baseflow to 7.1 mm (10.8%) (Table 9). This difference might be attributed to the prevalence of negative urbanization impacts for the LEC watershed, in contrast to the significant positive effects of climate variation for the UWBDR watershed.
Similar to the climate scenario, we observed that the average monthly water yield was concentrated in the late fall and early spring in the X4 scenario, totaling 39% of the annual yield. In general, positive changes were detected in all months under different scenarios except for November, January, and February. However, the variation due to the joint effects tended to be higher in all months with a higher precipitation pattern in the second period of time (TS2) than in TS1 (Figure 11). For instance, the effect of land use scenario (X2) was higher than those of X3 and X4 in August, as the average monthly rainfall was 127.2 mm in TS1, while it was only 99.8 mm in TS2. Meanwhile, the combined effect of land use change and climate variability and the sole effect of climate change had greater impacts on water yield in July, as the average monthly precipitation was 83.7 in TS1, increasing to 114.8 mm in TS2. Baseflow variations showed increasing trends in warm months from May to September, then decreased from October to April in conjunction with the joint effect of climate variation and land use change (Figure 12). The increase in baseflow may be mostly due to an increase in rainfall, and could be explained by fluctuations in both precipitation and temperature between TS1 and TS2. The freeze–thaw processes of the active layer could have changed the soil infiltration capacity and the volume of subsurface water storage, thus impacting baseflow as well [72].
Results from the X4 scenario in the Walzem Creek Watershed indicate that the average annual water yield decreased by 25.7%, while the average annual baseflow showed a consistent reduction by 67.9% (Table 10). Additionally, the annual evapotranspiration was negatively impacted by the joint effect of climate variation and land use change, decreasing by 4.3%. The decline in evapotranspiration is mainly caused by reduction in green cover (Table 7). Compared to X1, the combined effects of land use change and climate variability under X4 decreased surface runoff by 4.2 mm (1.7%). Therefore, with the concurrent reduction in evapotranspiration, average annual water yield, and baseflow had a significant decrease under the X4 scenario These findings indicate that changes of average annual water yield and surface runoff under the joint effects of climate variation and urbanization were smaller than that the changes due to the impacts of the sole impact of climate variation. In other words, the climate alteration had a dominant role, while the land use variation experienced a counteractive role affecting water yield and runoff. The land use change reduced the negative impacts of climate variation by 3.0% and 6.5% for annual water yield and surface runoff, respectively. However, the joint effect of climate variability and land use change on baseflow was higher than the sole impact of the land use change scenario and climate variation scenario. X4 had the greatest negative impact on evapotranspiration. Thus, when the impacts of individual land use change scenario and sole climate alteration happened in the same direction (increase/decrease), the impacts will be intensified when both changes occur at the same time. Of note, the joint effect of climate variation and the urbanization scenario are not a simple summation of each of the individual impacts; however, it represents the interaction of both climate and land use change represented by the SWAT model.
On a monthly basis, the contribution of the joint effects of both climate variability and land use change tended to be similar to the contribution of climate change impacts but with a smaller magnitude for the monthly water yield, with the maximum difference recorded in September (19.0 mm) due to a significant increase in rainfall. Monthly water yield exhibited the highest reduction in June (18.6), due to the notable decline in precipitation in this month. It should be noted that the impact of land use change played a counteractive role for water yield from the Walzem Creek Watershed (Figure 13). Similarly, the highest average monthly change for baseflow was recorded in September (2.5 mm) as a result of the combined impacts of urbanization and climate variation (Figure 14). Generally, the behavior of average monthly baseflow under the combined impact of land use and the climate change scenario was consistent with the changes under the individual impact of climate variation. At a monthly timescale, the streamflow for Walzem Creek Watershed increased only in August, September, and November considering the climate change scenario and the combined scenario of land use and climate variation, with the highest increase in September by 63.7% (X3) and 71.3% (X4), while it exhibited a reduction in all other months with the highest decline recorded in June by 35.0% (X3) and (33.8) (X4). Note that streamflow in Walzem Creek showed a minor increase in all months when considering the impacts of land use change scenario (X2), except for July and August that showed a minor reduction of 0.7% and 1.1%, respectively (Figure 16). Moreover, the streamflow rate tended to decrease when considering scenarios X3 and X4, except in August, September, and November, due to the increase in precipitation during these months. The impact of the combined effect of land use change and climate variability showed the same behaviors as the sole impact of climate variability in the Walzem Creek watershed. This situation is well demonstrated by the monthly streamflow variation in the watershed (Figure 16), with the greatest streamflow increase estimated in September at 71.3%. Meanwhile, the highest reduction in monthly streamflow when considering the individual impacts of climate change was estimated in June at 35.0%. These changes were mainly the result of incremental, dynamic precipitation patterns between the two periods, TS1 and TS2. For instance, September experienced the highest increase in rainfall with an amount of 40.4 mm, while June showed the highest reduction in monthly precipitation with an amount of 45.2 mm in TS2 as compared to TS1.
Compared to the LEC watershed, in which the urbanization had the prevailing negative effect on baseflow while climate changes caused increases in both flows, in Walzem Creek watershed, both land use change and climate change had an impact on streamflow and baseflow. However, our study showed that the climate change impacts played a more important role than land use dynamics and urban expansion on streamflow and baseflow in semi-arid regions.

