Interrelationships among water cycle fluxes and stores in a semi-arid urban environment

This thesis research was performed to quantify and analyze the interrelationships among water cycle fluxes and stores in a semi-arid urban environment. In September 2007, a hydrologic observation network was installed in a residential area of the Salt Lake City metropolitan area to collect continuous observations of precipitation, dry and wet weather runoff, and soil moisture. In addition, companion studies in the same catchment were simultaneously collecting evapotranspiration and water use data. For the time period studied, precipitation was normal; conversely, the ratios of runoff to rainfall (i.e., runoff coefficient) for the storm events observed were much smaller than expected (0.04 for spring and fall months and 0.006 for summer months), with significant variability noted. Soil moisture was found to have at least a 20% variation between irrigated and nonirrigated areas, which can be expected; however, the increase of nonirrigated soil moisture in response to rainfall was surprisingly small. Dry weather runoff was found to be nearly 8% of the total outflow volume (wet weather plus dry weather), and outdoor water use was shown to influence soil moisture, which in turn was observed to influence the evapotranspiration rate. The continuous observations quantified several uncertain fluxes and stores in semi-arid cities and identified relationships among stores and fluxes useful for developing predictive models of the interactions among urban hydrologic fluxes, stores, and climate.

INTERRELATIONSHIPS AMONG WATER CYCLE FLUXES AND STORES IN A SEMI-ARID URBAN ENVIRONMENT by Nicholas Walter Augustus A thesis submitted to the faculty of the University of Utah in partial fulfillment of the requirements for the degree of Master of Science Department of Civil and Environmental Engineering The University of Utah December 2008 Copyright © Nicholas Walter Augustus 2008 All Rights Reserved THE U N I V E R S I T Y OF UTAH G R A D U A T E SCHOOL SUPERVISORY COMMITTEE APPROVAL of a thesis submitted by Nicholas Walter Augustus This thesis has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. Chair: Steven J. Bdrian Christine Pomeroy Brian J/McPherson UNIVERSITY GRADUATE SCHOOL IO-ll-).a~ Christinepomeroy (5 THE U N I V E R S I T Y OF UTAH G R A D U A T E SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read the thesis of Nicholas Walter Augustus in its final form and have found that (1) its format, citations, and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the supervisory committee and is ready for submission to The Graduate School. Date Steven J. Burian Chair: Supervisory Committee Approved for the Major Department Paul Tikalsky Chair/Dean Approved for the Graduate Council J£c.J4 rQ- c ,^ David S. Chapman Dean of The Graduate School UNIVERSITY UT AH GRADU A TE SCHOOL APPROVAL }o -10-09 ead~ C/£ ChairlDean ABSTRACT This thesis research was performed to quantify and analyze the interrelationships among water cycle fluxes and stores in a semi-arid urban environment. In September 2007, a hydrologic observation network was installed in a residential area of the Salt Lake City metropolitan area to collect continuous observations of precipitation, dry and wet weather runoff, and soil moisture. In addition, companion studies in the same catchment were simultaneously collecting evapotranspiration and water use data. For the time period studied, precipitation was normal; conversely, the ratios of runoff to rainfall (i.e., runoff coefficient) for the storm events observed were much smaller than expected (0.04 for spring and fall months and 0.006 for summer months), with significant variability noted. Soil moisture was found to have at least a 20% variation between irrigated and nonirrigated areas, which can be expected; however, the increase of non-irrigated soil moisture in response to rainfall was surprisingly small. Dry weather runoff was found to be nearly 8% of the total outflow volume (wet weather plus dry weather), and outdoor water use was shown to influence soil moisture, which in turn was observed to influence the evapotranspiration rate. The continuous observations quantified several uncertain fluxes and stores in semi-arid cities and identified relationships among stores and fluxes useful for developing predictive models of the interactions among urban hydrologic fluxes, stores, and climate. non­irrigated hydrologic fluxes, stores, and climate. TABLE OF CONTENTS ABSTRACT iv LIST OF TABLES vii LIST OF FIGURES viii ACKNOWLEDGEMENTS xi 1. INTRODUCTION 1 1.1 Conservation 6 1.3 Energy Budget 8 2. LITERATURE REVIEW 11 3. OBSERVATION NETWORK AND METHODS 15 3.1 General 15 3.2 Murray Urban Water Cycle Observation Network 15 3.3 Data Collection and Analysis 33 4. RESULTS AND DISCUSSION 34 4.1 Precipitation 34 4.2 Flow Rates 37 4.3 Soil Moisture 40 4.4 Evapotranspiration 44 4.5 Water Use 44 4.6 Processes and Relationships 45 5. SUMMARY AND CONCLUSIONS 57 APPENDICES A. TEMPERATURE 60 ....................................................................................................................... .............................. .. .............. ..... ... ... .......................................... ......... ......................................................................................................... ............................................................................................... CHAPTERS ......................................................................................................... ............................................................... .. ......................... .. ............ 1.2 Climate Variability ... ................... ...................................................... ... ................ 7 1.3 Energy Budget ...... ......................................................... ......... ...... .. ..................... 8 2. LITERATURE REVIEW ... ...... ....................................................................... ............ 11 3. OBSERVATION NETWORK AND METHODS ...................................................... 15 3.1 General ................................................................................................................ 15 3.2 Murray Urban Water Cycle Observation Network ............................................. 15 3.3 Data Collection and Analysis .............................................................................. 33 4. RESULTS AND DISCUSSION .................................................................................. 34 4.1 Precipitation ...................................................... .................................................. 34 4.2 Flow Rates .......................................................................................................... 37 4.3 Soil Moisture ................. ..... ...... .... ..................... ... ... ............................ ... ............. 40 4.4 Evapotranspiration .............................................................................................. 44 4.5 Water Use ................... ......................................................................................... 44 4.6 Processes and Relationships ............................................................................... 45 5. SUMMARY AND CONCLUSIONS ..................................................................... ..... 57 APPENDICES A. TEMPERA TURE ........................................................................................................ 60 B. TEMPERATURE AND PRECIPITATION REFERENCES vi ........................................................ .. .... 63 ....................................................... .......................................................... 65 VI LIST OF TABLES Table Page 2: County Site dataset summary of errors 36 3: Storms with significant precipitation (over 12.7 mm) 36 4: Storm water flow meter experiment results 38 44 45 48 1: Instrument Summary ......................................................................... 32 ....... , ............................................ .................................. .36 ................................. ' ............. 5: ET compared with other common climate variables ................................... .44 6: Outdoor water use monthly totals ................. , .... .................................. .45 7: Runoff coefficients for major storm events ............................... , .............. .48 LIST OF FIGURES Figure Page 1: Urban water cycle before and after urbanization 3 2: Pre-urban water cycle 3 3: Posturbanization water cycle 4 4: Salt Lake Metropolitan area average water cycle 6 5: Location of the Murray neighborhood 16 6: Murray catchment land cover characteristics 17 7: Murray catchment land cover 18 8: Storm water drainage system and catchment boundary 19 9: Site overview of monitoring instrument locations and catchment boundary 20 10: Tipping bucket rain gauge (Hydrological Services, Type TB-3) 22 11: Field setup of rain gauge and power inverter 22 12: Stingray area-velocity flow meter 24 13: Capabilities of flow meter 24 14: Field setup of the flow meter 25 15: Soil moisture data logger and probe 27 16: Site P.H.- Late afternoon hours in August 2008 28 17: Location P.C. (Power Corridor) of moisture sensors September of 2007 28 18: Location P.C. of moisture sensors August of 2008 29 ........................................................... .................................................................................................... ......................................................................................... .......................................................... ......................................................................... .............................................................. ...................................................................................... ............................................... ............... .............................. inverter. ........................................................... meter. ............................................................................. .......................................................................................... .................................................................................... .......................................................................... ........................................................ of2007 ................. of2008 ................................................... 19: Placement of soil moisture sensor data logger 30 20: Installation of soil moisture sensor by trenching 30 21: Precipitation comparison of the three rain gauges 35 22: Depth adjustment for storm water flow 39 23: Weekly runoff totals 40 24: Monthly runoff totals 40 25: Soil moisture content at sites P.C. and P.H. from September through October 2007 41 26: Soil moisture comparison of irrigated lawn areas at both sites P.H. and P.C 42 27: Soil moisture comparison of irrigated areas at site P.H 42 28: Soil moisture comparison of nongrassy areas for both sites P.H. and P.C 43 29: Seasonal variation of monthly runoff coefficients 47 30: Monthly irrigation flow compared with average outdoor water use 50 31: Monthly irrigation waste flow compared with maximum outdoor water use 50 32: Soil moisture response to precipitation from September through October 2007 51 33: Soil moisture response to precipitation for site P.H 52 34: Soil moisture response to precipitation at site P.C 53 35: Soil moisture content vs precipitation for site P.H. garden in September 2007 53 36: Nonirrigated soil moisture vs. precipitation for isolated time frame 54 37: Nonirrigated soil moisture content vs ET rate response for September 26, 2007 55 38: Irrigated soil moisture content compared to the ET rate response for September 26, 2007 55 39: Urban water budget for Murray Catchment 56 ix ........................................................... ....................................................... ..................................................... ..................................................................... .................................................................................................. ................................................................................................. 2007 ......................................................................................... 41 26: Soil moisture comparison of irrigated lawn areas at both sites P.H. and P.C ............. 42 27: Soil moisture comparison of irrigated areas at site P.H. ............................................. 42 28: Soil moisture comparison of non grassy areas for both sites P.H. and P.C ................ 43 29: Seasonal variation of monthly runoff coefficients ..................................................... 47 30: Monthly irrigation flow compared with average outdoor water use ......................... 50 31: Monthly irrigation waste flow compared with maximum outdoor water use ............ 50 32: Soil moisture response to precipitation from September through October 2007 ..................................................................................... 51 33: Soil moisture response to precipitation for site P.H .................................................. 52 34: at site P.C .................................................... 53 35: content vs precipitation for H. garden in September 2007 ........ 36: Nonirrigated soil moisture vs. precipitation for isolated time frame ......................... 