Estimating the effects of potential climate and land use changes on hydrologic processes of a large agriculture dominated watershed

Accepted Manuscript Estimating the effects of potential climate and land use changes on hydrologic processes of a large agriculture dominated watershed Ram P. Neupane, Sandeep Kumar PII: DOI: Reference:

S0022-1694(15)00556-9 http://dx.doi.org/10.1016/j.jhydrol.2015.07.050 HYDROL 20616

To appear in:

Journal of Hydrology

Received Date: Revised Date: Accepted Date:

27 April 2015 22 July 2015 29 July 2015

Please cite this article as: Neupane, R.P., Kumar, S., Estimating the effects of potential climate and land use changes on hydrologic processes of a large agriculture dominated watershed, Journal of Hydrology (2015), doi: http:// dx.doi.org/10.1016/j.jhydrol.2015.07.050

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Estimating the effects of potential climate and land use changes on hydrologic processes of a large agriculture dominated watershed

Ram P. Neupanea, Sandeep Kumara,*

a

Department of Plant Science, South Dakota State University, Brookings, SD 57007, USA

*

Corresponding author: E-mail: [email protected] (S. Kumar)

Abstract: Land use and climate are two major components that directly influence catchment hydrologic processes, and therefore better understanding of their effects is crucial for future land

use planning and water resources management. We applied Soil and Water Assessment Tool (SWAT) to assess the effects of potential land use change and climate variability on hydrologic processes of large agriculture dominated Big Sioux River (BSR) watershed located in North Central region of USA. Future climate change scenarios were simulated using average output of temperature and precipitation data derived from Special Report on Emission Scenarios (SRES) (B1, A1B, and A2) for end-21st century. Land use change was modeled spatially based on historic long-term pattern of agricultural transformation in the basin, and included the expansion of corn (Zea mays L.) cultivation by 2, 5, and 10%. We estimated higher surface runoff in all land use scenarios with maximum increase of 4% while expanding 10% corn cultivation in the basin. Annual stream discharge was estimated higher with maximum increase of 72% in SRESB1 attributed from higher groundwater contribution of 152% in the same scenario. We assessed increased precipitation during spring season but the summer precipitation decreased substantially in all climate change scenarios. Similar to decreased summer precipitation, discharge of the BSR also decreased potentially affecting agricultural production due to reduced future water availability during crop growing season in the basin. However, combined effects of potential land use change with climate variability enhanced for higher annual discharge of the BSR. Therefore, these estimations can be crucial for implications of future land use planning and water resources management of the basin.

Keywords: Climate change, Hydrologic process, Stream discharge, SWAT, Land use change 1.

Introduction

Land use and climate are two main factors that affect watershed hydrologic processes (Costa et al., 2003; Brath et al., 2006; Wang et al., 2006; Wu et al., 2012; Chien et al., 2013). Land use, a major global research issue (Foley et al., 2005; Marshall and Shortle, 2005), is considered as one of the most important components of terrestrial environment system (Lin et al., 2007; Lin et al., 2009) that affects surface runoff, stream discharge, and sediment transportation influenced by rainfall interception, evapotranspiration, and surface soil hydraulic conductivity (He et al., 2008; Germer et al., 2009; Scheffler et al., 2011; Munoz-Villers and McDonnell, 2013; Yan et al., 2013). Recently, land use change impacts associated with deforestation and agricultural transformation on water resources have created social and political problems at both local and national levels. Considerable stress on water supply including seasonal variations and downstream water quality issues are widely observed. Changes in water supply and quality caused by land use change have become very critical issues (Kundzewicz et al., 2007) that affect hydrologic functions of surface water and groundwater resources (Fohrer et al., 2005; Stonestrom et al., 2009). However, these effects vary as functions of seasonality and the changing climate (Huxman et al., 2005). Knowing these responses, we can address the questions of how the on-going land use and climate changes may have influenced the annual and seasonal hydrologic components, and nutrients transportation in the system. Answers to these questions will improve the predictability of hydrologic consequences of these changes that directly influence the daily life of a large number of population downstream of the watershed. Therefore, we require the knowledge of how water resources are affected by these changes of various aspects of regional hydrologic cycle. Global climate change and associated impacts on water resources are the most urgent challenges facing mankind today and will have enduring societal implications for generations to

come. Potential impacts may include the changes in watershed hydrologic processes including timing and magnitude of surface runoff, stream discharge, evapotranspiration, and flood events, all of which would influence other environmental variables such as nutrient and sediment flux on water sources (Simonovic and Li, 2004; Zhang et al., 2005; Zierl and Bugmann, 2005) affecting every aspect of human life (Xu 1999, 2000). Water resources managers are facing difficult decisions regarding current water resources and future management strategies based on dwindling freshwater supplies and human population growth. Therefore, better estimation of hydrologic responses to both land use change and climate variability is crucial, and separation of their effect is of great importance for future land use planning and water resources management, specifically for large agriculture dominated watershed. Though a number of studies have highlighted the concerns of river water qualities, there is clear lack of assessing the impacts of potential land use and climate changes on hydrologic processes of large-scale watersheds. Therefore, this study is one of the first to estimate potential land use and climate change effects on hydrologic processes of large agriculture dominated Big Sioux River (BSR) watershed that may be crucial for better future land use planning and sustainable water resources management of the region. The major objectives of this study are: (1) to assess how the potential land use change affects surface runoff and total water yield of the BSR basin through simulation modeling, (2) to estimate the effects of climate variability on key hydrologic processes including precipitation, stream discharge, water yield, groundwater contribution to stream discharge, water percolation, evapotranspiration, snowfall, and snowmelt in the basin, and (3) to model the combined effects of potential land use change and climate variability on future water availability of the basin.

2.

Materials and methods

2.1.

Study site The BSR watershed located in north-central part of the United States (Figure 1) covers an

area of 21,033 km2 with low elevation ranges between 284 and 663 m above mean sea level. The BSR is a permanent, natural river that flows north to south along eastern edge of South Dakota and eventually drains into the Missouri River. Geologically, the watershed is composed of Precambrian Sioux Quartzite, Dakota Sandstone, Granerous Shale, Greenhorn Limestone, Carlile Shale, Niobrara Chalk, and Pierre Shale (Pirner, 2004), that form a complex geologic setting of the basin potentially influencing the interaction between groundwater and surface water reservoirs. The land use is primarily dominated by crop cultivation including corn, soybean, wheat, grassland, water, wetland, and urban with contributions of 36, 24, 1, 27, 3, 1, and 8%, respectively (Table 1). The dominant soils of the basin, derived from Soil Survey Geographic Data (SSURGO) (http://websoilsurvey.sc.egov.usda.gov/App/WebSoilSurvey.aspx), are silt clay loam, sandy loam, and silt loam. The average annual temperature of the region is 7.6oC with minimum and maximum values of -7.3 and 22.7oC during January and July months, respectively (SDSU, 2003). The average annual precipitation is 544 mm with major contribution (40%) during summer months (June-August), and the seasonal snowfall is about 1062 mm with 48% contribution during winter months (December-February). The Big Sioux River (BSR) has been subject to large fluctuation in water quantity in recent decades. These natural fluctuations, coupled with land use change and climate variability have led to deteriorate water quality of the river. About 60% of the BSR watershed is used for agricultural production with high application of chemical fertilizers, so the leading source of water pollution is fertilizer runoff leading to the river being ranked 13th most polluted in the United States (Kerth and Vinyard, 2012).