5. Summary and Conclusions

Urbanization and climate change play an important role in altering the spatiotemporal distribution of water resources and hydrologic components. Streamflow and baseflow are two critically important components of hydrology that are essential to sustain water demands by various sectors, such as agriculture and industry, and are vulnerable to these changes. Therefore, it is of vital significance to understand the behaviors of these components under the separate and combined impacts of climate variation and land use dynamics in different climate regions. In this research, we followed the methodology discussed by Aboelnour et al. [8] for computing streamflow and baseflow for diverse watersheds.
Findings of this research indicate that the climate became warmer and wetter for both the UWBDR and LEC watersheds evaluated by Aboelnour et al. [8] but warmer and drier at the Walzem Creek watershed. The combined effect of these changes showed nonlinear responses to the water balance component. Changes at the UWBDR watershed were remarkably similar to those for the LEC watershed, with the exception that the climate variation was shown to have the greater impact on streamflow, surface runoff, and baseflow, while land use change exerted a relatively small influence on the flow. In other words, in the UWBDR watershed, when the direction of the changes caused by urbanization and climate variation occur in the same direction, the changes of the combined impacts will be intensified. Of note, increasing surface runoff was considered a negative impact as it further strengthened environmental stress by generating more surface erosion and sedimentation. On the other hand, urbanization influenced streamflow positively, while it affected baseflow negatively in the semi-arid Walzem Creek Watershed. However, the climate change had negative impacts on all water components in the area. This might be attributed to the change in rainfall pattern between the two climate periods. The small reduction in mean annual precipitation in the TS2 produced a considerable reduction in runoff. Therefore, the impact of the combined scenario will be amplified when the individual impacts of land use alteration and climate variation are in the same direction (positive/negative). These findings indicate the necessity of evaluating the influences of urbanization and climate alteration separately when assessing the hydrologic effects in urban catchments.
Generally, with the variation in spatiotemporal properties of precipitation, and increasing hazardous events associated with water, such as droughts and floods, stress on water resources will increase and will further encourage the development of mitigation approaches. Based on this research, findings will provide practical suggestions for policy makers on how to sustain water resources more efficiently in relation to its variability as a response to urbanization, land use, and climate change. These changes can be problematic and incur great cost to establish new infrastructure, especially in undeveloped nations. Therefore, policy makers need to develop policies to address these types of changes, taking into account the individual influences of human activities and climate variation, for instance, improving infrastructure to be more resilient to human activities, constructing dams following proper regulations on water resources, and limiting the amount of deforestation, which threatens some hydrological components. In addition, outcomes of this study can be used in quantifying the potential impacts of future projected climate change and land use change. Nevertheless, it might be found that the driving factors interact to impact streamflow and baseflow through chain effects, in which one factor is trying to increase/decrease the magnitude of the other. Hence, more studies are crucial to evaluate this potential future impact on the hydrological system, with the emphasis on the interactive effect of environmental change drivers when predicting future change.
While this research showed the separate and combined impacts of human activity and climate alteration using the SWAT model, modelers should be aware that other types of uncertainties associated with the model exist that may result from observed data, the parameterization process, or from the conceptual model itself. One of the potential shortcomings of this study is that the urbanization processes is an integrated part of the watershed, along with climate alteration. Therefore, it is difficult to discern whether the separate effects of human action and climate change were able to be truly simulated and this issue might therefore create a biased condition. Thus, a suggestion to avoid this limitation in future research is to hypothesize an extreme land use/land cover change that is sensitive to the change instead of a natural system simulated by the model.

Author Contributions