54 37: content vs ET rate response for September 26,2007 ..... 55 38: Irrigated soil moisture content compared to the ET rate response for September 26, 2007 ............................................................................................................... 55 39: Urban water budget for Murray Catchment.. ............................................................. 56 IX Al: Daily Maximum Temperatures from September 2007 to August 2008 62 A2: Monthly Average Maximum Temperature from Salt Lake County Site 62 Bl: Temperature compared to precipitation events 64 AI: ................... .................. BI: ......................................................... x ACKNOWLEDGEMENTS There is nothing greater in life than having the support of family and friends when life's mountains are difficult to climb. I would like to thank all those that have contributed to this thesis. I would like to especially thank my wife Kristin for supporting me and allowing me to dedicate such long nights and weekends to this work. I am very blessed to be married to someone so understanding. I have truly been honored throughout my life to have worked with such great scholars, colleagues and friends. I express my sincere appreciation to Dr. Steven Burian for his support and ability to lift and encourage students such as myself to achieve their greatest potential. Finally, I would like to convey my appreciation to all those that have spent time collecting data and blundering around in a manhole; this could not have been completed without their help. life 's CHAPTER 1 INTRODUCTION The urban growth rate in the Salt Lake City, Utah metropolitan area is near the highest among metropolitan areas in the U.S. According to the 2000 Census, Utah was ranked as the 4 t h fastest growing state in the nation, and has recently (2006-2007) achieved the 3 r d fastest growth rate. Growth is expected to continue with an estimated increase in population of 48% by 2020, with most of this growth occurring along the Wasatch front (adjacent to the Salt Lake City metropolitan area) where over 80% of Utah's population currently resides (Envision Utah, 2007). Industrial and commercial growth is expected to be even greater with the number of people employed in Utah predicted to reach two million by 2020, or double the number employed in 1994 (Utah's Water Resources, 2003), a 200% increase in 34 years. The population growth in Utah will expand the urban footprint significantly, with the current number of buildings expected to double by 2050. However, the resulting impact on the urban footprint is difficult to estimate because much depends on the pattern and characteristics of the urban development. Researchers such as Dr. Arthur Nelson at the University of Utah are estimating that better use of the current building footprint will limit the impacts of urban sprawl that may otherwise be observed (Leonard, 2008). Summers (1999) stated that by 2020, Envision Utah, in their plan to conserve water 4th 3rd 2 through efficient urban planning, will have decreased the size of the average residential lot from 0.129 ha (0.32 acres) to 0.117 ha (0.29 acres) and will have reduced the water usage by 6%. Urbanization affects the water cycle by altering the natural landscape. Indigenous vegetation is removed, impervious surfaces are introduced, pervious surfaces are modified, and ornamental landscapes are created. The degree of urbanization has been suggested as the most dominant factor altering the hydrology of an area (Claessens et al., 2006). Initially the removal of indigenous vegetation, grading of land surfaces, construction of impervious surfaces and landscaped areas, and importation of water supply change the fluxes and stores of water from predevelopment conditions (see Figure 1). The graphics in Figure 1 suggest there will be greater surface runoff caused by the increased impervious surfaces. This should logically suggest a decrease in infiltration; however, when imported water is considered, the results may be different. Berg et al. (1996), for example, suggested area available for infiltration may indeed be reduced, but the management of stormwater and the over-irrigation of landscapes may cause the net infiltration/groundwater recharge to actually increase. Figures 2 and 3 provide a hypothetical quantification of the impacts to the water cycle fluxes caused by urbanization. Figure 2 shows the predevelopment fluxes with the size of the arrow indicating the relative magnitude of the flux. Figure 3 shows the postdevelopment water cycle. The change in size of the arrows (i.e., water cycle flux magnitude) represents one possible hypothetical set of conditions. The effects are uncertain and cannot be generalized without consideration of climate and site-specific geophysical features, urban characteristics, and human behavior. omamentallandscapes aI., cycle. The change in size of the arrows (i.e., water cycle flux magnitude) represents one possible hypothetical set of conditions. The effects are uncertain and cannot be generalized without consideration of climate and site-specific geophysical features, urban characteristics, and human behavior. 3 Figure 1: Urban water cycle before and after urbanization. ^ / I N EVAPDTRANSPIRATIDN A EVAPDTRANSPIRATIDN RUNDFF INFILTRATION \ GROUNDWATER Figure 2: Pre-urban water cycle. VATER CYCLE BEFORE URBANIZATION WATER CYCLE AFTER URBANIZATION PRECIPITATION EVAPOTRANSPIRA, TIO N RUNOFF , ~APDTRANSPIRATIDN ~ '" 1 INF,IL TR A TION GROUND\.,! ATER RECHARGE 4 / IMPORTED WATER E V A P OT R. A NS PIR A TI • N \ EVAPDTRANSPIRATIDN RUNOFF Figure 3: Posturbanization water cycle. INFILTRATION I GROUNDWATER RECHARGE In western U.S. cities, the effects of urban areas on the water cycle are generally more pronounced than those observed in eastern U.S. cities because the changes to the landscape are more extreme and precipitation is less. In the Salt Lake City (SLC) metropolitan area, for example, the average annual rainfall is 406 mm, for a total volume of about 776,150 million liters (assuming Salt Lake County area of 1909.8 km" based on the 2000 Census). Urban water use in SLC is approximately 1011 liters per capita per day (lpcd) (267 gpcd) (Utah Division of Water Resources, 2003). 575 lpcd (152 gpcd), or 188,160 million liters, on average is used outdoors (assuming a Salt Lake County population of 898,387 based on the 2000 Census). Focusing on residential, daily water use in Salt Lake County is approximately 693 lpcd, of which 428 lpcd is used outdoors (Utah Division of Water Resources, 2003). On an annual basis, the residential outdoor water use is nearly 20% of the precipitation falling in the Salt Lake County boundary. Stream flow observations from a gage located at 500 North in SLC report varied volumes of flow from 138,000 million liters during the 2005 water year (October 2004 PRECIPITATION / ~IMPORTED ~ATER EV APDTRANSPIRA nON " ~ APDTRANSPIRA nON \ JI' [~TR!IDN RUNOFF CiROUNDVATER RECHARGE km2 5 through September 2005) to 186,000 million liters during the 2006 water year (October 2005 through September 2006) for an average annual streamflow volume of 162,000 million liters. The evapotranspiration (ET) flux in the SLC area represents a significant component of the water budget. Using the Penman-Monteith equation and considering the land use in the entire Salt Lake County extent, a volume of 1,481,465 million liters of water is exchanged from the ground surface to the atmosphere via ET. Figure 4 captures the water budget components of the SLC metropolitan area (as defined by the Salt Lake County boundary). The figure is organized to show water entering and exiting the SLC metropolitan area. To create the figure, it was assumed that outdoor water use is imported from mountainous areas outside of the SLC boundary and the streamflow represents the ultimate destination of the wastewater collection and treatment system. However, this is also assuming that no additional water enters the stream prior to reaching Salt Lake County (which is not entirely true). Figure 4 suggests water leaving the SLC metropolitan area is greater than water entering. The message of this graphic is existing data sources do not provide accurate estimates of the urban water budget in SLC, not do they provide the quantities at the levels useful for many neighborhood to city-scale decisions. The SLC water budget represents average conditions, but variability introduced by human behavior, public policy, climate, and energy budget interactions may influence the budget significantly, creating problems for urban water management planning and system design. To aid the understanding of the impact of external influences on water cycle processes, the relationships between the water cycle and water conservation, climate variability, and the energy budget are described in the next three subsections. city -6 1481000 Volume of Water (Million Liters) Salt Lake Metropolitan Water Budget Figure 4: Salt Lake Metropolitan area average water cycle. 1.1 Conservation The state of Utah, in their plan to conserve water, has a goal of reducing the per capita water use from 1215 lpcd to 908 lpcd, approximately a 25% decrease. This is nearly 493,500 million liters (400,000 acre-feet) each year that would be made available for future use if conservation can be enacted (Utah Division of Water Resources, 2003). An additional document was created by the Utah Water Conservation Advisory Board in 1995 titled Water Conservation Recommendations. They provided recommendations such as requiring management plans, restricting landscape irrigation to nighttime hours, metering implementation, water pricing to promote conservation, and annual accounting to quantify system losses and repair these losses. As a part of developing a water management plan, the Utah communities must identify areas of inefficiencies. The term "inefficiency" could be described as any resource that is not being used to its maximum capacity and usefulness or is being used more than is required. In the water budget, this could signify that the rainfall in the area is not being used to its fullest extent, or that Water Out Water In 776000 Preci~itation I Evapotranspiration 189000 Outdoor Water Use -,-----,-- -,.- 162000 Streamflow " inefficiency" could be described as any resource that is not being used to its maximum capacity and usefulness or is being used more than is required. In the water budget, this could signify that the rainfall in the area is not being used to its fullest extent, or that 7 sprinkler watering systems are not being used to their capacity, or are being used past what is necessary to maintain healthy landscaping. The Jordan Valley Water Conservancy District (JVSCD) in a recent news report indicated that water managers had been making progress and actual water use in that particular water district had been reduced from about 965 lpcd (255 gpcd) to about 784 lpcd (207 gpcd) between the years 2000 and 2005, but has since increased to nearly 950 lpcd. The article stated "Utahns need to remember the difference between weather and climate: Even if it rains, Utah is still the second-driest state" (Henetz, July 12, 2008). The program in place appeared to be helping conserve water and thus, serving its purpose. A decrease of nearly 9% in per capita water use was observed from 2000 to 2002. But, with the increased water usage there has been a net increase in demand of 12%; due in part to population growth. This brief analysis suggests water conservation practices if continued to be implemented widespread would reduce water imported to the SLC metropolitan area, possibly reduce outdoor water use, and then have subsequent cascading effects on microclimate, air quality, and more. 1.2 Climate Variability Climate variability might further perturb urban water cycle fluxes and stores, creating variability not represented in annual averages. A study described by Robert Kunzig in the February 2008 edition of National Geographic explains the notion of "normal" temperature and precipitation in the West is not correct. He explains that researchers have been able to determine that the 1900s contained some of the wettest years ever recorded. Through tree ring analyses, the so called "drought" conditions appear to be more normal than the recent data indicates (Kunzig, 2008). Water managers NSCD) 8 must plan to avoid a water shortage crisis. They seek to identify a cost effective combination of water use reduction (conservation) and new water supply development (e.g., new reservoirs and pipelines) to meet future water demands. Again, climate change only compounds the problem. In fact, one problem they are facing is the prediction that there will be a shortfall of 986,785 million liters (800,000 acre-ft) of water per year by 2050 for use in the Great Salt Lake Basin area (Great Salt Lake Basin, 2007) caused only by climate change and not taking into account the drought conditions that may be the "norm" for the area. At the state level, the limited availability of the Colorado River and the potential for future reductions further complicate the task. Even though recent agreements to manage the Colorado River allocations in times of drought have been put in place (Phenetz 2007), urban water managers must continue to consider the effects of climate change on their systems. For example, might climate change alter vegetation impacting ET and, in turn, soil moisture? Might the human response to climate change be to modify outdoor irrigation patterns, which would alter water use, irrigation, and ET? The problem transcends scales and all scales must identify the impacts and develop solutions. As Dirk Kempthorne stated in response to the Colorado River drought management strategy, "there will be no true winners unless everyone gets something and everyone gives up something" (Henetz, 2007). 