2.2.

Modeling approach For this study, we used the SWAT (version 2012) (Arnold et al., 1998), a river basin

model developed for the U.S. Department of Agriculture (USDA), that incorporates the effects of weather, surface runoff, evapotranspiration, groundwater flow, crop growth, and nutrient yield as well as the long term effects of varying agricultural management practices. The water balance equation which governs the hydrologic components of the model (Neitsch et al., 2011) is as follows.

𝑆𝑊𝑡 = 𝑆𝑊0 + (𝑅𝑑𝑎𝑦 −𝑄𝑠𝑢𝑟𝑓 − 𝐸𝑎 − 𝑤𝑠𝑒𝑒𝑝 − 𝑄𝑔𝑤 )

(1)

where SWt is final soil water content (mm), SW0 is initial soil water content (mm), t is time in days, Rday is amount of precipitation (mm), Qsurf is amount of surface runoff (mm), Ea is amount of evapotranspiration (mm), Wseep is amount of water entering the vadose zone from the soil profile (mm), and Qgw is amount of return flow (mm). The surface runoff was generated using a modified Soil Conservation Service (SCS) curve number (CN) method (USDA, 1986) based on local land use, soil type, and antecedent moisture conditions. Actual evapotranspiration was computed using an exponential function of soil depth and water content (Ritchie, 1972) whereas the potential evapotranspiration (PET) was modeled through Penman-Monteith approach (Monteith 1965). However, Snowmelt in the model was estimated through mass balance approach as presented in the Equation 2. 𝑆𝑁𝑂 = 𝑆𝑁𝑂 + 𝑅𝑑𝑎𝑦 − 𝐸𝑠𝑢𝑏 − 𝑆𝑁𝑂𝑚𝑙𝑡

(2)

where SNO is total amount of water in snowpack on a given day (mm H2O), Esub is amount of sublimation (mm H2O), and SNOmlt is amount of snowmelt (mm H2O). The changes in snowpack volume depends on additional snowfall or release of meltwater in the basin. The kinematic storage model (Sloan and Moore, 1984) that primarily accounts for soil hydraulic conductance, soil moisture, and slope was used to compute soil interflow. The baseflow components for this study were modeled as in Equation 3 (Arnold et al., 1998), all expressed in millimeter.

𝑄𝑔𝑤 ,𝑖 = 𝑄𝑔𝑤 ,𝑖−1 𝑒𝑥𝑝 −𝛼𝑔𝑤 𝛥𝑡 + 𝑤𝑟𝑐ℎ𝑟𝑔,𝑠ℎ 1 − 𝑒𝑥𝑝 −𝛼𝑔𝑤 𝛥𝑡

(3)

where Qgw is groundwater flow into the main channel on day i, Qgw,i-1 is groundwater flow into the main channel on day i-1, αgw is baseflow recession constant, Δt is time step (1 day), and Wrchrg is amount of recharge entering the shallow aquifer on day i. Finally, the net water yield (WYLD in mm) to the stream channel was computed by the following equation.

𝑊𝑌𝐿𝐷 = 𝑆𝑈𝑅𝑄 + 𝐿𝐴𝑇𝑄 + 𝐺𝑊𝑄 − 𝑇𝐿𝑂𝑆𝑆

(4)

where SURQ is surface runoff (mm), LATQ is lateral flow contribution to stream discharge (mm), GWQ is groundwater contribution to stream discharge (mm), and TLOSS is transmission losses from the system (mm).

2.3.

Required input data The SWAT model requires the data for topography, land use, soil, weather/climate, and

stream discharge. For this study, we used 30 × 30 m resolution global digital elevation model (GDEM) data derived from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (http://gdem.ersdac.jspacesystems.or.jp/search.jsp) for delineating sub-basins and also for defining stream, area, and slope of the sub-basins. Land use data for this study was

obtained from the National Agricultural Statistics Service (NASS) Cropland Data Layer (CDL) (http://nassgeodata.gmu.edu/CropScape/) of the United States Department of Agriculture (USDA). The NASS CDL is derived from satellite imagery and maps agricultural land cover at very high crop-type specificity at 30 and 56 m spatial resolutions (Boryan et al., 2011). However, for our simulations, we used the land use/cover data from 2006 since the complete NASS dataset for the study site was available from that year onward. Soil data was obtained from the Soil Survey Geographic Data (SSURGO), the most detailed soil database currently available for the US at a range of scales between 1:12,000 and 1:24,000 (USDA, 2011). This dataset can be crucial for better representation of soil characteristics for larger scale projects (NRCS, 2009) which might reflect in model performance outputs. Finally, all the spatial datasets were coregistered to the projection of WGS-1984 UTM Zone 14N using ArcGIS 10.0 for further simulations. Spatial variability within the watershed was represented by 25 sub-watersheds, which were further sub-divided into hydrologic response units (HRUs) based on specific land use, soil, and slope characteristic features. Finally, we defined 227 HRUs using thresholds of 5, 10, and 20% for land use, soil, and slope classes, respectively since characterization of multiple HRUs is the best way for accurate streamflow simulations in a watershed system (Haverkamp et al., 2002).

2.4.

Hydro-meteorological data SWAT requires daily precipitation, maximum and minimum air temperature, solar

radiation, wind speed, and relative humidity as forcing data for hydrologic simulation. For this

study, daily precipitation and temperature data were derived from Daily Surface Weather and Climatological Summaries (DAYMET) Single Point Data Extraction (SPDE) (http://daymet.ornl.gov/dataaccess) (Thornton et al., 1997; 2012) for five different spatial locations as Stations 1, 2, 3, 4, and 5, and presented in the Table 2. The DAYMET dataset is available on a daily timescale from 1980 to the present at high (1-km) spatial resolution that uses a weighting algorithm and digital elevation model to generate daily surface air temperature and precipitation by interpolation from observations of ground-based meteorological stations (e.g. National Oceanic and Atmospheric Administration and National Weather Service). Also, this is a commonly used meteorological dataset for hydrologic simulation of large-scale watersheds (Manter et al., 2005; White et al., 2006; Keane et al., 2008; Oubeidillah et al., 2014). However, remaining meteorological inputs were automatically generated within SWAT using daily precipitation and temperature data. To compare with model simulation outputs, we used 15 years (1987-2001) of daily measured stream discharge data taken from the USGS site no. 06485500 (http://waterdata.usgs.gov/nwis) located at Akron (Table 2).

2.5.

Management operations Since major portion of agricultural land in the BSR watershed covered by cornfield

(about 36%), corn was designated as the major crop type to apply for management operations in the basin. The Urea (46-0-0), Di-ammonium Phosphate (DAP) (18-46-0), and Mono-ammonium Phosphate (MAP) (11-52-0) were applied as primary fertilizers for growing corn with application rates of 168, 56, and 56 kg/ha, respectively (the data was obtained from soil expert at SDSU) with no-till operation. The management operations including fertilization, planting, and

harvesting were adopted based on the potential heat unit values (Neitsch et al., 2011) for corn cultivation in the basin (Table 3).

2.6.