Writing original draft, M.A.; Methodology, M.A. and M.W.G.; Formal analysis, M.A. and B.A.E.; Editing and reviewing, M.W.G. and B.A.E.; Conceptualization, M.W.G.; and Supervision, B.A.E. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Egyptian Government General Scholarship Programme (Ministry of Higher Education), administrated by the Egyptian Cultural and Education Bureau, Washington, DC, USA (Call 2013/2014).

Acknowledgments

The authors would like to thank Erin Sorlien, Celena Alford, and Mirac Eryigit for their critical reviews of the manuscript, and the Egyptian Government for financial support through the General Scholarship Program.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Ficklin, D.L.; Robeson, S.M.; Knouft, J.H. Impacts of recent climate change on trends in baseflow and stormflow in United States watersheds. Geophys. Res. Lett. 2016, 43, 5079–5088. [Google Scholar] [CrossRef]
  2. Zhang, D.; Liu, X.; Liu, C.; Bai, P. Responses of runoff to climatic variation and human activities in the Fenhe River, China. Stoch. Environ. Res. Risk Assess. 2013, 27, 1293–1301. [Google Scholar] [CrossRef]
  3. Tong, S.T.Y.; Sun, Y.; Ranatunga, T.; He, J.; Yang, Y.J. Predicting plausible impacts of sets of climate and land use change scenarios on water resources. Appl. Geogr. 2012, 32, 477–489. [Google Scholar] [CrossRef]
  4. Duan, W.; He, B.; Nover, D.; Fan, J.; Yang, G.; Chen, W.; Meng, H.; Liu, C. Floods and associated socioeconomic damages in China over the last century. Nat. Hazards 2016, 82, 401–413. [Google Scholar] [CrossRef]
  5. Zhang, Y.K.; Schilling, K.E. Increasing streamflow and baseflow in Mississippi River since the 1940 s: Effect of land use change. J. Hydrol. 2006, 324, 412–422. [Google Scholar] [CrossRef]
  6. Kumar, S.; Merwade, V.; Kam, J.; Thurner, K. Streamflow trends in Indiana: Effects of long term persistence, precipitation and subsurface drains. J. Hydrol. 2009, 374, 171–183. [Google Scholar] [CrossRef]
  7. Aboelnour, M.; Engel, B.A. Responses of Streamflow and Baseflow Hydrology to Climate Variability and Land Use Dynamics in an Urban Watershed. In Proceedings of the AGU Fall Meeting Abstracts, Washington, DC, USA, 9–14 December 2018. [Google Scholar]
  8. Aboelnour, M.; Engel, B.A.; Gitau, M.W. Hydrologic Response in an Urban Watershed as Affected by Climate and Land-Use Change. Water 2019, 11, 1603. [Google Scholar] [CrossRef] [Green Version]
  9. Abdi, R.; Yasi, M. Evaluation of environmental flow requirements using eco-hydrologic-hydraulic methods in perennial rivers. Water Sci. Technol. 2015, 72, 354–363. [Google Scholar] [CrossRef]
  10. Jin, H.; Zhu, Q.; Zhao, X.; Zhang, Y. Simulation and prediction of climate variability and assessment of the response of water resources in a typical watershed in china. Water 2016, 8, 490. [Google Scholar] [CrossRef] [Green Version]
  11. Novotny, E.V.; Stefan, H.G. Stream flow in Minnesota: Indicator of climate change. J. Hydrol. 2007, 334, 319–333. [Google Scholar] [CrossRef]
  12. Frans, C.; Istanbulluoglu, E.; Mishra, V.; Munoz-Arriola, F.; Lettenmaier, D.P. Are climatic or land cover changes the dominant cause of runoff trends in the Upper Mississippi River Basin? Geophys. Res. Lett. 2013, 40, 1104–1110. [Google Scholar] [CrossRef]
  13. Duan, W.; He, B.; Takara, K.; Luo, P.; Nover, D.; Hu, M. Impacts of climate change on the hydro-climatology of the upper Ishikari river basin, Japan. Environ. Earth Sci. 2017, 76, 1–16. [Google Scholar] [CrossRef]
  14. Chen, J.; Theller, L.; Gitau, M.W.; Engel, B.A.; Harbor, J.M. Urbanization impacts on surface runoff of the contiguous United States. J. Environ. Manag. 2017, 187, 470–481. [Google Scholar] [CrossRef]
  15. Xu, X.; Scanlon, B.R.; Schilling, K.; Sun, A. Relative importance of climate and land surface changes on hydrologic changes in the US Midwest since the 1930s: Implications for biofuel production. J. Hydrol. 2013, 497, 110–120. [Google Scholar] [CrossRef]
  16. Gebert, W.A.; Radloff, M.J.; Considine, E.J.; Kennedy, J.L. Use of streamflow data to estimate base flowground-water recharge for Wisconsin. J. Am. Water Resour. Assoc. 2007, 43, 220–236. [Google Scholar] [CrossRef]
  17. Neff, B.P.; Day, S.M.; Piggott, A.R.; Fuller, L.M. Base Flow in the Great Lakes Basin. Scientific Investigations Report; US Geological Survey: Reston, VA, USA, 2005; p. 32.
  18. Charles, D. Assessing regional land-use/cover influences on New Jersey Pinelands streamflow through hydrograph analysis. Hydrol. Process. 2007, 21, 185–197. [Google Scholar]
  19. King, R.S.; Scoggins, M.; Porras, A. Stream biodiversity is disproportionately lost to urbanization when flow permanence declines: Evidence from southwestern North America. Freshw. Sci. 2016, 35, 340–352. [Google Scholar] [CrossRef]
  20. Price, K.; Jackson, C.R.; Parker, A.J.; Reitan, T.; Dowd, J.; Cyterski, M. Effects of watershed land use and geomorphology on stream low flows during severe drought conditions in the southern Blue Ridge Mountains, Georgia and North Carolina, United States. Water Resour. Res. 2011, 47. [Google Scholar] [CrossRef]
  21. Price, K. Effects of watershed topography, soils, land use, and climate on baseflow hydrology in humid regions: A review. Prog. Phys. Geogr. 2011, 35, 465–492. [Google Scholar] [CrossRef]