1.3 Energy Budget The water cycle is inexorably linked to the energy budget. In urban environments, changes to the water cycle have complex, cascading effects on vegetation, surface fluxes, microclimate, energy use, air quality, and public health. Studies of this linkage between water cycle and urban energy budgets are beginning to expand in depth effective of986,785 management strategy, "there will be no true winners unless everyone gets something and everyone gives up something" (Henetz, 2007). 9 to reduce uncertainties and uncover hidden effects. The results from this thesis research will help to inform at the first order level the linkage between the water cycle and the energy budget through the influence of soil moisture on the ET flux. The urban water cycle is comprised of fluxes and stores influenced by urban form, human behavior, and climate. The modifications made to the water cycle have cascading effects on other components of the urban system, including microclimate, energy use, air quality, and more. Continuous observations of the urban water cycle are needed to quantify key fluxes and stores, and to better define the uncertain relationships among water cycle fluxes, stores, climate, and human behavior. The study described in this thesis takes the first steps to address this need in semi-arid urban environments. The goal of the research was to reduce uncertainties of urban water cycle fluxes and stores by quantifying their values through continuous observation and defining their seasonal variability and interrelationships. This goal has been achieved by accomplishing the following objectives: • Developed and installed a continuous urban water cycle observation network in a SLC residential area (Murray, Utah). • Compiled 14 months of precipitation, dry weather and wet weather runoff, and soil moisture data along with several days of ET flux data and combined these with historical water use data and relevant meteorological observations. • Analyzed trends and assessed seasonal variability of precipitation, runoff, soil moisture, and ET and related these to climate variation. Assessed relationships between precipitation and wet weather runoff, precipitation and soil moisture, water use and irrigation runoff, irrigation and soil • runoff, 10 moisture, and soil moisture and ET. The next chapter provides a brief summary of key observation studies of the urban water cycle. The third chapter summarizes the urban water cycle observation plan and methods implemented in this research. The fourth chapter presents the research results and aims to answer the following overarching research questions: • What is the seasonal variation of precipitation-to-runoff conversion? • What is the seasonal variation of irrigation waste flows and is it related to common climate variables? • What is the soil moisture response to rainfall and irrigation for different land cover/sun exposure combinations? • What is the effect of soil moisture on ET? The thesis concludes with a summary of key findings and recommendations for future work. runoff CHAPTER 2 LITERATURE REVIEW Long-term continuous observations of water cycle fluxes and stores and their response to urbanization are necessary to quantify the impacts of future urbanization on the water cycle. Given observations, processes can be characterized and models can be developed and validated. In addition, the relationships among fluxes, stores, and processes can be studied and defined to further refine the models for use in predicting impacts of urbanization and planning for their mitigation. The interconnection of the water cycle fluxes and stores in an urban environment requires a comprehensive approach to observation. For the thesis research described herein, a monitoring system was installed to observe the primary components of the urban water cycle. The data collected and the analysis and synthesis of the data constitute the major contributions of the research. To set these results in the context of other urban water cycle studies, whether completed or underway, key studies with similar objectives were selected and are summarized in this chapter. Additional references to these studies will be made in Chapter 4 when the results are presented. There have been a few published studies of urban water cycle observation networks. Here we review the studies with the similar goal of studying the urban water cycle fluxes and stores and their interrelationships. Grimmond et al. (1986), for example, used a water balance model approach with readily available data to analyze the interaction of the urban hydrologic cycle components. Assumptions such as assuming the suburban area consists only of three surface types (pervious irrigated, pervious non-irrigated, and impervious) and assuming specific empirical coefficients for the ET model were necessary to complete their study. The data obtained was site-specific to the Vancouver, British Columbia area, which may not be transferable to other locations when applying the model. This indicates that site-specific models and data are imperative when studying the urban hydrologic system. A study completed by Berthier et al. (1999) provided a set of urban water cycle observations, similar to the set collected for this thesis research. They collected continuous observations of rainfall, stormwater discharge, soil moisture, water table level, and meteorological characteristics at the Reze catchments for 7 years when their paper was published. They were able to set up a continual monitoring system for two side-by-side catchments. The time increment of most observations was 1 minute, providing one of the more substantial and comprehensive urban hydrologic databases available anywhere. One key aspect missing from their study that was included in this thesis is the ET flux. In Los Angeles, California, Xiao et al. (2006) studied the urban hydrologic cycle and the relationship of hydrologic processes to best management practices (BMPs). Their objective was to quantify the effect of BMPs on water cycle fluxes and stores. They focused observations at the microscale level - at two residential land parcels - to observe microclimate (temperature and relative humidity), outdoor water use, soil moisture, and surface runoff. There have been other urban water cycle observation networks established that 12 non­irrigated, micro scale have not been published for their complete urban water cycle observations. For example, there are instrumented urban catchments in Baltimore and Phoenix associated with the National Science Foundation Long-Term Ecological Research (LTER) project. The Gwynn Falls catchment in Baltimore is one of the more well-instrumented catchments, but does not have dedicated ET flux measurement. Comprehensive urban water cycle modeling studies have also been performed and in some cases have been linked to a limited set of observations. One has been associated with the AquaCycle modeling system introduced by Mitchell et al. (2001). This model has been tested for a watershed in the Woden Valley area of Canberra, Australia (Mitchell et al., 2001). The hydrological observation networks established in the studies mentioned above have accomplished their objectives. The study by Xiao et al. (2006) proved that BMPs are an effective way to reduce runoff, water use, and the spread of contaminants downstream. They also illustrated that best irrigation management practices would help to further reduce water use. The main goal for the study by Berthier et al. (1999) was to set up a monitoring system in order to provide a database for additional urban hydrological studies. This was a success and additional authors have studied items such as the variability of runoff, effectiveness of rainfall-runoff models, and infiltration simulations. Ramier et al. (2006) discovered through their monitoring of street runoff that the hydrological behavior of streets is not well defined and that there is infiltration and evaporation on some urban surfaces that is not correctly modeled and does not contain sufficient data from which to draw a valid conclusion. A previous study by Ragab et al. (2003) confirms there is a portion of precipitation that infiltrates through the urban road surfaces. The relative limited number of studies, and the lack of applicability 13 aI. aI., aI. aI. runoff aI. aI. urban road surfaces. The relative limited number of studies, and the lack of applicability 14 to conditions in the semi-arid mountain west U.S., support the need to quantify urban hydrologic fluxes and stores. The next chapter will present the observation network that was set up to address the research questions posed in Chapter 1. CHAPTER 3 OBSERVATION NETWORK AND METHODS 3.1 General A water cycle observation network was established in a 17.7 ha mature residential neighborhood in the SLC metropolitan area. The observation network was established to capture the primary hydrologic inflow to the system (precipitation), a key outflow (runoff), the ET flux, and the variation of the soil moisture store. The temporal resolution of the observations was set fine enough to capture diurnal variations and individual events. Observations not collected by the network but pertinent to the research (e.g., temperature and water use) were acquired from other sources or companion studies. 3.2 Murray Urban Water Cycle Observation Network 3.2.1 Catchment Characteristics The Murray catchment is located towards the south end of the Salt Lake Valley near where the Jordan River divides the valley into east and west halves (Figure 5). The catchment consists of 162 houses built during the late 1970s and early 1980s. Lots are characterized by mature landscaping, with approximately 700 mature trees covering nearly 3.5 ha (roughly 20% of the total catchment area). The catchment is comprised of numerous residential streets and 9 cul-de-sacs. The catchment is 48% impervious with the greatest fraction of land covers being comprised of building rooftops and roadways OBSERV ATION A water cycle observation network was established in a 17.7 ha mature residential neighborhood in the SLC metropolitan area. The observation network was established to capture the primary hydrologic inflow to the system (precipitation), a key outflow (runoff), the ET flux, and the variation ofthe soil moisture store. The temporal resolution of the observations was set fine enough to capture diurnal variations and individual events. Observations not collected by the network but pertinent to the research (e.g., temperature and water use) were acquired from other sources or companion studies. 16 Figure 5: Location of the Murray neighborhood. The study location within the Salt Lake City metropolitan area (solid line) and the Jordan River basin (dashed line). The proximity to the Great Salt Lake and the major interstate highways is shown for reference. (Figure 6). A visual representation of the different land covers can be seen in Figure 7. The drainage system is configured with storm drains located at the inner portion of the cul-de-sacs, causing storm drain lines to cross through private property (Figure 8). The catchment elevation is between 1314 and 1318 meters above sea level and, in general, has a relatively flat slope. The slope between one end of the catchment and the other is about 2.0% with most of the drop occurring in the last 100 meters (11.0% slope) before reaching the outfall location. The outfall from the catchment lies approximately 0.34 km east of the Jordan River with a power corridor situated between the catchment ..­" -" /' West Bench " Catchments "'- \ we~ench .... /' " ,R eference Site -" " I \ \ ~ __ Salt Lake City ropolitan Area .) Boundary (2000) _/--~,) _ /' /' J"- ~rdan River Basin o_ .:=5 ==10- ___20 K ilometers 17 Characteristics Figure 6: Murray catchment land cover characteristics. and the Jordan River. The stormwater drainage network connects the hydrological aspects of the catchment and consists of 12 catch basins and 3200 meters of pipe. The SLC weather pattern consists of four distinct seasons, with cold winters (average maximum temperatures ranging from 0°C to 12°C) and warm summers (average maximum temperatures ranging from 28°C to 33°C). The spring and fall months are shorter and create a transition period between the two extremes. The spring usually generates the highest amount of rainfall, whereas snowfall occurring in the winter accounts for a large fraction of the annual precipitation. The mountain snowpack that either drains into streams and is diverted to treatment facilities or recharges groundwater accessed by production wells provides the majority of the water for urban supply in SLC. 35.00% III CII 30.00% ... III 30.00% ~ 25.00% .0.. 20.00% 20.00% 20.00% .... 20.00% 0 CII 15.00% C) .cfI: 10.00% 10.00% CII C,) ... 5.00% CII a.. 0.00% (1,0 ~o ftt' ~ ",?-i?J Murray Catchment Characteristics O°Figure 7: Murray catchment cover. Land Cover "tYpe 0 Reflective Rooftop _ Asphalt Road [·~\".l Grass .~ - Residential Rooftop Tree boundary land 100 18 N + 50 o 100 Meters 19 A sd_manhole boundary _JV M catch_basin If drnjine 100 50 0 100 Meters Figure 8: Storm water drainage system and catchment boundary. 