Calibration, confirmation, and sensitivity analyses Before calibration, 15 years (1987-2001) of model simulated stream discharge using

SWAT default parameter values were compared with measured discharge data, and referred as pre-calibration simulation. Then, simulation periods were organized by establishing a calibration (1987-1996), and confirmation (1997-2001) periods. However, for calibration, the initial three years (1987-1989) of simulated outputs were disregarded for our analysis as warm-up period that allows the model to cycle multiple times to minimize the effect of user estimated parameter values (Zhang et al., 2007). The SWAT-CUP (Abbaspour et al., 2007; Abbaspour, 2012) was applied for both calibration and confirmation analyses. Due to better capability of handling large number of parameters in less number of model runs, the Sequential Uncertainty Fitting (SUFI-2) algorithm, a semi-automatic inverse modeling approach, was used for this study (Yang et al., 2008). The global sensitivity analysis integrated within SUFI-2 was used to test 19 SWAT hydrologic parameters for stream discharge simulation in parallel with the calibration procedure. Thus derived new parameter values obtained from calibration and confirmation analyses were incorporated with the SWAT database for further simulations. For calibration and confirmation analyses, simulated stream discharge data were compared to the measured values on both daily and monthly basis and the model performance was judged by using Nash-Sutcliffe efficiency (NSE) (Nash and Sutcliffe, 1970; Moriasi et al., 2007), and percentage of bias (PBIAS) (Gupta et al., 1999) indices as follows.

(𝑌 𝑜𝑏𝑠 −𝑌 𝑠𝑖𝑚 )2

𝑁𝑆𝐸 = 1 −

𝑃𝐵𝐼𝐴𝑆 =

(𝑌 𝑜𝑏𝑠 −𝑌 𝑚𝑒𝑎𝑛 )2

(𝑌 𝑜𝑏𝑠 −𝑌 𝑠𝑖𝑚 ) 𝑌 𝑜𝑏𝑠

× 100

(5)

(6)

where Yobs is the measured data, Ysim is the model simulation output, and Ymean is the mean of measured data. The NSE is widely applied in hydrologic models that ranges from zero to one, one being the ideal value (Moriasi et al., 2007). The PBIAS measures the average deviation of simulated outputs from observed values with zero as the ideal value (Gupta et al., 1999). A positive (negative) PBIAS value shows an underestimation (overestimation) bias of the simulated variable compared to the measured variable. Furthermore, coefficient of determination (R2) and root mean square error (RMSE) (m3/s) (Equation 7) were also used to analyze the goodness of calibration. The R2 shows correlation between model simulation and measured values, one being the ideal value. The model performance can be categorized as satisfactory (good) if NSE>0.5 (0.65) and PBIAS<±25% (±15%) (Moriasi et al., 2007). We evaluated these statistical measures for both daily and monthly stream discharge simulations. However, monthly simulation values were used to assess potential land use and climate change impacts on watershed hydrologic processes. 𝑅𝑀𝑆𝐸 =

1 𝑁

𝑁 ′ 𝑖=1(𝑄𝑖

− 𝑄𝑖 )2

(7)

where 𝑄 is the simulated discharge value at time step i and Qi is the observed discharge value.

2.7.

Potential land use and climate change scenarios

Analysis of multi-year land use data obtained from the NASS dataset showed the increase in agricultural land by 5% within 7 years (2006-2013) of time period in the BSR basin similar to the findings of Wright and Wimberly (2013). The change was substantially higher for corn cultivation with the increased value of 9% during same period of time. Considering all these real situations of land use change, we developed three potential land use scenarios which include: 2, 5, and 10% increases in corn cultivation as Scenario-I, Scenario-II, and Scenario-III, respectively, and new land use maps were derived (Figure 2). Thus derived land use data were incorporated into the SWAT model for determination of new HRUs for each sub-basin, with the same set of parameters used during calibration period. To estimate the effect of these potential land use scenarios on watershed hydrologic processes including surface runoff and total water yield, model simulations were run for the same time period from calibration; however, we ignored the initial three years data for analysis as model warm-up period. For future climate change simulations, we used daily Bias-Correction Constructed Analogues (BCCA) average temperature and precipitation data estimated for the Special Report on Emission Scenarios (SRES) including low (B1), medium (A1B), and high (A2) emission scenarios derived from 8 different models (except mpi_echam5) for 2090s having the spatial resolution of 1/8 degree (http://gdo-dcp.ucllnl.org/downscaled_cmip_projections/) (Maurer et al., 2010; Brekke et al., 2013). To assess climate changes effects on watershed hydrologic processes including precipitation, stream discharge, water yield, groundwater contribution to stream discharge, water percolation, evapotranspiration, snowfall, and snowmelt, the daily average temperature and precipitation data obtained from climate change scenarios were incorporated into the baseline SWAT database. Then, SWAT simulations were run for the period of 2090 to 2099 at monthly basis; however, we ignored the initial three years simulation data for analysis as

model warm-up period. Finally, we also estimated the effect of future climate on stream discharge and total water yield of the BSR basin combined with potential land use scenarios. Simulation results obtained from all these scenarios were compared with the baseline simulation output as in the following equation. % 𝐶ℎ𝑎𝑛𝑔𝑒 = (

𝑌 𝑠𝑖𝑚 −𝑌 𝑏𝑎𝑠𝑒 𝑌 𝑏𝑎𝑠𝑒

) ×100

(8)

where Ybase is the data for baseline scenario.

3.

Results and discussion

3.1.

Analysis of long term precipitation and stream discharge Figure 3 shows the hydrographs of long term precipitation and stream discharge on both

annual and seasonal basis. Observation of 34 years (1980-2013) of precipitation data obtained from DAYMET dataset demonstrated that higher amount of precipitation occurs from April to September (Figure 3a). Winter months are dry with little precipitation. The mean minimum January precipitation is 13 mm and the maximum June precipitation reaches up to 111 mm. Mostly, the rate of precipitation corresponded with stream discharge values. The minimum stream discharge of 13 m3/s was corresponded with the rate of precipitation. However, the higher stream discharge was measured from March to June with maximum value of 123 m3/s on April potentially due to higher spring snowmelt to cause higher runoff in these months (Groisman et al., 2001). Seasonal analysis of precipitation and stream discharge data indicated that 42% of total annual precipitation in the BSR basin occurred during summer season that corresponded with 30% of total stream discharge in the same season (Figure 3b). The minimum precipitation contribution of 6% occurred during winter season corresponding to 8% discharge contribution

during the same season. The most effective season for runoff is the early spring when snowmelt and precipitation rates generally exceed the potential for evapotranspiration. We analyzed the maximum spring discharge contribution of 47% primarily due to dominant meltwater contribution during the spring season (Groisman et al., 2001).

3.2.