  22. Dey, P.; Mishra, A. Separating the impacts of climate change and human activities on streamflow: A review of methodologies and critical assumptions. J. Hydrol. 2017, 548, 278–290. [Google Scholar] [CrossRef]
  23. Khoi, D.N.; Thom, V.T. Impacts of climate variability and land-use change on hydrology in the period 1981–2009 in the central highlands of vietnam. Glob. Nest J. 2015, 17, 870–881. [Google Scholar]
  24. Li, Z.; Liu, W.; Zhang, X.; Zheng, F. Impacts of land use change and climate variability on hydrology in an agricultural catchment on the Loess Plateau of China. J. Hydrol. 2009, 377, 35–42. [Google Scholar] [CrossRef]
  25. Jothityangkoon, C.; Sivapalan, M.; Farmer, D.L. Process controls of water balance variability in a large semi-arid catchment: Downward approach to hydrological model development. J. Hydrol. 2001, 254, 174–198. [Google Scholar] [CrossRef]
  26. Gitau, M.W.; Chaubey, I. Regionalization of SWAT Model Parameters for Use in Ungauged Watersheds. Water 2010, 2, 849–871. [Google Scholar] [CrossRef]
  27. Arnold, J.G.; Srinivasan, R.; Muttiah, R.S.; Williams, J.R. Large area hydrologic modeling and assesment Part I: Model development. JAWRA J. Am. Water Resour. Assoc. 1998, 34, 73–89. [Google Scholar] [CrossRef]
  28. Liu, G.; He, Z.; Luan, Z.; Qi, S. Intercomparison of a lumped model and a distributed model for streamflow simulation in the Naoli River Watershed, Northeast China. Water 2018, 10, 1004. [Google Scholar] [CrossRef] [Green Version]
  29. Arnold, J.G.; Moriasi, D.N.; Gassman, P.W.; Abbaspour, K.C.; White, M.J.; Srinivasan, R.; Santhi, C.; Harmel, R.D.; Van Griensven, A.; Van Liew, M.W.; et al. SWAT: Model Use, Calibration, and Validation. Trans. ASABE 2012, 55, 1491–1508. [Google Scholar] [CrossRef]
  30. Neitsch, S.; Arnold, J.; Kiniry, J.; Williams, J. Soil and Water Assessment Tool Theoretical Documentation—Version 2009; Technical Report No 406; Texas Water Resources Institute: College Station, TX, USA, 2009; p. 618. [Google Scholar]
  31. Abbaspour, K.C.; Rouholahnejad, E.; Vaghefi, S.; Srinivasan, R.; Yang, H.; Kløve, B. A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model. J. Hydrol. 2015, 524, 733–752. [Google Scholar] [CrossRef] [Green Version]
  32. Zhang, X.; Srinivasan, R.; Arnold, J.; Izaurralde, R.C.; Bosch, D. Simultaneous calibration of surface flow and baseflow simulations: A revisit of the SWAT model calibration framework. Hydrol. Process. 2011, 25, 2313–2320. [Google Scholar] [CrossRef]
  33. Luo, Y.; Arnold, J.; Allen, P.; Chen, X. Baseflow simulation using SWAT model in an inland river basin in Tianshan Mountains, Northwest China. Hydrol. Earth Syst. Sci. 2012, 16, 1259–1267. [Google Scholar] [CrossRef] [Green Version]
  34. Yan, T.; Bai, J.; Lee Zhi Yi, A.; Shen, Z. SWAT-simulated streamflow responses to climate variability and human activities in the Miyun Reservoir basin by considering streamflow components. Sustainability 2018, 10, 941. [Google Scholar] [CrossRef] [Green Version]
  35. Mwakalila, S.; Feyen, J.; Wyseure, G. The influence of physical catchment properties on baseflow in semi-arid environments. J. Arid Environ. 2002, 52, 245–258. [Google Scholar] [CrossRef]
  36. Zhang, L.; Nan, Z.; Xu, Y.; Li, S. Hydrological impacts of land use change and climate variability in the headwater region of the Heihe River Basin, northwest China. PLoS ONE 2016, 11, e0158394. [Google Scholar] [CrossRef]
  37. Burke, C.B.; West, E.; Street, W.M.; Charles, S.; Burke, C.B. West Branch Dupage River Watershed Plan; Christopher B. Burke Engineering West, Ltd.: Rosemont, IL, USA, 2006; Volume 60174. [Google Scholar]
  38. Hejazi, M.I.; Markus, M. Impacts of Urbanization and Climate Variability on Floods in Northeastern Illinois. J. Hydrol. Eng. 2009, 14, 606–616. [Google Scholar] [CrossRef] [Green Version]
  39. Clean River Program San Antonio River Basin. Watershed Characterizations for the Upper San Antonio River, Salado Creek and Upper Cibolo Creek Watersheds; Texas Commission on Environmental Quality: Austin, TX, USA, 2017.
  40. Drury, B.; Rosi-Marshall, E.; Kelly, J.J. Wastewater treatment effluent reduces the abundance and diversity of benthic bacterial communities in urban and suburban rivers. Appl. Environ. Microbiol. 2013, 79, 1897–1905. [Google Scholar] [CrossRef] [Green Version]
  41. Kottek, M.; Grieser, J.; Beck, C.; Rudolf, B.; Rubel, F. World map of the Köppen-Geiger climate classification updated. Meteorol. Z. 2006, 15, 259–263. [Google Scholar] [CrossRef]
  42. Mehan, S.; Neupane, R.P.; Kumar, S. Coupling of SUFI 2 and SWAT for Improving the Simulation of Streamflow in an Agricultural Watershed of South Dakota. Hydrol. Curr. Res. 2017, 8. [Google Scholar] [CrossRef]
  43. Jung, Y.; Shin, Y.; Won, N.I.; Lim, K.J. Web-based BFlow system for the assessment of streamflow characteristics at national level. Water 2016, 8, 384. [Google Scholar] [CrossRef] [Green Version]
  44. Lee, J.; Kim, J.; Jang, W.S.; Lim, K.J.; Engel, B.A. Assessment of baseflow estimates considering recession characteristics in SWAT. Water 2018, 10, 371. [Google Scholar] [CrossRef] [Green Version]