3.2.2 Observation Network Overview The observation network infrastructure included eight soil moisture sensors, a stormwater flow meter, a heated rain gauge, and an Eddy-covariance flux tower to measure ET. Monthly water billing records were obtained to provide estimates of indoor and outdoor water use and temperature data were also obtained for use in analyzing the interrelationships of the water cycle to common climate variables. The monitoring system is illustrated in Figure 9. To capture soil moisture dynamics in sprinkler irrigated • III - drnJine N + o Stonn stonnwater 20 Rain Gauge & ET Monitoring Location Soil Moisture Sensor Locations at Site P.C. 1 1 * " 1 #* y ' Flow meter location Sensor P.Salt Lake County Site (2.36 km from outfall) Monitoring Sensor Locations I | Boundary N Instruments 0 0.2 0.4 0.8 1.2 1.6 • kilometers P.C . • Taylorsville Rain Gauge (3.75 km from outfall) Soil Moisture Sensor • Locations at Site P.H. D • Insbuments + o_ _C 0J.2l _0C.4I I_-=0=.8= =:::1J.2_ _1.6l < ilomelers Figure 9: Site overview of monitoring instrument locations and catchment boundary. Catchment boundary shown in orange. 20 21 landscapes, drip line irrigated landscapes, and nonirrigated vegetated landscapes, soil moisture sensors were positioned both at a residential parcel containing sprinkler and drip line irrigation, and at a nearby nonirrigated power corridor. Precipitation data were obtained from three different locations; one was in a neighboring catchment to the north of the study site and two other locations were selected at nearby recording sites (see Figure 9) with a gauge monitored by Salt Lake County located 2.36 kilometers from the outfall location and another labeled as the Taylorsville site located 3.75 kilometers from the outfall. Both the on-site rain gauge and ET monitoring equipment are located in a protected government facility for accessibility and security purposes. 3.2.3 Rain Gauge A heated rain gauge was installed in June of 2007 at the government facility mentioned, slightly to the north of the outfall location of the catchment. The heated rain gauge is a tipping bucket style (Hydrological Services Type TB-3, Figure 10) and is calibrated so that each tip of the bucket measures 0.01 inches of precipitation. The gauge sums precipitation depth at a time interval of 5 minutes and records the value to the data logger. Figure 11 shows the installed rain gauge, which required mounting atop a steel pole and securing the pole. An inverter was placed inside a water tight box (also shown in Figure 11) and is used to supply 12-volt power to the rain gauge heater in times of extreme cold to prevent excess precipitation from freezing and restricting movement of the tipping bucket. A 200 foot cable connects the rain gauge to the data logger located inside a secure facility. of2007 I2-22 Figure 10: Tipping bucket rain gauge (Hydrological Services, Type TB-3). Figure 11: Field setup of rain gauge and power inverter. 23 Due to power outages and other unforeseen problems with the data logging of the rain gauge measurements, the precipitation data was not accurately recorded until October 2007 when some of the issues relating to these outages were fixed. These types of gauges are best used in areas that are somewhat protected from the effects of wind and overhead obstructions. This site was chosen, however, mainly due to accessibility constraints in order to access the data logger set up at the government facility. It was also chosen for protection from vandals. The risk of wind affecting the results at this site was considered to be less of an issue than the overhead obstructions (based on consideration of wind speed variation with height at the site). Thus, to limit the impact of obstructions, the high profile was necessary. Since power outages affected the data logging and precipitation depths may have also been affected by the high profile of the setup, additional precipitation data were obtained from two other nearby locations (Figure 9) using the same type of heated rain gauge. The rain gauge located in Taylorsville is monitored by Weather Underground (www.wunderground.com). A comparison of these three rain gauges will be shown in the results portion of this thesis. 3.2.4 Runoff Flow Meter A Greyline Stingray area-velocity meter (Figure 12) was installed in a manhole in the storm drain pipe draining the catchment. The Stingray measures both velocity and depth of flow using two built-in sensors (the angled sensor measuring velocity and the vertical facing sensor measuring depth (Figure 13) (www.grevline.com/howitwk av.htm)). Flow may be determined automatically by the instrument or externally by specifying the shape and size of pipe. (www.greyline.comlhowitwk 24 Figure 12: Stingray area-velocity flow meter. Figure 13: Capabilities of flow meter. The flow meter was installed in an 18-inch storm drain by attaching it to the bottom center of the flow line with two concrete anchors (Figure 14). The location of the storm drain flow meter was selected in order to capture all the stormwater drainage from the catchment of interest. The flow meter was installed and began recording continuously June 22, 2007. The data was recorded at a time step of 1 minute until August 31, 2007, when it was switched to a 2-minute time step to conserve battery life. 3.2.5 Evapotranspiration ET is a component of the water cycle that is normally found to be the largest output in the average urban water balance (Cleugh et al., 2007). Three main methods are 31 , aI., 25 Figure 14: Field setup of the flow meter. normally used to estimate ET. These are energy budget formulations, water budget formulations, and mass transfer formulations. Mass transfer equations are based on the concept of the turbulent transfer of water vapor from an evaporating surface to the atmosphere. The Eddy-correlation method is one of the prominent mass transfer methods employed to estimate ET (it was also used for this study). the Murray catchment, ET was measured by Dr. Eric Pardyjak's research group using a set of sonic anemometers and sensors that were mounted to a tall tower to measure concentrations of water vapor and other gases. The latent heat flux was determined from the 3-D velocity observations and vapor/gas concentrations. The equation used to determine the volumetric ET rate given the flux is shown below: ET = - - - * 1000 * 180 Equation 3.1 Where ET is the evapotranspiration rate in mm, L is the latent heat flux in W-s"1, p w is the density of water (1000 kg-m"3), X v is the latent heat of vaporization (2.5* J-kg"1), 1000 In = L * * 180 Pw * Av Equation 3.1 W -s -I , pw kg_m-3 ) , Av 106 kg- I ), 26 is a conversion factor, and 180 is the time conversion from seconds for the 30 minute recording increment of this particular data. The ET value was then converted to units of 1000 liters by multiplying by the area of the catchment (17.7 ha) and converting to 1000 L units. The ET monitoring devices are located in close proximity to the rain gauge at the same government facility, but were set up and maintained by students of Dr. Eric Pardyjak of the University of Utah. The ET flux data was provided for three dates for comparison and analysis purposes in this study: September 25 and 26, 2007 and October 1,2007. 3.2.6 Soil Moisture Sensors Soil moisture sensors were installed in the catchment to observe differences in soil moisture in sprinkler irrigated landscapes, drip line irrigated landscapes and non-irrigated areas, and to quantify the effects of shading on soil moisture. Onset Echo EC-5 sensors and a HOBOware microstation data logger were used to collect soil moisture readings continuously during warm weather (sensors were removed when temperatures dropped below freezing). The sensors use the concept of time domain reflectometry to determine volumetric water content to +/- 3% accuracy in most soils. Figure 15 shows the sensors and the microstation data logger. The sensors installed were set to record data every 2 minutes in order to view diurnal fluctuations in the moisture content. A total of four sensors were installed on September 14, 2007 and removed on November 2, 2007 when temperatures dropped below freezing. The sensors used in 2007, along with four additional sensors, were re­installed in April of 2008 when the ground thawed for the summer months. At site P.H. (Figure 9), one soil moisture sensor was located in a shaded grassy area that received a non­irrigated of2008 27 Figure 15: Soil moisture data logger and probe. lot of irrigation water, whereas the other was located in a garden area that was watered using an irrigation drip line. At site P.C., one sensor was placed in an irrigated grassy, sunny area and the other was installed in a nonirrigated, sparsely vegetated sunny area of the power corridor. Figure 9 shows sensor locations as installed in April of 2008 whereas Figure 16 shows a much clearer picture of the specific location of the lawn area at site P.H. for 2008. Figures 17 and 18 show the grassy area in the power corridor and the soil moisture contrast between the two years. In 2007, the lawn area of the power corridor was irrigated frequently and remained green throughout the year, whereas in 2008 the lawn was not watered as much, resulting in dry looking lawn and assumed lower soil moisture content. The sensors installed in the spring of 2008 were strategically placed in different landscapes to observe distinctions. At site P.H., one sensor is located in a sunny, grassy area (Figure 16) and one is located in the garden just a few feet from the lawn and near a drip line. The other two are also located in the garden area; one is in a sunny section of different P .Figure 16: Site P.H.- Late afternoon hours in August 2008. Figure 17: Location P.C. (Power Corridor) of moisture sensors September of 2007. 28 of2007. 29 Figure 18: Location P.C. of moisture sensors August of 2008. the garden whereas the other is shaded, and both are close enough to irrigation drip lines that the excess moisture should be observed. The additional four moisture sensors are located at site P.C. with two sensors, again located in an irrigated lawn area whereas the other two are in the nonirrigated power corridor. The sensors were set up by placing the data loggers in a water tight irrigation box (Figure 19) and trenching out to the desired location of the sensor (see Figure 20). This served not only to protect the data loggers against the elements, but also was used as a protection from tampering and other hazards such as lawn care equipment. 3.2.7 Water Use Data such as the external or outdoor water use are difficult to obtain, especially for smaller areas (Grimmond and Oke, 1999). Since a sewer monitor will not be set up to of2008. Figure 20: Installation of soil moisture sensor by trenching. 30 Figure 19: Placement of soil moisture sensor data logger. Figure 20: Installation of soil moisture sensor by 31 determine total indoor water use, a percentage of water used for outdoor purposes must be assumed based on differences between summer and winter water use records. Additional water used during the summer months such as that for evaporative coolers that does not contribute to outdoor irrigation should be accounted for if the water use is determined as the difference between winter and summer month usage rates. These evaporative coolers, for example, generally use between 11.4 and 56.8 liters a day during summer months depending on the size of the house, contributing to the overall ET rate and making the use of evaporative coolers an important part of the water budget. Water use data included in this study are from Murray City in the form of monthly billing records from November 1997 through September 2005. 3.2.8 Water Budget Observations were collected from June 2007 to November 2007 to establish baseline conditions for comparison to subsequent seasonal monitoring. The soil moisture sensors were removed in November; however, the rain gauge, the flux tower, and the pipeflow meter monitored the precipitation, ET, and runoff continuously throughout the winter months between November 2007 and March 2008. The soil moisture sensors were re-installed during April of 2008. Listing of instruments and summary details of installation dates and recording frequencies are contained in Table 1. Collectively, the instruments provided data of the urban water budget from September 1, 2007 and August 31, 2008. This duration is sufficient to provide preliminary quantification of fluxes and stores and to define relationships and trends. Also, seasonal variability can be assessed because the duration spans the four seasons (essentially the entire 2007 water year). detennine detennined summer months depending on the size of the house, contributing to the overall ET rate and making the use of evaporative coolers an important part of the water budget. Water use data included in this study are from Murray City in the fonn of monthly billing records from November 1997 through September 2005. 1,2007 stores and to define relationships and trends. Also, seasonal variability can be assessed because the duration spans the four seasons (essentially the entire 2007 water year). 