Calibration and confirmation analysis of the SWAT model For calibration, the key parameters that govern hydrologic processes of the study

watershed were selected based on previous literature sources (e.g. Muleta and Nicklow, 2005; Rostamian et al., 2008; Neupane et al., 2014, 2015). Evaluation coefficients of the simulated stream discharge by various indices are presented in the Table 4. Pre-calibration outputs for daily simulation showed a lower correspondence with the values of -9.95, -275, 371, and 0.16 for NSE, PBIAS, RMSE, and r2, respectively. For monthly simulation, the values were -8.15, -275, 282, and 0.38 for these statistics, respectively. We found a surprisingly good correlation between measured and simulated discharge values during both calibration and confirmation periods. However, the model showed a better performance for monthly simulations with NSE = 0.70, PBIAS = -3, RMSE = 53, and r2 = 0.75. Similar to calibration outputs, confirmation analysis also followed the higher correlation of simulated stream discharge vs measured values with NSE = 0.60, PBIAS = -14, RMSE = 59, and r2 = 0.67 at daily simulation. The model showed better performance in monthly simulations with the values of 0.70, -14, 48, and 0.86 for these statistics, respectively. We analyzed a higher negative PBIAS values during confirmation period indicating higher average difference between measured and model simulated values for both daily and monthly simulations. Finally, the mean daily stream discharge of 69 m3/s matched surprisingly with the measured value of 70 m3/s for the calibration period. The hydrographs obtained after

model calibration substantially improved the fit of modeled vs measured stream discharge values (Figure 4). Therefore, all these analyses might show the applicability of SWAT model in largescale BSR basin for future hydrologic assessment of potential land use change and climate variability. However, the model was not able to simulate some of the higher flows that might be attributed to accuracy in measured precipitation and stream discharge data, especially during high flow seasons causing substantial errors of discharge values at extreme low and high flow seasons affected by recording and rating errors, respectively (Rossi et al., 2009). Analysis of 7 years (1990-1996) of model simulation outputs indicated that about 85% of total annual stream discharge of the basin was contributed by rain (Figure 5) highlighting the importance of rainfall to hydrologic processes in the basin. Based on different source water contribution values estimated from total water input of the basin, Snowmelt and groundwater had lower contribution values of 9 and 6%, respectively. These estimations will be affected by potential climate variability influencing hydrologic processes and future water availability of the basin (Fontaine et al., 2001; Wu et al., 2012). However, we estimated a loss of 26 mm (about 4% of total inputs) of water from the basin, potentially indicating water flux into deep groundwater reservoirs, not accounted by the SWAT model (Arnold et al., 1993).

3.3.

Model sensitivity and uncertainty analysis The most sensitive input parameters regarding stream discharge simulation were

identified on the basis of global sensitivity analysis, and presented in Table 5. The most 16 sensitive parameters for this study were SMTMP (snowmelt base temperature), ALPHA_BNK (baseflow alpha factor for bank storage), SFTMP (snowfall temperature), CN2 (curve number for moisture condition II), SMFMX (maximum snowmelt rate), GW_DELAY (groundwater delay

time), CH_N2 (Manning’s n value for the main channel), TIMP (snowpack temperature lag factor), OV_N (Manning’s n value for overland flow), RCHRG_DP (deep aquifer percolation fraction), REVAPMN (threshold depth of water in the shallow aquifer for revap), ESCO (soil evaporation compensation factor), EPCO (plant uptake compensation factor), ALPHA_BF (baseflow alpha factor), CH_K2 (hydraulic conductivity in main channel), and GWQMN (threshold depth of water in shallow aquifer required for return flow). Among these, the SMTMP was found as the most sensitive parameter to simulate stream discharge of the basin that was supported by similar higher sensitivity of the parameter studied in northwestern Minnesota watershed (Wang and Melesse, 2005). SMTMP is sensitive since it is the indicator of starting and ending of melt, thus affecting the availability of snow for melting on a specific day. As a result, model simulated stream discharge, mainly the peaks, are substantially influenced by variations in SMTMP. However, the SMFMN (minimum snowmelt rate), SURLAG (surface runoff lag time), and GW_REVAP (groundwater revap coefficient) were analyzed as the least sensitive parameters to simulate stream discharge of the basin. Due to higher seasonal snowfall in the basin, SMTMP and SFTMP play an important role for accurate estimation of rate and volume of meltwater released (Fontaine et al., 2002) that contributes to stream discharge; however, it depends on snowpack conditions (Martinec and Rango, 1986). The higher ALPHA_BNK value showed a slow recession of sub-surface water reservoirs (Arnold et al., 2011), potentially indicating the importance of shallow groundwater sources for annual water budgets of the basin. The uncertainty analysis performed for one year model predicted stream discharge values from calibration period showed the significant model over and under estimations, primarily during summer season but showed only overestimation vales during spring season (Figure 6) that might be attributed to measurement errors occurred during high flow

seasons (Rossi et al., 2009). However, we analyzed better model performance during low-flow winter months.

3.4.

Modeling the hydrologic response to potential land use change The dominant land use types of the BSR basin are agriculture, grassland, and urban

(Table 1) which accounts for about 95% of the entire area. There was a dramatic land use change between 2006 and 2013, primarily agricultural transformation of grassland area in the basin (NASS, http://nassgeodata.gmu.edu/CropScape/). For example, the maximum change of +9% was analyzed for the corn cultivation within the same time period. Therefore, to evaluate such effects of land use change on some hydrologic processes including surface runoff and water yield, the calibrated SWAT model was simulated in three potential land use scenarios for the period of 1990 to 1996. Projected changes of these hydrologic parameters are presented in Figure 7. Simulation outputs indicated that the mean annual surface runoff was 2% higher while increased 2% corn cultivation in the basin but, it was 3% higher when the corn cultivation increased to 5%. The maximum change of +4% was analyzed when the corn cultivation expanded to 10%, potentially due to lower soil water interception of the cornfield compared to the grassland area. This is because, the less dense crops with shallow-root system generally have less storage capacity than grasses which ultimately increases the surface runoff to the basin (Zhang et al., 2001). Despite minimal, corresponding to higher surface runoff, the total water yield of the basin was also higher in all land use scenarios with maximum increase of about 0.5% in the Scenario-III. Therefore, these trends of rapid agricultural transformation with application of potential higher amount of chemical fertilizers deteriorate the quality of water resources,

ultimately increasing the maintenance cost for drinking water supply in the region (Palaniappan et al., 2010; Koschak, 2014).

3.5.

Modeling the hydrologic response to future climate variability Future changes of temperature and precipitation derived for B1, A1B, and A2 scenarios

for the BSR basin are presented in Table 6. We found that the mean annual temperature of the basin was higher than current value with the ranges between 2.9 and 5.7oC. Monthly temperatures varied substantially with the ranges from 1.5 to 8.2 oC. These estimations demonstrate overall warming of the BSR basin with higher temperatures in the SRES-A2 scenario. However, the mean annual precipitation of the basin decreased in all scenarios with maximum change of -5% in the SRES-A2. Substantial change of -25 to +61% was estimated in monthly precipitation. Therefore, these potential climate variations might induce for risks of drought and water quality degradation in the basin (Wu et al., 2012). In order to study some hydrologic processes including precipitation, stream discharge, total water yield, groundwater contribution to stream discharge, water percolation, evapotranspiration, snowfall, and snowmelt in the BSR watershed, the calibrated model was simulated using potential changes of temperature and precipitation derived for the B1, A1B, and A2 scenarios, and is presented in Figure 8. Due to potential higher temperature, precipitation of the BSR basin decreased in all scenarios with maximum change of -11% in the SRES-A2 scenario. Despite decrease in precipitation, stream discharge increased to 72, 47, and 13% for the SRES-B1, A1B, and A2, respectively that corresponded with similar increases of 67, 45, and 8%, respectively for total water yield of the basin. Potential higher discharge was primarily gained by groundwater sources with maximum increase in discharge contribution of 152% in the SRES-B1.