  45. Eckhardt, K. A comparison of baseflow indices, which were calculated with seven different baseflow separation methods. J. Hydrol. 2008, 352, 168–173. [Google Scholar] [CrossRef]
  46. Hamed, K.H.; Rao, R.A. A modified Mann-Kendall trend test for autocorrelated data. J. Hydrol. 1998, 204, 182–196. [Google Scholar] [CrossRef]
  47. Abbaspour, K.C. SWAT-CUP: SWAT Calibration and Uncertainty Programs; EAWAG: Dübendorf, Switzerland, 2015; Volume 130. [Google Scholar]
  48. Welde, K.; Gebremariam, B. Effect of land use land cover dynamics on hydrological response of watershed: Case study of Tekeze Dam watershed, northern Ethiopia. Int. Soil Water Conserv. Res. 2017, 5, 1–16. [Google Scholar] [CrossRef]
  49. Ghazal, K.A.; Leta, O.T.; El-Kadi, A.I.; Dulai, H. Assessment of wetland restoration and climate change impacts on water balance components of the Heeia coastalwetland in Hawaii. Hydrology 2019, 6, 37. [Google Scholar] [CrossRef] [Green Version]
  50. Veith, T.L.; Van Liew, M.W.; Bosch, D.D.; Arnold, J.G. Parameter Sensitivity and Uncertainty in SWAT: A Comparison Across Five USDA-ARS Watersheds. Trans. ASABE 2010, 53, 1477–1486. [Google Scholar] [CrossRef]
  51. Yuan, E.; Nie, W.; Sanders, E. Problems and Prospects of SWAT Model Application on an Arid/Semi-arid Watershed in Arizona. In Proceedings of the 2015 SEDHYD Conference, Reno, NV, USA, 19–23 April 2015. [Google Scholar]
  52. Moriasi, D.N.; Arnold, J.G.; Van Liew, M.W.; Bingner, R.L.; Harmel, R.D.; Veith, T.L. Model Evaluation Guidelines for Systematic Quantification of Accuracy in Watershed Simulations. Trans. ASABE 2007, 50, 885–900. [Google Scholar] [CrossRef]
  53. Gupta, H.V.; Kling, H.; Yilmaz, K.K.; Martinez, G.F. Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling. J. Hydrol. 2009, 377, 80–91. [Google Scholar] [CrossRef] [Green Version]
  54. Nie, W.; Yuan, Y.; Kepner, W.; Nash, M.S.; Jackson, M.; Erickson, C. Assessing impacts of Landuse and Landcover changes on hydrology for the upper San Pedro watershed. J. Hydrol. 2011, 407, 105–114. [Google Scholar] [CrossRef]
  55. Moriasi, D.N.; Gowda, P.H.; Arnold, J.G.; Mulla, D.J.; Ale, S.; Steiner, J.L.; Tomer, M.D. Evaluation of the Hooghoudt and Kirkham Tile Drain Equations in the Soil and Water Assessment Tool to Simulate Tile Flow and Nitrate-Nitrogen. J. Environ. Qual. 2013, 42, 1699–1710. [Google Scholar] [CrossRef] [Green Version]
  56. Thirel, G.; Andréassian, V.; Perrin, C.; Audouy, J.N.; Berthet, L.; Edwards, P.; Folton, N.; Furusho, C.; Kuentz, A.; Lerat, J.; et al. Hydrology under change: An evaluation protocol to investigate how hydrological models deal with changing catchments. Hydrol. Sci. J. 2015, 60, 1184–1199. [Google Scholar] [CrossRef]
  57. Sen, P.K. Estimates of the Regression Coefficient Based on Kendall’s Tau. J. Am. Stat. Assoc. 1968, 63, 1379–1389. [Google Scholar] [CrossRef]
  58. Hirsch, R.M.; Slack, J.R.; Smith, R.A. Techniques of trend analysis for monthly water quality data. Water Resour. Res. 1982, 18, 107–121. [Google Scholar] [CrossRef] [Green Version]
  59. Lerner, D.N. Identifying and quantifying urban recharge: A review. Hydrogeol. J. 2002, 10, 143–152. [Google Scholar] [CrossRef]
  60. Bhaskar, A.S.; Beesley, L.; Burns, M.J.; Fletcher, T.D.; Hamel, P.; Oldham, C.E. Will it rise or will it fall? Managing the complex effects of urbanization on base flow. Freshw. Sci. 2016, 35, 293–310. [Google Scholar] [CrossRef] [Green Version]
  61. Sekaluvu, L.; Zhang, L.; Gitau, M. Evaluation of constraints to water quality improvements in the Western Lake Erie Basin. J. Environ. Manag. 2018, 205, 85–98. [Google Scholar] [CrossRef]
  62. Aboelnour, M.; Engel, B.A. Application of Remote Sensing Techniques and Geographic Information Systems to Analyze Land Surface Temperature in Response to Land Use/Land Cover Change in Greater Cairo Region, Egypt. J. Geogr. Inf. Syst. 2018, 10, 57–88. [Google Scholar] [CrossRef] [Green Version]
  63. Gitau, M.; Bailey, N. Multi-Layer Assessment of Land Use and Related Changes for Decision Support in a Coastal Zone Watershed. Land 2012, 1, 5–31. [Google Scholar] [CrossRef] [Green Version]
  64. Castilla, G.; Hay, G.J. Uncertainties in land use data. Hydrol. Earth Syst. Sci. 2007, 11, 1857–1868. [Google Scholar] [CrossRef] [Green Version]
  65. Lazzari Franco, A.C.; Bonumá, N.B. Multi-variable SWAT model calibration with remotely sensed evapotranspiration and observed flow. Rbrh-Revista Brasileira Recursos Hidricos 2017, 22. [Google Scholar] [CrossRef] [Green Version]
  66. Van Liew, M.W.; Arnold, J.G.; Bosch, D.D. Problem and Potential of Autocalibrating a Hydrlogic Model. Trans. ASABE 2005, 48, 1025–1040. [Google Scholar] [CrossRef] [Green Version]
  67. Van Liew, M.W.; Veith, T.L.; Bosch, D.D.; Arnold, J.G. Suitability of SWAT for the Conservation Effects Assessment Project: Comparison on USDA Agricultural Research Service Watersheds. J. Hydrol. Eng. 2007, 12, 173–189. [Google Scholar] [CrossRef] [Green Version]