32 Table 1: Instrument Summary Instrument Description Date Installed Date Removed Date Installed Date Removed Recording Frequency Tipping Bucket Rain Gauge On-site-heated to record precipitation during winter months Oct-07 - - - 5 minute increments Tipping Bucket Rain Gauge Monitored by Salt Lake County - - - Event Specific- sub-daily increments Tipping Bucket Rain Gauge Monitored by Weather Underground - - - 15 minute increments Stingray Flow Meter Area-Velocity Flow Meter Jun-07 - - - 2 minute increments Soil Moisture Sensors PH-1 9/14/2007 11/2/2007 4/24/2008 - 2 minute increments Soil Moisture Sensors PH-2 9/14/2007 11/2/2007 4/24/2008 - 2 minute increments Soil Moisture Sensors PH-3 - - 4/24/2008 - 2 minute increments Soil Moisture Sensors PH-4 - - 4/24/2008 - 2 minute increments Soil Moisture Sensors PC-1 9/14/2007 11/2/2007 4/24/2008 - 2 minute increments Soil Moisture Sensors PC-2 9/14/2007 11/2/2007 4/24/2008 - 2 minute increments Soil Moisture Sensors PC-3 - - 4/24/2008 - 2 minute increments Soil Moisture Sensors PC-4 - - 4/24/2008 2 minute increments ET On-site Oct-07 - - - 15 minute increments Freguenc~ sub- daily 212007 4124/911412007 1112/412412008 4124/212007 4124/212007 4124/4124/2412008 Equipment 33 3.3 Data Collection and Analysis The observations and data collection results and analyses will be described in the next chapter. The results contain observations of the following water cycle components: • Precipitation • Dry weather and wet weather flow rates • Soil moisture • Evapotranspiration • Water use After presenting the data, the results are analyzed to quantify the trends and inter­relationships among water cycle fluxes and stores. These analyses include an error analysis, diurnal and seasonal variability, the relationship among water cycle components, and their relationship to climate. • Precipitation • Soil moisture • Evapotranspiration • Water use CHAPTER 4 RESULTS AND DISCUSSION The first set of subsections of this chapter describes the individual data measurements along with possible errors and trends. The second set of subsections describes the synthesis of the data to study the relationships among water cycle components. 4.1 Precipitation The installed rain gauge, as mentioned previously, likely had data errors caused by power outages in the data logger. In fact, there were a total of 138 days that contained suspected data errors, 91 of which recorded no data entries (out of 335 total days). Also, the rain gauge being placed amongst obstructions could have caused errors. Consequently, additional precipitation data were acquired from nearby rain gauge sites. The two sites selected to provide additional rainfall data are located within 4 kilometers of the study location. The data for the Taylorsville location was downloaded from www.wunderground.com by selecting the site listed as KUTTAYL04 and the data from Salt Lake County was also available via the web http://www.pweng.slco.org/flood/precip /index.cfm). A comparison of the precipitation data obtained from the three sites is shown in Figure 21. As can be seen in Figure 21, the monthly precipitation totals for the Taylorsville RESUL TS describes the synthesis of the data to study the relationships among water cycle components. Salt Lake County was also available via the web http://www.pweng.slco.orglfloodlprecip lindex.cfm). A comparison of the precipitation data obtained from the three sites is shown in Figure 21. 35 _ 180.00 £ g | a W On-g °- Q. & $ & $ & & 0> & & 5? ^ ^ <f y ^ ^ ^ ^ ^ Rain Gauge Comparison Figure 21: Precipitation comparison of the three rain gauges. site were lower than those of the Salt Lake County site and the on-site gauge for November and December of 2007. These months contained the least amount of data gaps for the on-site gauge. The Taylorsville site recorded higher precipitation values than the on-site gauge for the months of January, February, and March, which were months with high amounts of data gaps. The on-site gauge did not record any precipitation for September and October 2007 and July and August 2008. In fact, July and August 2008 had large amounts of data gaps. However, when comparing the Taylorsville gauge with the Salt Lake County gauge, the data indicate that monthly precipitation totals depend highly on the location. The yearly totals are similar with the county gauge recording a total of 390 mm and the Taylorsville gauge recording 410 mm. Since the Salt Lake County rain gauge was located closest to the site, the county precipitation data was used for comparison with the runoff and soil moisture data when analyzing the water budget. It should be noted that although the data obtained from the county site were relatively 210.00 _ 180.00 • Taylorsville E • SL County §. 150.00 :WOn-Site 0c 120.00 ~ ·JC!!o 90.00 ·u 60.00 .C..I.) a.. 30.00 0.00 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ , ~ ~ (f ~ ":sf ~ ~ ;;\ ~ e:,1lJ~ OV ~o <;)1lJ )'If «,10 ~'lf ~~ ~fl) )V Date ofthe of2007. the Salt Lake County gauge, the data indicate that monthly precipitation totals depend highly on the location. The yearly totals are similar with the county gauge recording a total of 390 mm and the Taylorsville gauge recording 410 mm. Since the Salt Lake County rain gauge was located closest to the site, the county precipitation data was used for comparison with the runoff and soil moisture data when analyzing the water budget. It should be noted that although the data obtained from the county site were relatively 36 Maximum Daily Precipitation Temperature (°C) (mm) 10/12/2007 Data error Data error Table 3: Storms with significant precipitation (over 12.7 mm). 9/29/2007 15.0 12/1/2007 17.5 12/7/2007 23.4 12/8/2007 13.5 1/21/2008 13.2 4/24/2008 14.7 18.5 6/4/2008 18.8 8/31/2008 20.1 Total: 154.7 void of errors, there were 5 days as shown in Table 2 that contained data gaps or errors and were excluded from the overall dataset. The monthly precipitation (Figure 21) peaked in December at the county site and in July at the Taylorsville site. The peak at the Taylorsville site is due to isolated July and August thunderstorms. Shown in Table 3 is the total precipitation for the nine largest storm events between September 1, 2007 and August 31, 2008. As can be seen, there were three storms in December 2007 that fell into this category, all receiving more than 12.7 mm of precipitation. These nine storms accounted for 40% of the total annual precipitation amount. Table 2: County Site dataset summary of errors. precipitation amount. Date 10/13/2007 10/14/2007 2/29/2008 4/8/2008 (DC) Data error Data error 0.00 Data error Data error Data error Data error Data error Date 5/26/2008 % of Total Yearly Precipitation: Precipitation (mm) 40% 37 4.2 Flow Rates The storm water flow meter was installed to observe wet weather and dry weather flow rates. After installation, it was determined that adjustments were needed to correct the low flows associated with dry weather flow. The Stingray area-velocity flow meter requires a minimum depth of approximately 1 inch, but the dry weather flow from the small catchment area produces a depth less than an inch. An experiment was needed to calibrate the flow meter using manually collected observations of the low flow rates. To adjust the sensor measurements to obtain the flow, Manning's equation (eq. 4.1) was used: where v is the velocity, Rh is the hydraulic radius, S is the slope of the pipe, and n is the manning coefficient corresponding to the roughness of the pipe. This equation and method is used assuming the pipe remains only partially full. Experiments were performed using a calibrated bucket and stopwatch to measure the flow on two separate occasions. At the same time the flow was being measured, real-time velocity data was being recorded from the flow meter. A summary of the recorded data is shown in Table 4 The depth shown in Table 4 was determined using the flow and velocity values that were measured in these experiments and by applying eqs. 4.2 and 4.3. v = -R„2,iS Equation 4.1 n <9 = 2cos_ 1[l-2(^-)] d Equation 4.2 Area (0-sin(0))* - Equation 4.3 inch, but the dry weather flow from the - ~R 2/3SI / 2 v- h n 4. B = 2 cos -I [1- 2( Y )] d2 = B-sin(B))*- 8 38 Table 4: Storm water flow meter experiment results. Trial Time (sec) Volume (L) Velocity (m/s) Depth (m) 1 7 7.571 1.082 0.411 0.021 2 7 7.571 1.082 0.416 0.021 3 8.28 7.571 0.914 0.411 0.018 4 8.37 7.571 0.905 0.389 0.019 5 7.39 8.517 1.153 0.421 0.021 6 6.99 8.517 1.218 0.416 0.022 7 8.34 8.517 1.021 0.431 0.019 00 7.06 7.571 1.072 0.426 0.020 14 27.25 6.200 0.228 0.138 0.015 15 29.94 6.500 0.217 0.141 0.014 16 28.56 5.500 0.193 0.126 0.014 17 29.62 5.700 0.192 0.123 0.014 18 27.06 5.200 0.192 0.126 0.014 19 30.62 5.600 0.183 0.121 0.014 In these two equations, y is the depth of the flow, d is the diameter of the pipe, which in this case is 0.4572 m (18 inch), and e is the angle between the edges of the flow line and the center of the pipe. Equation 4.2 was inserted into eq. 4.3 and multiplied by the velocity to determine the flow as shown in eq. 4.4. This equation was then used to determine the depth of flow by inserting the flow and velocity values obtained through experimentation and solving for the depth. The results of these trials are shown in Figure 22. Flow = (2 cos- 1 [1 - 2( y )] - sin(2 cos- 1 [1 - 2( y )])) * * v Equation 4.4 The equation and corresponding adjustment appears to be reasonable given the depth output by the equations at the time of experimentation; thus, it was applied to the recorded flow data when the flow depths were insufficient. Flow (L) (Us) 8 = 2cos-I[I-2( y )]-sin(2cos-I[I-2( y )]))*°.4572 2 *v 0.4572 0.4572 8 39 0.025 • £ 0.015 0.02 Q. <3 0.01 0.005 Adjustment Factor - Linear (Adjustment Factor) 0 0 Manning's Flow Adjustment Figure 22: Depth adjustment for storm water flow. The wet weather flows and dry weather flows were recorded continuously from June 22, 2007 to August 30, 2008. A summary of the weekly outflow is shown in Figure 23 and the monthly runoff totals are shown in Figure 24. The weekly runoff totals indicate that the weeks with the highest runoff totals are the weeks from May 13, 2008 to May 30, 2008, the week of December 7, 2007, and the week of January 4, 2008 with the highest total being the week of May 30, 2008. However, the monthly totals show that February had the highest total runoff which suggests the importance of snowmelt runoff on the water budget of the Murray catchment. ~ 0.015 .-s::. Q) c 0.1 y = 0.0196x + 0.012 R2 = 0.917 ~ • Adjustm-ent Factor ~ - ,- - 0.2 0.3 0.4 Velocity (m/s) 0.5 0.6 13,2008 highest total being the week of May 30, 2008. However, the monthly totals show that February had the highest total runoff which suggests the importance of snowmelt runoff on the water budget of the Murray catchment. 40 - o ^ 1.1.1. hi. . | i J J I J 111 I I I hlluhllhhhkli v r f F rfS^ r& vO^ ^ <fT r,^ rfc eft r # r # r # ^ ^ r# r # r # r # rA> crt> 4r 4r 4r K ^ y ^ ^ & ^ ^ ^ <3> <r ^ ^ Date Runoff Totals Figure 23: Weekly runoff totals. o o o 500 400 300 • • $ $ & & & si <S° <3P 53= Runoff Totals Figure 24: Monthly runoff totals. 4.3 Soil Moisture Figure 25 shows the differences in soil moisture between the nonirrigated site P.C. to the lawn and garden irrigated areas at site P.H. The lawn area of site P.H. maintains high moisture content, whereas the garden area appears to be losing moisture. It is of interest that the nonirrigated site increases in moisture content, rising to the level of the drip line irrigation soil moisture. The drip line irrigation was shut off in late September, so this would lead one to believe that the soil moisture response of areas with §o ~ 150 o 100 ::J 250 00 200 -.0.. .. 150 ~ 0 100 LL ] 50 0 I- 0 Weekly 500 :r 0 0 400 0 ,~ ~ 300 ~ D~ u0:: : 200 - S 100 I0- 0 , 0 ,0, t:;:') b f;)'b r:;)'b r:;)'b t:;:)'b ~ :sf ~ ~ _\' )'li «ei ~'li '?"~ ~~.. .. Date Monthly c. 41 30.00% 25.00% $ 20.00% c S 1 CD 1 5 5.00% go 10.00% S ^ 5.00% (0 0.00% -5.00% - Site PH Garden Drip Line Site PH Lawn Sprinkler Irrigated - Site PC Nonirrigated 9/14/2007 9/22/2007 9/30/2007 10/8/2007 10/16/2007 10/24/2007 11/1/2007 Date Soil Moisture Content September to October 2007 Figure 25: Soil moisture content at sites P.C. and P.H. from September through October 2007. similar soil types and vegetation, when irrigation is not present, will be nearly the same. Figures 26-28 show the results of the soil moisture content for the spring and summer months of 2008. Figure 26 shows a comparison of the two lawn irrigated sites at site P.C. with the lawn irrigated area at site P.H. just a couple blocks away. As is shown, there is a jump in soil moisture for sensor PHM (P.H. Middle) and a slight jump in moisture at site PCE (P.C. East) near the first part of June. This may indicate either a change in watering practices or a storm event. In either case, the soil moisture level seems to remain high for sensor PHM, whereas the other two sensors indicate a continued decline in soil moisture. All lawn areas are mostly in open, sunny areas with the only shade coming in the morning and late afternoon hours at site P.H. The main difference when observing the two sites is that the lawn at site P.H. remained green throughout the summer months, whereas the lawn at site P.C. was dry (refer to Figures 16 - 18). Negative soil moisture values are assumed to have zero moisture content or represent an oven-dry state rather than an error in the recorded data. .. ji I: 8 15.00% ~ ~ ·0 ::i ·0 I/) --. ,.-. .. ... ....... -- /#'/""- ' . .- _ .... -_ ...... - ...... -_ ... .--- ............ .. ' . , flo .. ...... ' ............ - ..... ... ... .. ... '" ... ... - - - - - - --I I -- I -t-- I 26 - 28 ofthe of2008. P .change in watering practices or a storm event. In either case, the soil moisture level seems to remain high for sensor PHM, whereas the other two sensors indicate a continued decline in soil moisture. All lawn areas are mostly in open, sunny areas with the only shade coming in the morning and late afternoon hours at site P .H. The main difference when observing the two sites is that the lawn at site P .H. remained green throughout the summer months, whereas the lawn at site P.C. was dry (refer to Figures 16 - 18). Negative soil moisture values are assumed to have zero moisture content or represent an oven-dry state rather than an error in the recorded data. Bcc oo 1C_D 3 " 5 "o 55.0% 35.0% 15.0% 5.0% -5.0% PCWLawn -J 42 F J? 4J?? ^J ? ^J ? 