Groundwater contribution to total stream discharge is simulated by routing a shallow aquifer storage component to the stream (Arnold et al., 1993). For the BSR basin, groundwater discharge from the aquifers forms baseflow component that is important to sustain discharge of the BSR, mainly during low flow seasons when there is no significant amount of surface runoff (Ellis et al., 1969). Therefore, our estimation of potential higher groundwater contribution to stream discharge might be indicating the importance of sub-surface water resources (Nelson and Siegel, 1983) that may influence future water availability of the BSR basin. However, detailed study of local groundwater reservoirs is required for accurate estimation of groundwater contribution to discharge of the BSR. Model simulation outputs showed substantial change in evapotranspiration rate for all scenarios with the values of -20, -14, and -10% for the SRES-B1, A1B, and A2, respectively. The lower annual ET rates are mainly caused by the decrease in future precipitation (Kim et al., 2013) which might increase soil water content in the basin. These potential lower ET rates corresponded with higher soil water percolation of 133, 113, and 47% for the same scenarios, respectively. Percolation is the movement of water through each soil layer where the water content exceeds the corresponding field capacity (Neitsch et al., 2011). Due to projected higher soil water content and reduced evapotranspiration rate in the basin, we estimated substantial increase in soil water percolation in all climate change scenarios (Pennell, 2003). Groundwater reservoirs in the BSR basin provide a substantial proportion (about one-third) of the water supplied for municipal, rural water, irrigation, and various other uses (Nelson and Siegel, 1983). These reservoirs are shallow making water level very close to the surface that lead for quick recharge of the aquifers from saturated soil increasing the stream discharge due to higher transmissivity of porous outwash in the basin. Increased air temperature lead to decrease in the

ratio of snow to rain, that decreases snow water equivalent (water stored in snowpack) reducing snowmelt contribution to runoff. Due to potential higher temperature and reduced precipitation distribution in the basin, both snowfall and snowmelt decreased in all climate change scenarios (Figure 8g and 8h). Snowfall decreased by an average of 45% in SRES-B1 scenario and 59% in SRES-A2 scenario. Similar snowmelt changes of -36 and -52% were estimated in the same scenarios, respectively. Therefore, these estimations of potential decrease in snowfall coupled with continued climate change can have long term effects on stream discharge reducing future water availability of the basin, mainly during dry seasons (Barnett et al., 2005). On seasonal basis, we analyzed a mean annual precipitation contribution of 44% during summer season in the baseline scenario (Figure 9). This was higher than the values of 29, 21, and 5% contributed for the spring, fall, and winter seasons, respectively. Corresponding to precipitation, mean annual stream discharge contribution of 37, 26, 24, and 14% were estimated for the summer, spring, fall, and winter seasons, respectively. This is because, seasonal variation in stream discharge is directly related to the seasonal precipitation (Hansen, 1986). The precipitation contribution increased in climate change scenarios with maximum changes of +9 and +33% for winter and spring seasons, that corresponded to higher stream discharge values of 20 and 29% for these seasons, respectively. However, we found a substantial decrease of summer precipitation contribution to 36% in future scenario that resembled with reduced discharge contribution of 28% in the same season. Similarly, the snowfall contributions of 45, 20, and 35% were assessed for the winter, fall, and spring seasons, respectively. However, maximum snowmelt contribution (77%) was estimated for the spring season highlighting the importance of seasonal snowmelt in the basin. These assessments showed a large potential

seasonal variability on key hydrologic components that may affect future water availability to use for drinking water supply and agricultural production of the region.

3.6.

Modeling the hydrologic response to potential land use change combined with climate

variability The combined effects of potential land use change and climatic variability on hydrologic processes including stream discharge and total water yield of the BSR basin are presented in Figure 10. The discharge of the BSR was estimated higher in all combined scenarios with maximum increase of 77% in Scenario-III that changed by +72% in the SRES-B1. Due to higher discharge of the BSR, the amount of total water yield was also intensified in combined scenarios with maximum increase of 71% in Scenario-III that was changed by +67% in the SRES-B1. This is because, regarding the coupled effects of potential land use change and climate variability, the decrease in magnitude of hydrologic parameters caused by land use change will be offset by climate change and the magnitude will be enhanced in climate change scenarios if higher in land use scenario (Ma et al., 2009). Finally, potential land use change coupled with climate variability intensified both stream discharge and total water yield that might potentially enhance the risk of flood and sedimentation affecting drinking water supply and irrigation system of this large agriculture dominated watershed. Therefore, these assessments might be crucial for better future land use planning and water resources management of this region.

4.

Summary and conclusions (i) The SWAT model was applied to large agriculture dominated BSR watershed to

assess the effects of potential land use change and climate variability on hydrologic processes regarding the water quantity issues; (ii) Our results indicated that SWAT proved to be a powerful tool to simulate the effect of environmental change on hydrologic processes of the basin highlighting the importance of snow and groundwater related parameters to influence the accuracy of simulated discharge vs measured values; however, SMTMP and ALPHA_BNK were evaluated as the most important factors; (iii) We estimated higher surface runoff while expanding corn cultivation in the basin; (iv) Despite decrease in future precipitation distribution in the basin, discharge of the BSR was estimated higher that attributed to potential higher groundwater contribution in future scenarios; (v) However, to explore the pronounced role of future groundwater resources in the basin, it necessitates for further investigation on groundwater reservoirs and their contribution to discharge of the river; and (vi) Finally, understanding these effects of potential land use change and climate variability might be crucial for better future land use planning and water resources management for sustainable agricultural production of large watersheds.

Acknowledgements This study was part of a project supported by the United States Department of Agriculture-NIFA (Award Number 2014-51130-22593) and the project entitled ‘‘Integrated plan for drought preparedness and mitigation, and water conservation at the watershed scale.’’