  68. Feyereisen, G.W.; Strickland, T.C.; Bosch, D.D.; Sullivan, D.G. Evaluation of SWAT Manual Calibration and Input Parameter Sensitivity in the Little River Watershed. Trans. Am. Soc. Agric. Biol. Eng. 2007, 50, 843–856. [Google Scholar] [CrossRef]
  69. Tobin, K.J.; Bennett, M.E. Using SWAT to model streamflow in two river basins with ground and satellite precipitation data. J. Am. Water Resour. Assoc. 2009, 45, 253–271. [Google Scholar] [CrossRef]
  70. Zhang, L.; Karthikeyan, R.; Bai, Z.; Srinivasan, R. Analysis of streamflow responses to climate variability and land use change in the Loess Plateau region of China. Catena 2017, 154, 1–11. [Google Scholar] [CrossRef]
  71. Schilling, K.E.; Jha, M.K.; Zhang, Y.-K.; Gassman, P.W.; Wolter, C.F. Impact of land use and land cover change on the water balance of a large agricultural watershed: Historical effects and future directions. Water Resour. Res. 2008, 44. [Google Scholar] [CrossRef]
  72. Qin, J.; Ding, Y.; Han, T.; Liu, Y. Identification of the factors influencing the baseflow in the permafrost region of the northeastern qinghai-tibet plateau. Water 2017, 9, 666. [Google Scholar] [CrossRef]
Figure 1. Index map showing location of the study watersheds: (a) Upper West Branch DuPage River in Illinois; and (b) Walzem Creek in Texas.
Figure 1. Index map showing location of the study watersheds: (a) Upper West Branch DuPage River in Illinois; and (b) Walzem Creek in Texas.
Water 12 00191 g001
Figure 2. Flow chart showing the methodology used in this study (modified from Aboelnour et al. [8]).
Figure 2. Flow chart showing the methodology used in this study (modified from Aboelnour et al. [8]).
Water 12 00191 g002
Figure 3. The MK trends for average daily streamflow (a) and baseflow (b) in the UWBDR watershed; and average daily streamflow (c) and baseflow (d) in the Walzem Creek watershed.
Figure 3. The MK trends for average daily streamflow (a) and baseflow (b) in the UWBDR watershed; and average daily streamflow (c) and baseflow (d) in the Walzem Creek watershed.
Water 12 00191 g003
Figure 4. The MK trends for annual precipitation (a) and temperature (b) in the UWBDR watershed; and annual precipitation (c) and temperature (d) in the Walzem Creek watershed.
Figure 4. The MK trends for annual precipitation (a) and temperature (b) in the UWBDR watershed; and annual precipitation (c) and temperature (d) in the Walzem Creek watershed.
Water 12 00191 g004
Figure 5. Land use types in the UWBDR watershed in: (A) 1992; (B) 2011; and (C) the transition between 1992 and 2011.
Figure 5. Land use types in the UWBDR watershed in: (A) 1992; (B) 2011; and (C) the transition between 1992 and 2011.
Water 12 00191 g005
Figure 6. Land use types in the Walzem Creek watershed in: (A) 1992; (B) 2011; and (C) the transition between 1992 and 2011.
Figure 6. Land use types in the Walzem Creek watershed in: (A) 1992; (B) 2011; and (C) the transition between 1992 and 2011.
Water 12 00191 g006
Figure 7. Observed and simulated time series streamflow for the UWBDR watershed during calibration and validation periods.
Figure 7. Observed and simulated time series streamflow for the UWBDR watershed during calibration and validation periods.
Water 12 00191 g007
Figure 8. Observed and simulated time series baseflow for the UWBDR watershed during calibration and validation periods.
Figure 8. Observed and simulated time series baseflow for the UWBDR watershed during calibration and validation periods.
Water 12 00191 g008
Figure 9. Observed and simulated time series streamflow for the Walzem Creek watershed during calibration and validation periods.
Figure 9. Observed and simulated time series streamflow for the Walzem Creek watershed during calibration and validation periods.
Water 12 00191 g009
Figure 10. Observed and simulated time series baseflow for the Walzem Creek watershed during calibration and validation periods.
Figure 10. Observed and simulated time series baseflow for the Walzem Creek watershed during calibration and validation periods.
Water 12 00191 g010
Figure 11. Monthly water yield change for the UWBDR watershed under different scenarios.
Figure 11. Monthly water yield change for the UWBDR watershed under different scenarios.
Water 12 00191 g011
Figure 12. Monthly baseflow change for the UWBDR watershed under different scenarios.
Figure 12. Monthly baseflow change for the UWBDR watershed under different scenarios.
Water 12 00191 g012
Figure 13. Monthly water yield change for the Walzem Creek watershed under different scenarios.
Figure 13. Monthly water yield change for the Walzem Creek watershed under different scenarios.
Water 12 00191 g013
Figure 14. Monthly baseflow change for the Walzem Creek watershed under different scenarios.
Figure 14. Monthly baseflow change for the Walzem Creek watershed under different scenarios.
Water 12 00191 g014
Figure 15. Absolute change in mean monthly streamflow for the UWBDR watershed under different scenarios.
Figure 15. Absolute change in mean monthly streamflow for the UWBDR watershed under different scenarios.
Water 12 00191 g015
Figure 16. Absolute change in mean monthly streamflow for the Walzem Creek watershed under different scenarios.