4J?? J? Figure 26: Soil moisture comparison of irrigated lawn areas at both sites P.H. and P.C. c $ c O o ok_ 3 4-" (0 O O 0 ) 6> PHG Shaded, Nonirrigated PHG Sunny Garden 2 Drip Line 4 • A - / . N * <c> Figure 27: Soil moisture comparison of irrigated areas at site P.H. 75.0% - 65.0% c: -Q) c: 0 0 45.0% .Q..) -~ .!!! 0 25.0% ::E 0 en PHM Lawn Sprinkler Irrigated PCW Lawn Sprinkler Irrigated - - • PCE Lawn Sprinkler Irrigated ----~/~ ~'-------------­("" --.J ';:)q;, ';:)q;, ';:)q;, ",CPS ",CPS ",CPS C(}fJ, '\ \~ '\ f.' Date p.e. 78.0% - 68.0% c: oS 58.0% c: 0 0 48.0% -~~ 38.0% .!!! 0 28.0% :IE '0 18.0% en 8.0% -2.0% - - - . PHM Lawn Sprinkler Irrigated - - • PHM Sunny Garden 1 Drip Line It • - "I "\ • PHG SunnyGarden 2 Dlrip ~Lin~ ' • \.-', y y'' "/ \. ..., .',.. ", j ',J. .. .1'01 ../J ''\. ... .• Date Soil Moisture Comparison at Site PH -I 43 PHM Sunny Garden Figure 28: Soil moisture comparison of nongrassy areas for both sites P.H. and P.C. A comparison of all four soil moisture sensors for spring and summer of 2008 at site P.H. is shown in Figure 27. The tree and garden areas were irrigated by drip line irrigation, whereas the lawn area was irrigated using traditional lawn sprinkler methods. The magnitude of the difference is fairly consistent over the observation period. The variability in soil moisture is clearly the highest for the irrigated site, beginning with a low value of approximately 5% to greater than 50%. Figure 28 shows only the areas irrigated by drip line or that are otherwise nonirrigated. As can be seen, the soil moisture content of most of these sensors shows similar patterns of declining soil moisture content. In fact, some of the data dip into negative values, which are again an indication of an oven-dry state. One of the sensors located in the sunny area of the garden at site P.H. shows a rising trend in soil moisture, which could indicate that the drip line irrigation is affecting the moisture content of the area. -c: .s c: o o -e:::s .~ o ~ o en 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% ---PHMSunnyGarden 1 Drip Line --PCW Nonirrigated - • PCE Nonirrigated -PHG Shaded, Nonirrigated - - • . PHG Sunny Garden 2 Drip Line ~co ~CO ~CO ~CO ~~ ~~ ~~ ~~ ~Cl, '\ \~ '\ ~ ~~ Date fact, some of the data dip into negative values, which are again an indication of an oven-dry state. One of the sensors located in the sunny area of the garden at site P.H. shows a rising trend in soil moisture, which could indicate that the drip line irrigation is affecting the moisture content of the area. 44 Table 5: ET compared with other common climate variables. Date Total ET (1000 L) Average Temperature PC) Precipitation (1000 L) Runoff (1000 L) Average Non-irrigated Soil Moisture Content) Average Daily Outdoor Water Use (1000 L) 25-Sep-07 53.59 10.26 0 0.544 12.01% 463 26-Sep-07 85.77 12.87 0 0.305 11.52% 463 1-Oct-07 46.17 10.80 0 0.043 10.38% 207 4.4 Evapotranspiration ET data was available for 3 days in the fall of 2007; other days were not available due to sensor and logging errors. The data was acquired from Dr. Eric Pardyjak's research group and converted to a volume of water (see Chapter 3). A summary of the 3 days along with other common climate variables for those days is shown in Table 5. The water use for these 3 days was taken as an average day given the water usage rates of the entire month. Thus, the usage is not specific for the given day, but provides a reasonable value to use for comparison to ET. Given this data, however, it appears that temperature has the greatest impact on the ET rate with the higher ET rates being recorded on the day with the highest average temperature, although much more data is needed to have more confidence in this relationship. 4.5 Water Use Murray city monthly water use billing records were used to obtain outdoor irrigation for the summer months (between May and October) for the years 1998 to 2005, excluding only October of 2005. These records contain data from 742 lots of varying sizes ranging from 0.012 ha (0.03 acres) to 2.278 ha (5.63 acres) with an average lot size of 0.101 ha (0.25 acres). This average lot size was considered to be the average size of of2007; has the greatest impact on the ET rate with the higher ET rates being recorded on the day of2005. of 0.1 0 1 ha (0.25 acres). This average lot size was considered to be the average size of comQared Non- Preci pitation irrigated (0C) Content} {q 0ct-45 July October the lots within the catchment boundary; thus, the average water usage for the 742 lots was assumed to be the average usage for the study area. A summary of the maximum monthly outdoor water use for the study area as well as the average monthly outdoor water use is shown in Table 6. These values use the average household water use multiplied by 162, or the number of houses within the catchment boundary. As is shown, the highest maximum outdoor water use occurred in the month of July; however, August recorded the highest average water use totals, with both months being consistently dry months in the SLC metropolitan area, requiring irrigation to maintain healthy landscaping. One conclusion possible from this preliminary analysis may be the response of people to climate being observed in the water use pattern. 4.6 Processes and Relationships In the previous sections, the seasonal variability of the water cycle fluxes and stores was presented. In this section, the water cycle components will be analyzed in combination to ascertain relationships, quantify influences, and to address the following research questions: • What is the seasonal variation of precipitation-to-runoff conversion? Table 6: Outdoor water use monthly totals. Maximum Monthly Average Monthly Outdoor Water Use Outdoor Water (1000 L) Use (1000 L) landscaping. One conclusion possible from this preliminary analysis may be the response of people to climate being observed in the water use pattern. quantifY runoff May June August September 9057 20987 25236 23822 17438 9283 4597 13980 21197 21293 13886 6405 46 • What is the seasonal variation of irrigation waste flows and is it related to common climate variables? • What is the soil moisture response to rainfall and irrigation for different land cover/sun exposure combinations? • What is the effect of soil moisture on ET? 4.6.1 Seasonal Variation of Precipitation to Runoff Conversion Runoff coefficients were determined for monthly precipitation and runoff throughout the year to observe the differences between seasonal precipitation-to-runoff conversion, which may lead to the observance of antecedent soil moisture conditions in the urban environment as a contributing factor to the runoff coefficient. The stormwater outflow measured with the stormwater flow meter was compared to the rainfall data for monthly precipitation event totals in order to compare seasonal runoff coefficients for the area (Figure 29). The runoff totals exclude the runoff flow during times without precipitation, or dry weather flow periods. It was observed that there is a significant seasonal difference between the monthly runoff coefficients. As a comparison, runoff coefficient design values for the lot size in this residential area are normally assumed to be between 0.30 and 0.50. The month of February, as shown, had the highest runoff coefficient of 0.094, which is much less than the expected runoff coefficient values. It is assumed that February has the highest coefficients due to rising temperatures which impact the runoff by causing additional snowmelt at a time when the ground is frozen, preventing infiltration into the soil layer. The spring and autumn months are shown to have similar runoff coefficients, whereas the hottest months of July and August have storm water seasonal difference between the monthly runoff coefficients. As a comparison, runoff coefficient design values for the lot size in this residential area are normally assumed to be between 0.30 and 0.50. The month of February, as shown, had the highest runoff coefficient of 0.094, which is much less than the expected runoff coefficient values. It is assumed that February has the highest coefficients due to rising temperatures which impact the runoff by causing additional snowmelt at a time when the ground is frozen, preventing infiltration into the soil layer. The spring and autumn months are shown to have similar runoff coefficients, whereas the hottest months of July and August have 47 0.100 0.090 § 0.080 o 0.070 £ 0.060 o 0.050 ^ 0.040 o 0.030 = 0.020 °t 0n .0r1u0n 0.000 j £ ^ & & & & & & j £ # Oo V *° < / ^ ^ ^ * i& Figure 29: Seasonal variation of monthly runoff coefficients. extremely low runoff coefficients due to the antecedent soil moisture conditions which allow for rapid infiltration as the precipitation reaches the ground. The low values of coefficients could signify that since the catchment has such a minimal slope over the majority of the catchment, much of the water may pool and evaporate or infiltrate instead of being directed into the storm drainage system. Thus, the amounts of runoff may be an underestimation of what is actually lost downstream or back into the atmosphere; however, further analyses shown in Figure 29 may assist in answering this question. Runoff coefficients were determined for individual storm events to observe the variation between events and to compare these events to coefficients commonly used in practice. The storm event runoff coefficients are listed in Table 7. The runoff coefficients are higher in the warmer months and correspond to the monthly runoff coefficients; however, the runoff coefficients for the winter months are highly variant (between 0 and 0.033) which is a clear indication of snowfall remaining either on-site, or being converted to runoff in the days following the storm event as temperatures increased. _ 5i 'u IE o~ It: 0.040 § a: 0.010 0.046 0.094 0.051 <;:)'\ <;:')Q t:::)'Q t:::)'Q t:::)'Q <;:)'Q ()'IF ~~« «q1 ~f ~~" ~~~ Date Runoff coefficients 0.005 being converted to runoff in the days following the storm event as temperatures 48 Table 7: Runoff coefficients for major storm events. (1000 L) (1000 L) 9/29/2007 2652.5 108.0 0.041 0.2 18.4% 12/1/2007 3102.1 4.3 0.001 0.2 12/7/2007 4136.1 73.0 0.018 0.2 8.0% 12/8/2007 2382.8 78.0 0.033 0.2 1/21/2008 2337.8 0.0 0.000 0.2 4/24/2008 91.0 0.035 0.2 5/26/2008 3281.9 132.0 0.040 0.2 18.2% 6/4/2008 3326.9 161.0 0.048 0.2 21.9% 23827.7 647.3 0.027 0.2 12.3% One method normally used in design to determine a runoff coefficient is shown in eq. 4.5 below. This method uses the percentage of impervious area to estimate the runoff coefficient for the study area. C 0.858*/3 -0.78*/2 +0.774*/+ 0.04 Equation 4.5 The total impervious area was found to be nearly 25%, assuming the majority of the rooftops and sidewalks were not directly connected impervious areas or connected to the storm drainage system. Using eq. 4.5, this resulted in a runoff coefficient of 0.22 for the study area. This is higher than the runoff coefficients computed from the rainfall and runoff data (Table 7), but within reason. Although the storm events with their corresponding runoff coefficients shown in Table 7 were the largest rainfall events recorded during the study period, with a maximum rainfall depth of 18.8 mm, the 2-year 24-hour design storm is normally double that amount, or roughly 40 mm. It can thus be expected that these small storms do not produce a significant amount of runoff, accounting for the differences between the measured runoff coefficients and the estimated coefficients. Also, due to antecedent Estimated Percent Date Precipitation Runoff Runoff Runoff Difference Coeficient Coefficient (Actual to Estimated) 0.6% 14.8% 0.0% 2607.6 15.8% Total: = 0.858* i3 0.78* i2 0.774* i +runoff data (Table 7), but within reason. maximum rainfall depth of 18.8 mm, the 2-year 24-hour design storm is normally double that amount, or roughly 40 mm. can thus be expected that these small storms do not produce a significant amount of runoff, accounting for the differences between the measured runoff coefficients and the estimated coefficients. Also, due to antecedent 49 moisture conditions, the runoff for a small catchment such as this becomes difficult to estimate. 