References Abbaspour, K.C., 2012. SWAT-CUP 2012: SWAT calibration and uncertainty programs - a user manual. Eawag: Swiss Federal Institute Science and Technology. Abbaspour, K.C., Vejdani, M., Haghighat, S., 2007. SWAT-CUP calibration and uncertainty programs for SWAT. Modsim 2007: International Congress on Modelling and Simulation: Land, Water and Environmental Management: Integrated Systems for Sustainability, Christchurch, New Zealand. Arnold, J.G., Allen, P.M., Bernhardt, G., 1993. A comprehensive surface-groundwater flow model. Journal of Hydrology 142, 47 – 69. Arnold, J.G., Kiniry, J.R., Srinivasan, R., Williams, J.R., Haney, E.B., Neitsch, S.L., 2011. Soil and Water Assessment Tool Input/Output File Documentation. Texas Water Resources Institute. Arnold, J.G., Srinivasan, R., Muttiah, R.S., Williams, J.R., 1998. Large area hydrologic modeling and assessment Part I: Model development. Journal of the American Water Resources Association 34, 73-89. Barnett, T.P., Adam, J.C., Lettenmaier, D.P., 2005. Potential impacts of a warming climate on water availability in snow-dominated regions. Nature 438, 303-309. Boryan, C., Yang, Z., Mueller, R., Craig, M., 2011. Monitoring US agriculture: the US department of agriculture, national agricultural statistics service, cropland data layer program. Geocarto International 26, 341-358. Brath, A., Montanari, A., Moretti, G., 2006. Assessing the effect on flood frequency of land use change via hydrological simulation (with uncertainty). Journal of Hydrology 324, 141153. Brekke, L., Thrasher, B.L., Maurer, E.P., Pruitt, T., 2013. Downscaled CMIP3 and CMIP5 Climate Projections. Release of Downscaled CMIP5 Climate Projections, Comparison with Preceding Information, and Summary of User Needs. Chien, H., Yeh, P.J.F., Knouft, J.H., 2013. Modeling the potential impacts of climate change on streamflow in agricultural watersheds of the Midwestern United States. Journal of Hydrology 491, 73-88. Costa, M.H., Botta, A., Cardille, J.A., 2003. Effects of large-scale changes in land cover on the discharge of the Tocantins River, Southeastern Amazonia. Journal of Hydrology 283, 206217. Ellis, M.J., Adolphson, D.G., West, R.E., 1969. Hydrology of a part of the Big Sioux drainage basin, eastern South Dakota. US Geological Survey, 1969. Fohrer, N., Haverkamp, S., Frede, H.G., 2005. Assessment of the effects of land use patterns on hydrologic landscape functions: development of sustainable land use concepts for low mountain range areas. Hydrological Processes 19, 659-672. Foley, J.A., DeFries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., Chapin, S.T., Coe, M.T., Daily, G.C., Gibbs, H.K., Helkowski, J.H., Holloway, T., Howard, E.A., Kucharik, C.J., Monfreda, C., Patz, J.A., Prentice, I.C., Ramankutty, N., Snyder, P.K., 2005. Global Consequences of Land Use. Science 309, 570-574. Fontaine, T.A., Cruickshank, T.S., Arnold, J.G., Hotchkiss, R.H., 2002. Development of a snowfall-snowmelt routine for mountainous terrain for the soil water assessment tool (SWAT). Journal of Hydrology 262, 209-223.

Fontaine, T.A., Klassen, J.F., Cruickshank, T.S., Hotchkiss, R.H., 2001. Hydrological response to climate change in the Black Hills of South Dakota, USA. Hydrological Sciences Journal 46, 27-40. Germer, S., Neill, C., Vetter, T., Chaves, J., Krusche, A.V., Elsenbeer, H., 2009. Implications of long-term land-use change for the hydrology and solute budgets of small catchments in Amazonia. Journal of Hydrology 364, 349-363. Groisman, P.Y., Knight, R.W., Karl, T.R., 2001. Heavy Precipitation and High Streamflow in the Contiguous United States: Trends in the Twentieth Century. Bull. Amer. Meteor. Soc. 82, 219–46. Gupta, H., Sorooshian, S., and Yapo, P., 1999. Status of automatic calibration for hydrologic models: comparison with multilevel expert calibration. Journal of Hydrologic Engineering, 4, 135-143. Hansen, D.S., 1986. Water resources of lake and Moody counties, South Dakota. U.S. Geological Survey. Haverkamp, S., Srinivasan, R., Frede, H.G., Santhi, C., 2002. Subwatershed spatial analysis tool: Discretization of a distributed hydrologic model by statistical criteria. Journal of American Water Resources Association 38, 1723-1733. He, H., Zhou, J., Zhang, W., 2008. Modelling the impacts of environmental changes on hydrological regimes in the Hei River Watershed, China. Global and Planetary Change 61, 175-193. Huxman, T.E., Wilcox, B.P., Breshears, D.D., Scott, R.L., Snyder, K.A., Small, E.E., Hultine, K., Pockman, W.T., Jackson, R.B., 2005. Ecohydrological implications of woody plant encroachment. Ecological Society of America 86, 308-319. Keane, R.E., Holsinger, L.M., Parsons, R.A., Gray, K., 2008. Climate change effects on historical range and variability of two large landscapes in western Montana, USA. Forest Ecology and Management 254, 375-389. Kerth, R., Vinyard, S., 2012. Wasting Our Waterways 2012. Toxic Industrial Pollution and the Unfulfilled Promise of the Clean Water Act. Environment America Research and Policy Center. Kim, H.K., Parajuli, P.B., Filip To, S.D., 2013. Assessing impacts of bioenergy crops and climate change on hydrometeorology in the Yazoo River Basin, Mississippi. Agricultural and Forest Meteorology 169, 61-73. Koschak, M., 2014. Lower Big Sioux River Watershed Biotic Stressor Identification Report. Minnesota Pollution Control Agency. Kundzewicz, Z.W., Mata, L.J., Arnell, N.W., Döll, P., Kabat, P., Jiménez, B., Miller, K.A., Oki, T., Sen, Z., Shiklomanov, I.A., 2007. Freshwater resources and their management. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Parry, M.L., Canziani, O.F., Palutikof, J.P., Van der Linden, P.J., Hanson, C.E., (Eds.) Cambridge University Press, Cambridge, UK, 173-210. Lin, W., Zhang, L., Du, D., Yang, L., Lin, H., Zhang, Y., Li, J., 2009. Quantification of land use/land cover change in Pearl River Delta and its impact on regional climate in summer using numerical modeling. Regional Environmental Change 9, 75-82. Lin, Y.P., Hong, N.M., Wu, P.J., Lin C.J., 2007. Modeling and assessing land-use and hydrological processes to future land-use and climate change scenarios in watershed landuse planning. Environment and Geology 53, 623-634.

Ma, X., Xu, J., Luo, Y., Aggarwal, S.P., Li, J., 2009. Response of hydrological processes to landcover and climate changes in Kejie watershed, south-west China. Hydrological Processes 23, 1179-1191. Manter, D.K., Reeser, P.W., Stone, J.K., 2005. A Climate-Based Model for Predicting Geographic Variation in Swiss Needle Cast Severity in the Oregon Coast Range. Ecology and Epidemiology 95, 1256-1265. Marshall, E.P., Shortle, J., 2005. Urban development impacts on ecosystems. Chapter 7 in Goetz S, Shortle J, Bergstrom J (eds.), Land Use Problems and Conflicts: Causes, Consequences and Solutions. 61-72. Martinec, J., Rango, A., 1986. Parameter values for snowmelt runoff models. Journal of Hydrology 84, 197-219. Maurer, E.P., Hidalgo, H.G., Das, T., Dettinger, M.D., Cayan, D.R., 2010. “The utility of daily large-scale climate data in the assessment of climate change impacts on daily streamflow in California”. Hydrology and Earth System Sciences 14, 1125-1138. Monteith, J.L., 1965. Evaporation and environment. Symposia of the Society for Experimental Biology 19, 205-234. Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Binger, R.L., Harmel, R.D., Veith, T.L., 2007. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. American Society of Agricultural and Biological Engineers 50, 885-900. Muleta, M.K., Nicklow, J.W., 2005. Sensitivity and uncertainty analysis coupled with automatic calibration for a distributed watershed model. Journal of Hydrology 306, 127-145. MunozVillers, L.E., McDonnell, J.J., 2013. Land use change effects on runoff generation in a humid tropical montane cloud forest region. Hydrology and Earth System Sciences 17, 3543-3560. Nash, J.E., and Sutcliffe, J.V., 1970. River flow forecasting through conceptual models part I-A discussion of principles. Journal of Hydrology 10, 282-290. Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R., 2011. Soil and Water Assessment Tool: Theoretical Documentation, version 2009. Texas Water Resources Institute. Nelson, D., Siegel, J., 1983. Big Sioux aquifer water quality study. East Dakota Conservancy Sub-District 1983. Neupane, R.P., White, J.D., Alexander, S.E., 2015. Projected hydrologic changes in monsoondominated Himalaya Mountain basins with changing climate and deforestation. Journal of Hydrology 525, 216-230. Neupane, R.P., Yao, J., White, J.D., 2014. Estimating the effects of climate change on the intensification of monsoonal-driven stream discharge in a Himalayan watershed. Hydrological Processes 28, 6236-6250. NRCS. 2009. Soil Survey Geographic (SSURGO) Database. USDA Natural Resources Conservation Service. Available at: http://soildatamart.nrcs.usda.gov. Accessed 30 June 2009. Oubeidillah, A.A., Kao, S.C., Ashfaq, M., Naz, B.S., Tootle, G., 2014. A large-scale, highresolution hydrological model parameter data set for climate change impact assessment for the conterminous US. Hydrology and Earth System Sciences 18, 67-84. Palaniappan, M., Gleick, P.H., Allen, L., Cohen, M.J., Smith, J.C., Smith, C., 2010. Clearing the Waters. A focus on water quality solution. UNEP. The Pacific Institute, Nairobi, Kenya. Pennell, K.D., 2003. Soils, Water Percolation Through. Georgia Institute of Technology, Atlanta, Georgia, U.S.A.