Figure 16. Absolute change in mean monthly streamflow for the Walzem Creek watershed under different scenarios.
Water 12 00191 g016
Table 1. SWAT input parameters used for the UWBDR and Walzem Creek watersheds calibration of streamflow and baseflow [8].
Table 1. SWAT input parameters used for the UWBDR and Walzem Creek watersheds calibration of streamflow and baseflow [8].
Parameter1 Ext.DescriptionAdjustment1 IV1 LB1 UB
Parameters Controlling Water Balance
ESCOhruSoil evaporation compensation factorR0.950.011
EPCOhruPlant uptake compensation factorR10.011
CANMXhruMax canopy storage R0025
SFTMPbsnSnowfall temp R1−55
SMTMPbsnSnowmelt base temp R0.5−55
TIMPbsnSnow back temp lag factor R10.011
SMFMXbsnMelt factor for snow on 21 June R4.50.0110
SMFMNbsnMelt factor for snow on 21 DecemberR4.50.0110
Parameters Controlling Surface Water Response
CN2mgtInitial SCS Curve numberV--−0.250.25
SURLAGbsnSurface runoff lag coefficientR40.110
Parameters Controlling Sub-Surface Water Response
ALPHA_BFgwBaseflow alpha factor R0.0480.011
GWQMNgwDepth of water for return flowR10000.015000
GW_DELAYgwGroundwater delay timeR310.150
REVAPMNgwDepth of water for evaporationR7500.01250
GW_REVAPgwGroundwater evaporation coefficientR0.020.020.2
RCHRG_DPgwDeep aquifer percolation fractionR0.050.011
Parameters Controlling Soil’s Physical Properties
SOL_AWCsolAvailable water capacity of the soil waterV--−0.250.25
SOL_KsolSaturated hydraulic conductivity V--−0.150.15
Parameters Controlling Channel’s Physical Properties
CH_K2rteEffective hydraulic conductivityR05300
CH_N2rteMain channel Manning’s “n” R0.0140.010.15
1 Ext, Extension; R, Replace by value; V, Multiply by value; IV, Initial values; LB, Lower bound; UB, Upper bound.
Table 2. Top 10 optimized SWAT sensitive parameter values in the UWBDR watershed and Walzem Creek watershed.
Table 2. Top 10 optimized SWAT sensitive parameter values in the UWBDR watershed and Walzem Creek watershed.
UWBDR WatershedWalzem Creek Watershed
RankParameterFittedt-Statp ValueRankParameterFittedt-Statp Value
1ALPHA_BF0.8144.7101CN2−0.10−24.870.00
2CN20.0218.4702ESCO0.995.780.00
3CH_K228.39−13.3403SFTMP0.31−3.120.00
4CH_N20.08−4.7204SMFMN0.86−2.790.01
5SOL_AWC−0.17−4.1305ALPHA_BF0.23−2.510.01
6RCHRG_DP0.01−3.1606RCHRG_DP0.012.470.01
7EPCO0.16−2.9907SOL_AWC0.03−2.070.04
8SMTMP−1.512.480.018GW_DELAY32.14−0.780.44
9SFTMP4.90−2.240.039SURLAG0.920.750.45
10CANMX23.271.950.0510CANMX0.31−0.740.46
Table 3. Temporal trends in annual streamflow and baseflow in the study areas.
Table 3. Temporal trends in annual streamflow and baseflow in the study areas.
AreasStreamflowBaseflow
UWBDR watershedτ-Stat2.2381.848
Slope3.1951.301
α0.001>0.1
Walzem Creekτ-Stat0.277−1.961
Slope2.043−3.335
α>0.10.001
Table 4. Temporal trends in annual precipitation and temperature in the study areas.
Table 4. Temporal trends in annual precipitation and temperature in the study areas.
AreasPrecipitationTemperature
UWBDRτ-Stat0.5032.709
Slope0.8210.037
α>0.10.001
Walzemτ-Stat0.3273.640
Slope1.1790.036
α>0.10.001
Table 5. Summary of significance test and trend analysis for monthly precipitation and temperature in the UWBDR and Walzem Creek watersheds.
Table 5. Summary of significance test and trend analysis for monthly precipitation and temperature in the UWBDR and Walzem Creek watersheds.