4.6.2 Dry Weather Flows Capture of the dry weather flow in the storm drain was observed and can be seen during periods without any precipitation and where significant flows were recorded. Since the catchment was isolated from additional base flows or precipitation events further upstream, it was hypothesized that irrigation of impervious surfaces (roads and driveways) adjacent to landscaped areas would produce a measurable response in the storm drainage system. The timing of the dry weather flow also suggests irrigation waste flows are a major contributor to this, but other flows including infiltration or inflow (possibly again caused by landscape irrigation) may also play a role. Nearly 73% of the total discharge volume from the catchment during the month of August 2007 was dry weather flow. For the year of record, approximately 8% of the total annual runoff volume is from dry weather flow. This is comparable to 10-30% of the annual flow volume found for dry weather flows in the Ballona Creek watershed in Los Angeles, California (McPherson et al., 2002). The total outflow for these dry periods was compared to the total monthly water use for the summer months. As shown in Figures 30 and 31, the dry weather flow is a small fraction of the outdoor water use. Although not massive, the small irrigation waste flows for this small catchment account for approximately 225,000 L of discharge for the year, which is over half of the amount used in the entire catchment in one day for outdoor irrigation on average for the months of June and September. If this trend is similar across the entire SLC metropolitan area, the amount then becomes roughly 2,400 million liters of wasted water from inefficient during periods without any precipitation and where significant flows were recorded. Since the catchment was isolated from additional base flows or precipitation events driveways) adjacent to landscaped areas would produce a measurable response in the storm drainage system. The timing of the dry weather flow also suggests irrigation waste flows are a major contributor to this, but other flows including infiltration or inflow (possibly again caused by landscape irrigation) may also playa role. Nearly 73% of the total discharge volume from the catchment during the month of August 2007 was dry weather flow. For the year of record, approximately 8% of the total annual runoff volume is from dry weather flow. This is comparable to 10-30% ofthe annual flow volume found for dry weather flows in the Ballona Creek watershed in Los Angeles, California (McPherson et ai., 2002). The total outflow for these dry periods was compared to the total monthly water use for the summer months. As shown in Figures 30 and 31, the dry weather flow is a small fraction of the outdoor water use. Although not massive, the small irrigation waste flows for this small catchment account for approximately 225,000 L of discharge for the year, which is over half of the amount used in the entire catchment in one day for outdoor irrigation on average for the months of June and September. If this trend is similar across the entire SLC metropolitan area, the amount then becomes roughly 2,400 million liters of wasted water from inefficient 50 0.03% 0.16% 0.43% 0.27% • May • June • July • September • October 0.38% Monthly Irrigation Flow Wasted Compared with Average Outdoor Water Use Figure 30: Monthly irrigation flow compared with average outdoor water use. 0.13% 0 0 2 ° / o 0.22% • May • June • July • August • September G9 October Monthly Irrigation Flow Wasted Compared with Maximum Outdoor Water Use Figure 31: Monthly irrigation waste flow compared with maximum outdoor water use. landscape irrigation. 4.6.3 Soil Moisture vs. Irrigation and Precipitation Figure 32 shows the soil moisture response to precipitation. The only sensor that appears to be impacted by the precipitation is the nonirrigated sensor at site P.C., shown previously in Figure 27. There appeared to be a major storm event or change in the total water entering the lawn area, causing the soil moisture to increase from approximately 8% to more than 50%. Although this sensor indicates a response to precipitation, as 0.38% 0.31% f21May . .July • August mSeptember Ii!! October 0.02% 0.22% 0.24% 0.32% 0.21% fa May rn September Ii!! October e., 51 16 14 12 J 10 = 8 | 6 4 2 • Precipitation - - - Site PH lawn sprinkler irrigated I Site PH garden drip line Site PC non-irrigated 30.00% 25.00% 15.00% 20.00% f O £ 3 .2 So 5.00% 0.00% -5.00% 9/14/2007 9/24/2007 10/4/2007 10/14/2007 10/24/2007 Date Soil Moisture Content vs Precipitation Figure 32: Soil moisture response to precipitation from September through October 2007. shown in Figure 33, it appears that irrigation patterns have a greater impact on the soil moisture variability than do precipitation totals. Judging the soil moisture response based on the precipitation timing, the sensor located in the lawn and the sensor located in the shady part of the garden are the only sensors on which the precipitation seems to have a measureable impact. The remaining sensors do not indicate any impact from precipitation. There were significant storm events during the last part of May and the first part of June that are believed to have caused this increase in soil moisture content for the lawn area; however, in order to sustain the high soil moisture content, and based on the remaining sensors, high irrigation amounts were required to achieve such a high moisture content and to sustain it throughout the hot summer months. Since the lawn was extremely healthy during the hot summer months, it may be concluded that this location was perhaps overwatered. --Precipitation - - - - - •• ;-., ... / ... - .................................. - -_ ... - - ... - - - ... --- ... ..... E .§. i c - -, 'cu 6 DC 4 0 " . ---0,--------71 -~ ---- ----- I-,-------,-+-I ,u'UL-tLW - -c ~ o ~... .a 10.00% .!!! o :E 5.00% '0 II) 52 Drip - S i t e Drip Figure 33: Soil moisture response to precipitation for site P.H. Figure 34 presents similar results for site P.C. with precipitation making only minimal impacts to the overall soil moisture content. Figure 35 shows more explicitly, at a more refined time scale, that even though there were significant precipitation totals the week of September 22, 2007, this did not cause a significant change in soil moisture. There is, however, a regular pattern shown by the soil moisture sensor indicating the effect of drip line irrigation on soil moisture, which can be easily observed. Figure 36 shows the soil moisture response to rainfall for a select period near the end of May and the beginning of June 2008. This period included two of the largest rainfall events of the year. The only sensor indicating a response to the rainfall was the shaded garden area, which rose about 2% and tapered back off similar to the other sites. Overall, the only two locations that were able to maintain a high amount of moisture were lawn irrigated areas. Otherwise, the garden areas and nonirrigated areas whether shaded or unshaded generate the same moisture patterns. It could thus be concluded that irrigation has a much higher impact on the resulting soil moisture content of an area than do precipitation events. E .§. :! c ·iii IX: --Precipitation --Soil Moisture PH Lawn - - Site PH Garden 1 Drig Line Site PH Garden 2 Drip Line - - - Site Shaded Garden rip Line -,----------. --------------- 70% - 60% c S 50% c 0 40% u Q) 30% ... -::J 20% III ·0 10% :E 0% 0 -10% (f) Date Precipitation vs Soil Moisture Site PH e. the same moisture patterns. It could thus be concluded that irrigation has a much higher impact on the resulting soil moisture content of an area than do precipitation events. 53 1 Precipitation 18 16 14 12 10 8 6 4 2 0 _ _._ Soil Moisture PCE Nonirrigated - - - Soil Moisture PCW Nonirrigated Soil Moisture PCE Lawn PCW 4 Date Precipitation vs Soil moisture site PC Figure 34: Soil moisture response to precipitation at site P.C. | 1 3 20% £ 15% | 10% | 5% Z 0% o -5% h E E 8 infall 6 Ra 4 2 0 A * • Precipitation - - - - - Soil Moisture Site PH Garden Drip Line ! V r.'b- rib" . \ J .NJ XJ (b- r(b- nb- V AV AV AO-" # qfr' q>v Time 5.00% Figure 35: Soil moisture content vs precipitation for site P.H. garden in September 2007. ---- - peE ---peE Soil Moisture pew Lawn - Soil pew 20 40% 35% C . 30% ~ 14 25% 8 12 200/0 f 10 - 15% ~ ~ I I I 10%'0 4~_ _~........ 5%~ 2 - .. - .. - --.- .. ------------ o%:~ --_----- VI o -, , ~~, ~-~---_T----_r~--~~--_r----_r--~ -5% f;)CO RJCO ~~ ~~ A\'"' rV . \ '\,\ 12 10 --Precipitation - - Soil Moisture Site PH Garden Drip Line -...... Date & Soil Moisture Content vs Precipitation 25.00% -c 20.00% S c o 15.00% C..; .a 10.00% .!!! o ~ '0 (fJ 0.00% 54 Site PH Garden 1 Drip Line - - - Site Shaded Garden Drip Line Site PCW Nonirrigated Site PCE Nonirrigated 10% ^ £ £ i Figure 36: Nonirrigated soil moisture vs. precipitation for isolated time frame. 4.6.4 Soil Moisture and ET Figures 37 and 38 suggest a relationship exists between soil moisture, ET, and landscape irrigation. Although the outdoor irrigation flux was not measured, there was no precipitation on this particular day; thus, it can be concluded that since the soil moisture content increased, outdoor irrigation caused the increase. Also, the ET rate increased at times of presumed outdoor watering (early morning and late afternoon), resulting in a portion of the outdoor irrigation being returned to the atmosphere. For this particular date, based on average water use per day in September, 19% of the irrigation was lost to the atmosphere. Also, since the soil moisture content was relatively high, there was a sufficient supply of water from which to draw to evapotranspirate. Thus, a much higher percentage of outdoor irrigation used is lost to ET when compared to the percentage lost downstream via storm drain outflow. --Precipitation Site PH Garden 2 Drip Line --- - Drig - rip --20 10% _ 18 16 E 14 .§. 12 :! 10 c: 8 .iij 6 0:: 4 2 0 r::::,~ R) cl-rr; ~rr; ~""' .... --- - - .",,---"-"-~ Date Precipitation vs Soil Moisture Site PH -~ 8% 6% 4% 2% 0% -2% c: S c: 0 (,) I!! -:l UI ·0 ::::E ·0 en 55 Figure 37: Nonirrigated soil moisture content vs ET rate response for September 26, 2007. 25.00% 24.80% 24.60% nt 24.40% co u 24.20% <D i_ 3 24.00% -*-' </> O 23.80% E 23.60% Soi 23.40% 23.20% 0- # of hours on September 26, 2007 ET rate vs Soil moisture Figure 38: Irrigated soil moisture content compared to the ET rate response for September 26, 2007. ::J_ 0 II) o Q) 0:; ~.= -Q) E ~o t- M w 7 6 5 4 3 2 1 0 -1 -2 \:) ~ -, - '-.. -\,' - - - ET ---Nonirrigated ,~ I , , •, \' \ # of hours on September 26, 2007 ET rate vs Soil moisture 11 .90% 11 .80% -c 11 .70% .s c o (J 11 .60% E t -:::l 11.50% .~ ' .. \ ",' J\ 11.40% i " 11 .30% -- · 11 .20% 7 6 5 ::J 4 0 U) o Q) 0._.. .. -:C:: l 3 .s 'E 2 ~o t- M w 0 -1 -2 \:) - - - ET ---Irrigated I ~ 4 , .. A , '. I f , ' ,, I " # -. 1 ..... --\,' - --/ T 25.00% r C r 24.40% .8s \ t 24.20% ~ I 24.00% iii \ .,~ , 23. 80%·-~ , • .,J"" 23.60%'0 \' (/) • 23.40% 56 4.6.5 Murray Catchment Water Budget The overall water budget for the year between September 1, 2007 and August 30, 2008 is shown in Figure 39. This shows similar results to those shown in Figure 4 from Chapter 1. The ET rate is shown to be the governing factor in the overall water budget in the SLC metropolitan area. Cleugh et al. (n.d.) obtained similar results, determining that the ET rate was the greatest component of the water budget; however, they determined that there were additional benefits obtained from the high amount of water used to sustain the ET rate such as lower temperatures and energy savings. This was not the case in this thesis study as the air temperatures recorded were higher than normal values in the summer and lower than normal in the winter (see Appendix A). It should be noted that the water use in the urban area contributes to the large amount of water available for ET; however, both Figures 4 and 39 indicate that the ET rate is possibly an over-estimation due to actual amount of water available. 0 Volume of Water (1000 L) Murray Water Budget Figure 39: Urban water budget for Murray Catchment. l 2800 Water Out -----------------r Water In o 50000 Estimated Evapotranspiration 81400 e 100000 150000 Dry & Wet Weather Flow c::J Low Uncertainty c:::J Medium Uncertainty c:::J High Uncertainty 200000 250000 CHAPTER 5 SUMMARY AND CONCLUSIONS The observations and results have indicated that the fluxes and stores in the urban environment are extremely complex, interrelated, and are influenced by several factors. One of the factors influencing several water cycle components is outdoor water use. Outdoor water use was clearly shown to influence soil moisture and was shown to possibly influence ET rates. Outdoor water use and ET are the two most significant fluxes based on volume in the SLC metropolitan area and the case study residential catchment in Murray, Utah. Based on average data values for the area, September 2007 through August 2008 had slightly below average precipitation. The monthly runoff totals indicate that February generates the highest total runoff, which suggests urban snowmelt contributes significantly to the runoff volumes from SLC metropolitan area watersheds. Water use records indicated people use more water to irrigate during the hotter parts of the summer - supposedly to maintain healthy landscapes. Interestingly, the high outdoor water use influences the other water cycle components. For example, soil moisture, ET, and possibly groundwater recharge all are increased due to the large amount of water use for landscape irrigation. The potential groundwater recharge taking place in the Murray catchment would not replenish the Murray water sources, but would help to replenish the groundwater resources beneath the SLC metropolitan area that do provide on average approximately 50% of the urban outdoor water use. CHAPTERS ofthe the summer - supposedly to maintain healthy landscapes. Interestingly, the high outdoor water use influences the other water cycle components. For example, soil moisture, ET, and possibly groundwater recharge all are increased due to the large amount of water use for landscape irrigation. The potential groundwater recharge taking place in the Murray catchment would not replenish the Murray water sources, but would help to replenish the groundwater resources beneath the SLC metropolitan area that do provide on average approximately 50% of the urban outdoor water use. 58 When there is a sufficient water supply from landscape irrigation, the greatest impact on ET in the urban environment appears to be the average daily temperature, with a linear relationship between average temperatures and daily ET rates based on the 3 days of available ET data. It was also shown that soil moisture content of similar areas will generate the same patterns of moisture content when not impacted by human interaction. The seasonal variation of precipitation-to-runoff, expressed as runoff coefficients, suggests the coefficients calculated are significantly smaller than those used in practice. The largest runoff coefficients were computed for the month of February, due to the significant snowmelt occurring at that time. It was further observed that the hottest months of July and August yielded little runoff due to the antecedent soil moisture condition, which allowed for rapid infiltration. The rainfall storm totals were much lower than those that are normally used to design storm drain components, and it is assumed that the runoff coefficients were impacted due to significant depression storage that was able to limit the amount of runoff observed. The seasonal variation of dry weather flows was shown to increase as temperatures increased due to human response to increased temperatures. This logically leads to the conclusion that dry weather flows were predominantly comprised of irrigation excess. When comparing the dry weather flow volume to the total outdoor water use volume, the percentage lost is small, but may be significant over long periods. It was found that irrigation waste accounts for 8% of the total annual runoff from the catchment or about 225,000 Liters lost downstream. Assuming this same trend of waste occurs over the entire Salt Lake Valley, roughly 2,400 million Liters of water is wasted or lost downstream annually. of available ET data. was also shown that soil moisture content of similar areas will generate the same patterns of moisture content when not impacted by human interaction. to-runoff, significant snowmelt occurring at that time. was further observed that the hottest that the runoff coefficients were impacted due to significant depression storage that was able to limit the amount of runoff observed. irrigation excess. When comparing the dry weather flow volume to the total outdoor water use volume, the percentage lost is small, but may be significant over long periods. It was found that irrigation waste accounts for 8% of the total annual runoff from the catchment or about 225,000 Liters lost downstream. Assuming this same trend of waste occurs over the entire Salt Lake Valley, roughly 2,400 million Liters of water is wasted or lost downstream annually. 59 The soil moisture content was shown to respond more strongly to irrigation than to small rainfall events. Land cover and sun exposure were found to have only a small impact on soil moisture content. The only major difference was between lawn irrigated areas and all other locations, with the lawn areas having much higher moisture contents. The soil moisture theoretically contributes to higher ET rates by providing an available supply of water for the trees and vegetation. Landscape irrigation was suspected of increasing soil moisture, which corresponded to an increase in ET on one of the days ET data was available. Based on the volumetric ET and irrigation amounts, about 18.5% of the irrigation was lost to the atmosphere. Thus, a much higher percentage of outdoor irrigation is lost to ET when compared to the percentage lost downstream via storm drain outflow. There are many ways to extend the research initiated with this thesis. First, groundwater recharge is an important component of the water budget and in order to determine the recharge quantitatively through a water budget analysis, both the water use and the ET must be monitored more closely. Second, the problems with the ET monitoring equipment must be fixed to provide the ability to carefully analyze the relationships among landscape irrigation, soil moisture, and ET. Third, the continuous data collection is necessary to accumulate multiyear data for comparison. APPENDIX A TEMPERATURE The daily temperature was obtained from the same Salt Lake County site as were the additional precipitation data. These data were also monitored and checked for errors. There were only 5 days throughout the year, as was shown in Table 2, that did not contain accurate data. These days were thus excluded from the data set. Figure Al shows the maximum daily temperatures between September 2007 and August 2008, with the largest day-to-day swings in temperatures occurring in the spring and autumn months. July and August showed the least amount of temperature variation with temperatures remaining high for both months. To compare the temperature data with precipitation data, the monthly average maximum temperatures were extracted from the dataset. These temperature data were recorded at both the Salt Lake County site and the Taylorsville site, and are found in Figure A2. The seasons in Utah are fairly well defined based on the average monthly maximum temperatures with spring and autumn being intermediate temperatures when compared to the extremes of summer and winter. The values listed in Figure A2 are the temperature values for the County Site. Also shown in the figure are the average temperatures for Salt Lake City from 30 years of recorded data. Fall of 2007 and summer of 2008 are shown to have higher than average temperatures whereas the spring and maximum temperatures with spring and autumn being intermediate temperatures when temperature values for the County Site. Also shown in the figure are the average temperatures for Salt Lake City from 30 years of recorded data. Fall of 2007 and summer of 2008 are shown to have higher than average temperatures whereas the spring and 61 winter months of record had slightly lower than average maximum daily temperatures, resulting in larger annual swings in temperature. o 40.00 30.00 15.00 10.00 0.00 -5.00 -10.00 r8> 4? Maximum Daily Temperature Figure Al: Daily Maximum Temperatures from September 2007 to August 2008. 4 0 . 00 3 5 . 00 3 0 . 00 2 5 . 00 2 0 . 00 5.00 2 6 . 93 • County Site • Taylorsville Site • SLC Average 5.78 2.13 1.92 3 5 . 33 2 8 . 89 2 0 . 3 4 M * 33.22 of* O* ^ / ^ ^ ^ # Month Monthly A v e r a g e M a x i m u m T e m p e r a t u re Figure A2: Monthly Average Maximum Temperature from Salt Lake County Site. 40.00 l 35.00 30.00 l 25.00 20.00 1 5.00 . 5 .00 f;:)~ rV'D ,\\" Date Maximum Daily Temperature AI: -U ~ .Q..) -:::J .c.u. Q) Co E Q) I- 40.00 35.00 30.00 25.00 20.00 15.00 10.00 0.00 o 26.93 13 ~ 18.37 ~96 1469 ~~192 578~O~ I Tc::8J-,-- I I" 20.34 35.33 28.89 33.22 ~ .. ,~ .,~ ,~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ''-... ~ 'Sf ~' ~ ;:\ ~ ~ ~ 0(Q 0° ~o <::l ':,'l:f «(Q ~'l:f "(-~ ~'l:f 'loS 'loS "(-oS Month Monthly Average Maximum Temperature 62 APPENDIX B TEMPERATURE AND PRECIPITATION The temperature and precipitation events were analyzed to determine patterns between similar events. The number of occurrences of each category can be seen in Figure Bl. The breakpoints for the temperature were determined as the maximum daily temperature within the following ranges: • Cold: 4.44 °C and below • Cool: 4.45 °C-18.3 °C • Warm: 18.4 °C - 26.6 °C • Hot: 26.7 °C and above As the figure indicates, the majority of the precipitation occurs when the temperatures are in the "cool" category or when temperatures range between 4.45 °C and 18.3 °C; hence, spring and autumn months are usually the optimal temperature for generating the most precipitation throughout the year. This is surprising since it is usually thought that the majority of the precipitation occurs during the cold winter months. APPENDIXB PRECIPITA nON B 1. C - 18.3 • Warm: 18.4 °C - 26.6 °C • Hot: 26.7 °C and above DC; 100 tfl ay 80 a of 60 i_ a> 40 um 20 z 0 92 39 26 • 52 36 14 o*6 o<^ W e a t h e r Characteristics Temperature and Precipitation Figure Bl: Temperature compared to precipitation events. 64 100 92 93 1/1 >C\I- 80 -c 60 52 .0.. 39 36 QI 40 CL .c 26 E 20 14 10 ::::I Z 0 Q<A ~0\. Q<A ~0\. Q<A ~0\. Q<A ~0\. Vcr.~' Vcr.~' V0cr., V0cr., ~~~' ~~~' -(-O\.' -(-o\.' Weather Characteristics B 1: REFERENCES Berg, A.; Byrne, J.; Rogerson; R. An Urban Water Balance Study, Lethbridge, Alberta: Estimation of Urban Lawn Overwatering and Potential Effects on Local Water Tables. Canadian Water Resources Journal, 1996,21, 355-365. Claessens, L.; Hopkinson, C; Rasteter, E.; Vallino, J. Effect of Historical Changes in Land Use and Climate on the Water Budget of an Urbanizing Watershed. Water Resour. Res., 2006, 42, W03426. Cleugh, H.A.; Bui, E.; Simon, D.; Xu, J.; Mitchell, V.G. The Impact of Suburban Design on Water Use and Microclimate. http://mssanz.org.au/modsim05/papers/cleugh 2.pdf (accessed July 16, 2007) Envision Utah, http://envisionutah.org/toolbox pdf/Introduction.pdf (accessed December 11,2007) Grimmond, C.S.B.; Oke, T.R. Comparison of Heat Fluxes from Summertime Observations in the Suburbs of Four North American Cities. American Meteorological Society, 1995, 4, 872-889. Grimmond, C.S.B.; Oke, T.R. Evapotranspiration Rates in Urban Areas. Impacts of Urban Growth on Surface Water and Groundwater Quality. Proceedings of 10GO 99 Symposium HSS. Birmingham. July 1999. (IAHS. 1999, 259, 235-243.) Grimmond, C.S.B.; Oke, T.R.; Steyn, D.G. Urban Water Balance 1. A Model for Daily Totals. Water Resour. Res., 1986, 22, 1397-1403. Great Salt Lake Basin Hydrologic Observatory. Sustainability of Water Resources. http://greatsaltlake.utah.edu/ science/waterresources (accessed December 21, 2007) Havnes, M. Officials Hear Update on Plans for Pipeline Projects. The Salt Lake Tribune, August 9, 2008, p B3. Henetz, P. 7 Thirsty States Cut a Deal. The Salt Lake Tribune, December 14, 2007, ppAl,A8. " Henetz, P. Water Levels Way Up, but Dry Days Loom. The Salt Lake Tribune, July 12, 2008, pp Al, A5. 1996,21,355-365. Claessens, L.; Hopkinson, C.; Rasteter, E.; Vallino, J. Effect of Historical Changes in Land Use and Climate on the Water Budget of an Urbanizing Watershed. Water 2006,42, W03426. Cleugh, H.A.; Bui, E.; Simon, D.; Xu, J.; Mitchell, V.G. The Impact of Suburban Design papers/cleugh 16,2007) Envision Utah. http://envisionutah.org/toolbox pdf/Introduction.pdf (accessed December 11, 2007) Observations in the Suburbs of Four North American Cities. American 4, Grimmond, C.S.B.; Oke, T.R. Evapotranspiration Rates in Urban Areas. Impacts of Urban Growth on Surface Water and Groundwater Quality. Proceedings oflOGO 99 Symposium HSS. Birmingham. July 1999. (IAHS. 1999,259,235-1. Totals. Water Resour. Res., 1986,22, 1397-1403. Great Salt Lake Basin Hydrologic Observatory. Sustainability of Water Resources. http://greatsaltlake.utah.edui science/waterresources (accessed December 21, 2007) Havnes, M. Officials Hear Update on Plans for Pipeline Projects. The August 9, 2008, P B3. Henetz, P. 7 Thirsty States Cut a Deal. The Salt Lake Tribune, December 14, 2007, pp AI, A8. 12,2008, AI, Kunzig, R. Drying of the West. National Geographic, February 2008, pp 90-113. Leonard, W. National Expert on Growth Joins U. Architecture Faculty. Deseret News, July 2, 2008. http://findarticles.eom/p/articles/mi qn4188/is 20080702/ai n27887557/pg l?ta g=artBodv;coll (accessed July 2, 2008) McPherson, T.N.; Burian, S.J.; Turin, H.J.; Stenstrom, M.K.; Suffet, I.H. Comparison of the Pollutant Loads in Dry and Wet Weather Runoff in a Southern California Urban Watershed. Water Sci. Technol, 2002,45,255-261. Mitchell, V.G.; Mein, R.G.; McMahon, T.A. Modelling the Urban Water Cycle. Environmental Modelling and Software, 2001, 16, 615-629. Portoghese, I.; Uricchio, V.; Vurro, M. A GIS Tool for Hydrogeological Water Balance Evaluation on a Regional Scale in Semi-Arid Environments. Comput. Geoscl, 2005, 31, 15-27. Ragab, R.; Rosier, P.; Dixon, A.; Bromley, J.; Cooper, J.D. Experimental Study of Water Fluxes in a Residential Area: 2. Road Infiltration, Runoff and Evaporation. Hydrol. Processes, 2003,17, 2423-2437. Ramier, D.; Berthier, E.; Dangla, P.; Andrieu, H. Study of the Water Budget of Street: Experimentation and Modeling. Water Sci. Technol., 2006, 54, 41-48. State of Utah Natural Resources, Division of Natural Resources. Utah's Water Resources: Planning for the future, Utah State Water Plan. May 2001. http://www.water.utah.gov/waterplan/uwrpff/Chp-03a.htm (accessed December 11,2007) Summers, L. Incentive Pricing for Efficient Water Use in Urban Utah, Draft, January 14, 1999. Prepared for the Utah Water Conservation Forum, (Salt Lake City: Division of Water Resources, 1999), 7. Thomas, A.; Tellam, J. Modelling of Recharge and Pollutant Fluxes to Urban Groundwaters. Science of the Total Environment, 2006, 360, 158-179. USGS. 2004. Water Quality in the Great Salt Lake Basins Utah, Idaho, and Wyoming 1998-2001. National Water-Quality assessment Program. Circular 1236. Utah Division of Water Resources. Identifying Residential Water Use: Survey Results and Analysis of Residential Water Use for Thirteen Communities in Utah, (Salt Lake City: Division of Water Resources, 2000), 27. 66 comlp/articles/mi gn4188/1 ?artBody;1.; Techno!., 2002, 45,255-2001,16,615-VUITO, Geosci., 2005,31,15-Hydro!. 2003, 17,2423-Techno!., 2006,54,41-gov/waterplan/uwrpff/Chp-03a.htm 1. IdentifYing Utah Division of Water Resources. Utah's M&I Water Conservation Plan. Investing the Future. July 14, 2003. http://www.conservewater.utah.gov/Final71403AACC.pdf (accessed July 16, 2007) Utah Division of Water Resources. Municipal and Industrial Supply Studies. 2003. Xiao, Q.; McPherson, E.G.; Simpson, J.R.; Ustin, S.L. Hydrologic Processes at the Urban Residential Scale. Hydrol. Processes, 2006, 21, 2174-2188. 67 in utah.gov/FinaI71403AACC.pdf 2003. Hydro!. 2006,