Pirner, S.M., 2004. Phase I, Watershed Assessment Final Report and TMDL. South Dakota Department of Environment and Natural Resources. Ritchie, J.T., 1972. Model for predicting evaporation from a row crop with incomplete cover. Water Resources Research 8, 1204-1213. Rossi, C.G., Srinivasan, R., Jirayoot, K., Duc, T.L., Souvannabouth, P., Binh, N., Gassman, P.W., 2009. Hydrologic evaluation of the lower Mekong river basin with the Soil and Water Assessment Tool model. International Agricultural Engineering Journal 18, 1-13. Rostamian, R., Jaleh, A., Afyuni, M., Mousavi, S.F., Heidarpour, M., Jalalian, A., Abbaspour, K.C., 2008. Application of a SWAT model for estimating runoff and sediment in two mountainous basins in central Iran. Hydrological Sciences Journal 53, 977-988. Scheffler, R., Neill, C., Krusche, A.V., Elsenbeer, H., 2011. Soil hydraulic response to land-use change associated with the recent soybean expansion at the Amazon agricultural frontier. Agriculture, Ecosystems and Environment 144, 281-289. SDSU, 2003. South Dakota State University Climate and Weather Web Page. http://climate.sdstate.edu/climate_site/climate_page.htm. Simonovic, S.P., Li, L., 2004. Sensitivity of the Red River Basin Flood Protection System to Climate Variability and Change. Water Resources Management 18, 89-110. Sloan, P.G., Moore, I.D., 1984. Modeling Subsurface Stormflow on Steeply Sloping Forested Watersheds. Water Resources Research 20, 1815-1822. Stonestrom, D.A., Scanlon, B.R., Zhang, L., 2009. Introduction to special section on Impacts of Land Use Change on Water Resources. Water Resources Research 45, W00A00, doi:10.1029/2009WR007937. Thornton, P.E., Running, S.W., White, M.A., 1997. Generating surfaces of daily meteorology variables over large regions of complex terrain. Journal of Hydrology 190, 214-251. Thornton, P.E., Thornton, M.M., Mayer, B.W., Wilhelmi, N., Wei, Y., Cook, R.B., 2012. Daymet: Daily surface weather on a 1 km grid for North America,1980 - 2008. Acquired online (http://daymet.ornl.gov/) on [insert acquisition date here: DD/MM/YYYY] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. doi:10.3334/ORNLDAAC/Daymet_V2. USDA, 1986. Urban hydrology for small watersheds. Technical Release 55. Engineering Division, Soil Conservation Service, US Department of Agriculture. USDA, 2011. Soil Survey Geographic (SSURGO) Database, . United States Department of Agriculture, Natural Resources Conservation Service. Washington, DC. Wang, G., Zhang, Y., Liu, G., Chen, L., 2006. Impact of land-use change on hydrological processes in the Maying River basin, China. Science in China Series D: Earth Sciences 49, 1098-1110. Wang, X., Melesse, A.M., 2005. Evaluation of the SWAT model’s snowmelt hydrology in a northwestern Minnesota watershed. American Society of Agricultural Engineers 48, 1-18. White, M.A., Diffenbaugh, N.S., Jones, G.V., Pal, J.S., Giorgi, F., 2006. Extreme heat reduces and shifts United States premium wine production in the 21 st century. Proceeding of the National Academy of Sciences 103, 11217-11222. Wright, C.K., Wimberly, M.C., 2013. Recent land use change in the Western Corn Belt threatens grasslands and wetlands. Proceedings of the National Academy of Sciences 110, 41344139.

Wu, Y., Liu, S., Gallant, A.L., 2012. Predicting impacts of increased CO2 and climate change on the water cycle and water quality in the semiarid James River Basin of the Midwestern USA. Science of the Total Environment 430, 150-160. Xu, C.Y., 1999. Climate Change and Hydrologic Models: A Review of Existing Gaps and Recent Research Developments. Water Resources Management 13, 369-382. Xu, C.Y., 2000. Modelling the Effects of Climate Change on Water Resources in Central Sweden. Water Resources Management 14, 177-189. Yan, B., Fang, N.F., Zhang, P.C., Shi, Z.H., 2013. Impacts of land use change on watershed streamflow and sediment yield: An assessment using hydrologic modelling and partial least squares regression. Journal of Hydrology 484, 26-37. Yang, J., Reichert, P., Abbaspour, K.C., Xia, J., Yang, H., 2008. Comparing uncertainty analysis techniques for a SWAT application to the Chaohe Basin in China. Journal of Hydrology 358, 1-23. Zhang, G.H., Nearing, M.A., Liu, B.Y., 2005. Potential Effects of Climate Change on Rainfall Erosivity in the Yellow River Basin of China. American Society of Agricultural Engineers 48, 511-517. Zhang, L., Dawes, W.R., Walker, G.R., 2001. Response of mean annual evapotranspiration to vegetation changes at catchment scale. Water Resources Research 37, 701-708. Zhang, X., Srinivasan, R., Hao, F., 2007. Predicting hydrologic response to climate change in the Luohe River Basin using the SWAT model. American Society of Agricultural and Biological Engineers 50, 901-910. Zierl, B., Bugmann, H., 2005. Global change impacts on hydrological processes in Alpine catchments. Water Resources Research 41, W02028, doi:10.1029/2004WR003447.

Figure captions Figure 1. The Big Sioux River watershed located in north-central part of the United States is shown with the hydro-meteorological stations used in this study. Figure 2. Potential land use scenarios derived for the Big Sioux River basin: (a) existing land use class used as baseline for this study, (b) after increasing 2% corn cultivation as Scenario-I, (c) after increasing 5% corn cultivation as Scenario-II, and (d) after increasing 10% corn cultivation as Scenario-III. Figure 3. Hydrographs to show (a) mean monthly precipitation and stream discharge, and (b) seasonal contribution of precipitation and stream discharge in the Big Sioux River basin (The precipitation data have been averaged from the period of 1980 to 2013 and the stream discharge data have been averaged for the period of 1970 to 2012 (Note- Winter: December-February, Spring: March-May, Summer: June-August, and Fall: September-November). Figure 4. Hydrographs obtained during model calibration and confirmation periods: (a) daily, and (b) monthly. Figure 5. Mean yearly stream discharge contribution of the BSR basin by different source waters estimated from model simulation outputs (Note: the data have been averaged from the year of 1990 to 1996). Figure 6. Standardized residual error (z-scores) for SWAT model simulation outputs of stream discharge (m3/s) vs measured values for the Big Sioux River watershed. Figure 7. Mean annual changes of (a) surface runoff, (b) water yield, (c) total nitrogen, and (d) total phosphorus in the BSR basin from the baseline in different scenarios. Figure 8. Mean annual changes of (a) precipitation, (b) stream discharge, (c) total water yield, (d) groundwater contribution to stream discharge, (e) water percolation, (f) evapotranspiration, (g) snowfall, and (h) snowmelt in the Big Sioux River basin from the baseline in different climate change scenarios. Figure 9. Potential seasonal changes of (a) precipitation, (b) stream discharge, (c) snowfall, and (d) snowmelt in the Big Sioux River basin for different climate change scenarios (Winter: December-February, Spring: March-May, Summer, June-August, and Fall: SeptemberNovember). Figure 10. Mean annual changes of (a) stream discharge, and (b) total water yield in combined scenarios of potential land use change and climate variability in the Big Sioux River basin.

Figure 1

Figure 2

100

Precipitation (mm)

140

Precipitation Discharge

(a)

100

80

80 60 60

40

40

20

20

0

0 Jan

Feb Mar Apr May Jun

Jul

Aug Sep

Oct Nov Dec

Month 50

(b) % contribution

40 30

20 10

0 Winter

Figure 3

120

Spring

Summer

Fall

Stream discharge (m3/s)

120

Jan-90 Jun-90 Nov-90 Apr-91 Sep-91 Feb-92 Jul-92 Dec-92 May-93 Oct-93 Mar-94 Aug-94 Jan-95 Jun-95 Nov-95 Apr-96 Sep-96 Feb-97 Jul-97 Dec-97 May-98 Oct-98 Mar-99 Aug-99 Jan-00 Jun-00 Nov-00 Apr-01 Sep-01

Stream discharge (m3/s)

Stream discharge (m3/s) 1200

1000

Figure 4 Simulated Measured

800

600

400

200

0

700

600

Calibration Confirmation

(a)

Calibration Confirmation

(b)

500

400

300

200

100

0

Time

100

% contribution

80

60 40 20 0 Rain

Figure 5

Snow

Groundwater

60

Standardized residual, z

40

95 % c.i.

20

0 -20

Summer

Spring

Fall

Winter

-40 -60 -80 -100 0

50

100

150

200

Julian day

Figure 6

250

300

350

400

0.6

5 4 3 2

1

0.5 0.4

0.3 0.2 0.1

0

0 Scenario-I

Figure 7

(b) % change from baseline

% change from baseline

(a)

Scenario-II

Scenario-III

Scenario-I

Scenario-II

Scenario-III

(a)

-4 -6 -8

-10

80

% change from baseline

% change from baseline

0 -2

-12

% change from baseline

20

SRES-B1

(c)

60 40 20

SRES-A1B

SRES-A2

160

(d)

120 80 40 0

0 SRES-A1B

SRES-B1

SRES-A2

150

(e)

120 90 60 30

SRES-A1B

SRES-A2

0

% change from baseline

SRES-B1

% change from baseline

40

SRES-A2

% change from baseline

SRES-A1B

80

0

(f)

-5 -10 -15 -20 -25

SRES-A1B

SRES-A2

0

SRES-B1

(g)

-20 -40 -60 -80

SRES-A1B

SRES-A2

0

% change from baseline

SRES-B1

% change from baseline

60

0 SRES-B1

(h)

-10 -20 -30 -40 -50 -60

SRES-B1

Figure 8

(b)

SRES-A1B

SRES-A2

SRES-B1

SRES-A1B

SRES-A2

50

Winter

Spring

Summer

Fall

40

(b)

(a) % contribution

% contribution

40 30

20 10

20 10

0

0 Baseline

SRES-B1

SRES-A1B

Baseline

SRES-A2

120

SRES-B1

SRES-A1B

SRES-A2

100

(c)

100

(d) 80

% contribution

% contribution

30

80 60 40

60 40

20

20

0

0 Baseline

Figure 9

SRES-B1

SRES-A1B

SRES-A2

Baseline

SRES-B1

SRES-A1B

SRES-A2

% change from baseline

100

(a) 80 60 40 20

% change from baseline

0 80

(b) 60 40 20 0

Scenario

Figure 10

Tables

Table 1. Percent cover of land use class in the Big Sioux River watershed derived from analysis of National Agricultural Statistics Service (NASS) data for the period of 2007. Landuse class Percent cover Corn 36 Soybean 24 Wheat 1 Grassland 27 Water 3 Wetland 1 Urban 8

Table 2. The hydro-meteorological stations with their spatial locations used for hydrologic simulations of the Big Sioux River watershed that provided climate forcing data. Description Type Latitude (o) Longitude (o) Elevation (m) Station 1 Meteorology 43.00 -96.49 367 Station 2 Meteorology 43.58 -96.75 432 Station 3 Meteorology 44.33 -96.77 493 Station 4 Meteorology 44.55 -97.47 545 Station 5 Meteorology 44.90 -97.15 531 Akron Hydrology 42.84 -96.56 348

Table 3. The agricultural land management operations applied in the Big Sioux River watershed including the number of heat unit required to bring a plant to maturity. Date Operation Plant Heat Units Fraction of PHU (PHU) accumulated (PHU = 1,124) April 15 Fertilizer application Urea (46-0-0) DAP (18-46-0) MAP (11-52-0) May 5 Planting Corn (PHU = 1,124) 0 0.00 October 5 Harvest and Kill 1,124 1.20

Table 4. Model performance statistics for pre-calibration, calibration, and confirmation simulations in both daily and monthly time periods. Statistics Pre-calibration Calibration Confirmation Daily Monthly Daily Monthly Daily Monthly NSE -9.95 -8.15 0.63 0.70 0.60 0.70 PBIAS -275 -275 -5 -3 -14 -14 RMSE 371 282 54 53 59 48 r2 0.16 0.38 0.68 0.75 0.67 0.86

Table 5. Sensitivity results with the ranking of key SWAT parameters for stream discharge in the Big Sioux River watershed including the range of parameter values adopted from Muleta and Nicklow (2005), Rostamian et al. (2008), and Neupane et al. (2014, 2015).

Table 6. Projected scenarios for the Big Sioux River basin to use for future hydrologic assessment: (a) annual and monthly temperature change (oC), and (b) annual and monthly cumulative precipitation change (%).

44

Research highlights 

SWAT model was calibrated for a large agriculture dominated watershed.



The model simulated higher surface runoff with increase in corn cultivation.



Land use change and climate variability enhanced for higher annual discharge.

46