UWBDRWalzem Creek
τ-StatSlope1 Sigp-Valueτ-StatSlope1 Sigp-Value
JanuaryPRCP1.2450.551NS0.1060.7040.564NS0.240
TEMP0.8050.045NS0.2101.7220.051NS0.042
FebruaryPRCP0.9050.409NS0.183−0.905−0.142NS0.183
TEMP0.3390.005NS0.3672.3510.071S0.009
MarchPRCP0.1260.035NS0.4500.8550.248NS0.196
TEMP0.7290.043NS0.2331.6850.048NS0.046
AprilPRCP0.8050.425NS0.2110.6290.992NS0.265
TEMP1.3830.036NS0.0831.7220.038NS0.043
MayPRCP1.5840.988NS0.057−0.704−0.268NS0.241
TEMP0.9810.024NS0.1630.8930.017NS0.186
JunePRCP1.5341.364NS0.061−1.282−1.678NS0.100
TEMP2.0120.051S0.0221.7980.03NS0.036
JulyPRCP−0.4270.581NS0.3350.7290.741NS0.233
TEMP0.4650.014NS0.3211.0310.013NS0.151
AugustPRCP−0.805−1.349NS0.210−0.6410.493NS0.261
TEMP1.3580.021NS0.0871.5850.025NS0.056
SeptemberPRCP−0.855−0.442NS0.1961.3831.52NS0.083
TEMP2.3640.054S0.0091.2450.019NS0.107
OctoberPRCP0.1510.25NS0.440−0.930−0.547NS0.176
TEMP2.0870.058S0.0181.2070.031NS0.114
NovemberPRCP−2.024−1.669S0.021−1.471−0.504NS0.071
TEMP1.3200.043NS0.0931.8860.042NS0.030
DecemberPRCP−0.226−0.324NS0.4110.176−0.238NS0.430
TEMP0.5660.051NS0.2861.1190.04NS0.132
1 Significant level (α) = 0.05. S, Significant; NS, Not significant.
Table 6. Transition matrix (in percentages) of land use change in UWBDR from 1992 to 2011.
Table 6. Transition matrix (in percentages) of land use change in UWBDR from 1992 to 2011.
19922011
WaterDevelopedBarrenForestShrublandHerbsPlantedWetlandsTotal
Water0.770.820.010.080.000.070.140.112.00
Developed0.0443.560.000.220.000.080.140.0144.06
Barren0.012.590.070.030.010.650.210.023.59
Forest0.075.870.001.520.010.110.161.098.82
Herbs0.010.390.000.160.000.020.000.100.69
Planted0.2030.250.472.030.233.521.960.3539.01
Wetlands0.240.760.010.360.010.070.030.361.84
Total1.3484.240.564.400.264.522.652.04
Table 7. Transition matrix (in percentages) of land use change in Walzem Creek from 1992 to 2011.
Table 7. Transition matrix (in percentages) of land use change in Walzem Creek from 1992 to 2011.
19922011
WaterDevelopedBarrenForestShrublandHerbsPlantedWetlandsTotal
Water0.090.090.000.000.000.000.000.000.18
Developed0.0062.630.050.260.380.050.090.9564.40
Barren0.000.090.000.000.000.000.000.000.09
Forest0.003.420.001.150.610.050.113.328.66
Herbs0.001.320.000.050.430.120.210.002.13
Shrubland0.004.790.000.231.470.090.120.266.95
Planted0.0012.150.000.871.260.290.292.6417.50
Wetlands0.000.090.000.000.000.000.000.000.09
Total0.0984.570.052.554.160.590.827.17
Table 8. Statistical indicators for calibration and validation periods for streamflow and baseflow in the UWBDR watershed and Walzem Creek watershed.
Table 8. Statistical indicators for calibration and validation periods for streamflow and baseflow in the UWBDR watershed and Walzem Creek watershed.
PeriodStreamflow (m3/s)Baseflow (m3/s)
R2ENSPBIASKGER2ENSPBIASKGE
UWBDRCalibration(1984–1993)0.690.674.90.820.670.60−1.10.80
Validation(1994–1998)0.840.6823.10.670.790.588.40.79
WalzemCalibration(1984–1993)0.870.87−4.30.910.850.7621.60.70
Validation(1994–1998)0.830.70−3.80.540.700.68−5.120.79
Table 9. Average annual change in water balance components in the UWBDR watershed.
Table 9. Average annual change in water balance components in the UWBDR watershed.
ScenarioLand UseClimatePrecipitation (mm)Water Yield (mm)Baseflow (mm)Surface Runoff (mm)Evapotranspiration (mm)
Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)
X11992TS1890.8--312.6--65.9--238.3--568.6--
X22011TS1890.800314.11.50.567.92.03.0240.92.591.1567.8−0.8−0.1
X31992TS2972.281.49.1343.430.89.9119.753.881.6216.7−21.57−9.1608.6407.0
X42011TS2972.281.49.1347.134.611.173.07.110.8268.430.1112.6605.236.66.4
Table 10. Average annual change in water balance components in Walzem Creek watershed.
Table 10. Average annual change in water balance components in Walzem Creek watershed.
ScenarioLand UseClimatePrecipitation (mm)Water Yield (mm)Baseflow (mm)Surface Runoff (mm)Evapotranspiration (mm)
Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)Av.Ch. Δ(%)
X11992TS1857.7--326.1--95.2--223.3--528.3--
X22011TS1857.700334.18.02.787.3−7.9−26.0239.516.16.4520.0−8.3−1.5
X31992TS2802−55.7−6.5291.8−34.3−11.982.2−13.0−42.9202.6−20.8−8.2514.9−13.4−2.5
X42011TS2802−55.7−6.5300.5−25.7−8.974.6−20.7−67.9219.1−4.2−1.7505.1−23.2−4.3

Share and Cite

MDPI and ACS Style

Aboelnour, M.; Gitau, M.W.; Engel, B.A. A Comparison of Streamflow and Baseflow Responses to Land-Use Change and the Variation in Climate Parameters Using SWAT. Water 2020, 12, 191. https://doi.org/10.3390/w12010191

AMA Style

Aboelnour M, Gitau MW, Engel BA. A Comparison of Streamflow and Baseflow Responses to Land-Use Change and the Variation in Climate Parameters Using SWAT. Water. 2020; 12(1):191. https://doi.org/10.3390/w12010191

Chicago/Turabian Style

Aboelnour, Mohamed, Margaret W. Gitau, and Bernard A. Engel. 2020. "A Comparison of Streamflow and Baseflow Responses to Land-Use Change and the Variation in Climate Parameters Using SWAT" Water 12, no. 1: 191. https://doi.org/10.3390/w12010191

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop