Surveying and Projecting Sustainability and Urban-Water-Energy-Nexus Applications in Rhode Island

Rhode Island is both the smallest and 2nd most densely populated state, which already characterizes its unique situation within the United States of America. About 90.7% of the inhabitants live in urbanized areas, creating a more beneficial situation for the state’s cities and towns equates to establishing improved conditions for a majority of its citizens. This thesis constitutes a comprehensive approach on assessing sustainability in Rhode Island and its communities via implementation of a municipal ranking with 75 social, environmental and economic indicators. The rating is based on the best and worst performances of various indicators, thus allowing for concise comparison within the local context of the state. In this analysis, while the communities around Providence tend to perform unfavorably, the southeastern coastal communities are above average performers. The ranking results also show a certain link to both income and population density of the municipalities. The proposed tool allows for comprehensive evaluations and identification of areas for improvement for all municipalities of the state. The second research focus is to evaluate linkages between water and energy provision in the state with a distinct focus on the urban environment. Both water supply and power generation exhibit advantageous characteristics, but rely on adequate data gathering to enable more refined research approaches. In addition, interactions between these vital resources were assessed by evaluating pollution sources and urban heat island implications. The latter reveals a high share of people residing in areas with significantly increased temperatures, which results in considerable, potential benefits by mitigating the associated UHI. Accordingly, abatement thereof may increase resilience of the public water supply infrastructure.

In this introductory statement the general objective of the study along with associated research goals and challenges will be discussed. Additionally, the hypotheses, which will be scrutinized later in the paper, are explained.

Justification for the Study
Overall, this thesis aims to evaluate the communities of Rhode Island regarding their sustainability and assessing connections between energy and water procurement according to the corresponding urban environment. Therefore, the concept of sustainability and the importance of its ongoing progression will first be highlighted in section 1.1.1, where quantification approaches of this highly ambiguous term will be discussed as well. Next, nexus thinking, which generally describes joint management of interdependent resources, will be discussed in section 1.1.2, while particular challenges related to urban areas will be discussed in section 1.1.3. Furthermore, major upcoming challenges of the 21st century, such as urbanization, globalization and climate change, will be outlined in section 1.1.4. Lastly, sections 1.2 to 1.4 state the thesis objective, the hypotheses and the organization of the thesis.

Measuring Sustainability
Ongoing urbanization and climate change are among the key concerns of modern day politics as various scenarios predict a diversity of detrimental effects if these two trends are allowed to remain on their current trajectory. Furthermore, virtually all areas of everyday life such as health, well-being, the natural environment and secure access too food and water may be affected. As those concerns have already persisted for several decades, the concept of sustainability is often featured as a measure to potentially limit long term detrimental consequences. While its underlying principles have already been thought up centuries ago, political action and public interest have only very recently shifted in its favor with the oil crisis of 1973 being considered a starting point for environmental legislations in many different countries across the globe. [1] [2] A prominent example for the United States of America is the Energy Policy and Conservation Act from 1975, which was established to allow regulation of fuel prices to promote energy conservation. [3] According to this development, the most commonly used definition of sustainability, which is described as the ability to provide for one's needs without compromising future generations to do so, has been defined during the 1987 United Nations (UN) sessions on establishing a global agenda for change. Additionally, according to Portney even this early iteration of modern conferences on joint environmental agendas highlighted the importance of urban sustainability as is evident by the following statement: "cities [in industrialized countries] account for a high share of the worlds resource use, energy consumption and environmental pollution". [4] [5] While the aforementioned rather ambiguous definition of sustainability allows inclusion of wide ranging issues and sectors, it is abstract in its nature, thus hindering direct applicability and understanding. As a result, creating both palpable and legitimate goals, measures and concepts is paramount for enabling sustainable development and thus advancing the concept overall. [6] In general, sustainability measures shall aim to equally include all relevant sectors, which are nowadays commonly referred to in the modern sustainability paradigm, which is displayed in to life cycle assessments in order to quantify sustainability. [7] However, urban environments generally present a network of intricate interactions with concealed dependencies and it has been determined that carbon footprints are unsuitable for adequately capturing those complex processes. [8] In general, no well-founded decisions on future actions can be taken without extensive knowledge of the current situation, which is especially relevant for highly complex issues such as measuring sustainability. In the case of urban areas, the best practice for deriving and using appropriate indicators has been identified to constitute a holistic reporting tool to inform all involved parties such as local government, residents, businesses and other organizations in order to lay a path to achieving urban sustainability. Ideally, the overall endeavor should feature pertinent scoping, a critical review of the chosen indicators and finalization thereof, defining of goals, analysis of results and presentation to the targeted audience followed up by periodic reassessments. This approach is deemed to give vital information for communities in order to plan for sustainability and to work out constructive policies. [9] [10] Furthermore, planning is the most significant tool for sustainable development at the local level and may be seen as a political process with active public participation, which is said to benefit greatly from palpable goals such as clean air and water. As a result, further incentive is given to formulate sustainability as tangible and quantifiable measures. [11] As of 2012, there are more than 100 sustainability rating systems, of which the majority focuses on assessments of buildings and only a minority share features neighborhood or infrastructure projects. [12] Therefore, the following paragraphs are going to highlight a few selected rating schemes in order to report about the approaches and featured indicators. cover a variety of topics that can hardly be matched to the aforementioned pillars of sustainability but seem to mostly revolve around environmental issues. Additionally, in most cases the maximum attainable score is 110 points which is then further rated in the four categories of certified, silver, gold and platinum. [13] Furthermore, neighborhood development is the most suitable area of concern to evaluate communities from the already well established rating schemes, thus it will be discussed in more detail. For that purpose, figure 1.3 lists the credit categories and associated indicators for that specific rating system. It features the five credit categories of smart location and linkage, neighborhood pattern and design, green infrastructure and buildings, innovation design process and regional priority. While the latter category is clearly the least extensive one as it only awards four points in total, neighborhood pattern and design is the most significant one with 44 points in total. In general, the featured indicators are very explicitly devised and thus require extensive effort and attention to detail. However, both social, for instance access to public spaces or recreation facilities, and environmental aspects, such as building water and energy efficiency, are included. Nevertheless, economic considerations are apparently not accounted for in this rating scheme. [14]  The level of detail in the previously discussed LEED product is excessive and may often be too much to work out for rather small communities. Accordingly, the latest LEED project is targeted at evaluating entire cities, which is still in its pilot phase as of June 2017, features fewer and overall less intricate measures as displayed in figure 1.4. This approach is differentiated into nine individual categories that are seemingly much more oriented to match the three areas of environmental, social and economic. In addition, participating cities have to report progress of the featured indicators in an on-line interface, which is open to the public and thus greatly enhances the monitoring and information aspect in relation to sustainable development. [15] For instance, prosperity focuses on economic aspects with unemployment rate and median household income as individual indicators. Additionally, the three other categories of education, equitability and health and safety are designed to However, putting these numbers into perspective and interpreting them remains challenging and hinders the comparison of cities against one another. A possible approach to provide concise evaluations is to rank communities amongst a chosen sample size, which has been determined to be a sound procedure to evaluate the current status of development and quality of life. While the extreme values serve as references for the minimum or maximum score allocation, the remaining values are distributed according to their placement within that range. [16] This has been done for instance in the US and Canada Green City Index in 2011, where the featured communities were ranked in accordance with the worst and best performances across nine different categories and 31 individual metrics overall. By featuring topics such as CO 2 , energy, land use, buildings, transport, waste, water, air and environmental governance, this report, which has determined San Francisco to be the most environmentally sound city across the USA and Canada, primarily focuses on environmental aspects and lacks inclusion of social or economic measures. On the other hand, the inclusion of political aspects by assessing green action plans or management and the level of public participation in the different cities allows to compare the respective political efforts for promoting sustainability, which adds a novel and highly significant aspect to the findings. However, the report focuses on rather big cities as Orlando is the smallest featured community with roughly 240,000 inhabitants. [17] The Sustainable Cities Index 2016 takes a similar approach and compares 100 cities from around the world to one another by determining the best and worst performances for the three categories of people, planet and profit, which are composed of 32 individual indicators in total. Additionally, these metrics are grouped in sub-categories to allow comparison based on different areas of interest such as education, health and affordability, with the latter constituting the indicators of consumer price index and property prices. While this report is seemingly designed to match the areas of sustainability more closely, it is targeted at major population centers of the world, thus severely limiting its transferability to small or rural communities. [18] Additionally, figure 1  In conclusion, being able to properly quantify sustainability is crucial step in making this rather ambiguous term more tangible and thus advancing its integration and development. In general, measuring approaches may include rather conventional foot-printing, development of goals and rating systems with incorporated indicators. The latter has been done on numerous occasions and is primarily aimed for application at comparably large cities. However, when focusing on the USA, whose populace is already highly urbanized, the majority of people actually live in incorporated places with less than 100,000 inhabitants. As a result, it can be deduced that researching the suitability of sustainability rating approaches or derivation and application thereof to comparably smaller settlements may yield benefits that potentially affect a majority of the U.S. population.

Nexus Applications
As the secure provision of water, energy and food for all of the world's population has been determined to be a paramount issue, sustainable and resilient supply management frameworks are evermore increasing in importance. One rather recently developed approach to achieve overall increased efficiency of these three vital resources is the Water-Energy-Food (WEF)-Nexus, which was initially discussed during the 2011 Bonn Nexus conference. [21] It aims to promote jointly coordinated measures and policies as well as accounting for interdependencies and relationships between different supply sectors in order to achieve an overall improved use of resources. For instance, growing food crops requires water for irrigation and the conveyance thereof has an inherent energy demand. Furthermore, electric power generation often takes place in thermoelectric plants, which need water for cooling.
Additionally, that may lead to less water availability in the respective area and in turn shortages of water for irrigation and a limited supply of food.
Nexus thinking can be narrowed down to highlight the relationships between individual areas of which the dependencies between water and energy have most often been discussed as of 2015. [22] The connection of these two indispensable resources can be expressed directly, as water is required to provide electricity and energy is required to treat and transfer water, or embodied in other goods and services. For instance, thermoelectric power generation requires 80 liters of water withdrawal and two liters of consumption per provided kilowatt-hour of electricity in average in the United States. On the other hand, water treatment is rather energy intensive as provision of 60 million liters from surface water sources takes up to 60 kilowatt-hours of energy, which does not include distribution and varies with water source and required treatment. [23] Pressures on sustaining a high quality of life and providing adequate resources are amplified through densely populated environments, thus creating complex supply challenges. [24] Researching and identifying the interactions and implied consequences of individual measures on the overall supply network allows decisionmakers to determine the most beneficial course of action. The application of this approach to metropolitan areas is referred to as the urban nexus and its implementation into decision making may improve cross sectoral thinking and in turn help to advance sustainable urban design. [25] Furthermore, integrating and researching interactions of water and energy provision within an urban context is referred to as the Urban-Water-Energy (UWE)-Nexus. [26] Its importance is highlighted by the fact that water and wastewater utility operators are often the largest energy consumers in American municipalities and may be responsible for up to 40% of a city's total demand. [27] In addition, water and energy exhibit not directly evident connections in urban environments. For instance, runoff from impervious surfaces, which is the predominant land cover type in cities, contains a higher concentration of pollutants and thus has a lower water quality, which may result in an increased energy demand due to more extensive requirements for water treatment. [28] Another example is the heightened energy demand for cooling due to increased temperatures in downtown areas in comparison to the respective rural surroundings, which is referred to as the Urban Heat Island (UHI) effect. [29] [30] Furthermore, this entails an increased water demand hidden in the required amount for electricity generation.
Additionally, higher temperatures lead to accelerated evapotranspiration and, in turn, more water usage for irrigation, whose procurement and conveyance also has an embodied energy demand. [31] Understanding and accounting for those highly intertwined and somewhat concealed relationships will help to improve the overall sustainability of cities.
However, as metropolitan areas may be greatly different from one another, examinations of water and energy provision systems have to be embedded in local conditions and adopt a holistic approach in order to achieve optimal performance.
[32] This is emphasized by the substantial range of energy related to water supply systems in different cities, with 10 kilowatt-hours per capita and year for Melbourne, Australia and 372 kilowatt-hours per capita and year for San Diego, USA. [33] As a result, it is recommended that solutions have to be sought out to fit to each individual city, rather than establishing a universally applicable scheme. [34]

Pressures posed by Urban Environments
As mentioned previously, cities have relatively early been identified as major consumers of global resources. Accordingly, the term urban metabolism, which aims to holistically evaluate all the in and out coming resource flows of a city, has been described as early as 1965. [35] Quantification and research thereof requires and advanced understanding of the underlying processes, which has since been determined to be crucial for increasing resource efficiency and for limiting the negative consequences due to high consumption patterns. [ policies affect a high number of people embedded in a supply system that extensively exceeds the respective municipal confinements.

Future Trends and Developments
Research towards increasing sustainability of urban areas is highly important as ongoing global urbanization keeps enhancing pressures posed on resources and the natural environment. Accordingly, proper management of urban growth has been described as a major challenge for the 21st century. [44] Given this challenge, global sustainability has been determined to be closely linked to urban sustainability, as cities may potentially affect a majority of the global population and define the quality of life for their respective citizens. [45] Additionally, both climate change and globalization further frame the upcoming challenges of the 21st century.
As of 2014, 54% of all people were already living in urban areas, whereas in 1950 this was applicable to only 30% of the worlds population. Furthermore, this trend is projected to continue, leading to an increase in urbanized residency to 66% by 2050. While this ratio varies significantly by region, North America is one of the most highly urbanized regions as 82% of its citizens currently reside in metropolitan areas. [46] Furthermore, as figure 1.7 shows the USA has seen a rapid change in its settlement distribution. While, almost 95% of its citizens lived in rural areas in 1800, this status has almost reversed with over 80% of its current population currently residing in urban areas. Not only is the world population projected to become increasingly more urbanized, there will also be a higher number of extreme population agglomerations or mega-cities. While there were only two mega-cities, which refers to cities with Climate change, which describes significant long term deviations from recorded weather averages such as temperature or precipitation, will be a major challenge of the upcoming century. Furthermore, the recent and extreme shifts, which especially applies to temperature increases since the beginning of the 20th century, are largely attributed to anthropogenic activities, for instance, the exceedingly high releases of greenhouse gases like carbon-dioxide. [51] [52] Next to rising temperatures, which may exceed 5 • C on average till 2100 under a high emissions pathway according to figure 1.8, the occurrence of extreme weather events and significant shifts in regional climate may change as well. This will culminate in substantial impacts ranging from flooding, sea level rise, altered crop yields and increased water stress amongst many others. Additionally, risks associated with climate change generally increase with the temperature, as for instance a 2 • C average increase is thought to put unique ecosystems such as coral reefs and arctic species under very high threat, while an increase of about 0.5 • C will keep the risks at a medium level. [53] When focusing on North America, wildfires, loss of property and ecosystems, increased human mortality and morbidity due to heat, urban flooding, infrastructure damage and water quality impairment have been identified as the most relevant risks related to climate change. However, most of these factors will only reach high risk levels beginning in 2080 and all have the potential to be reduced to or even below medium risk by adequate adaptation measures. [53] As mentioned previously, the USA is a highly urbanized country, attributing additional importance to climate change impacts on cities. In general, reliability of service provision and economic stability will be challenged by a changing climate, which is especially true in areas of low elevation in coastal zones, where about 13% of the global urbanized population currently resides. Furthermore, as cities differ greatly in their structure and composition, there can be no universally applicable adaptation strategy and local conditions and potential impacts have to be taken into account for each individual case. However, it has been determined that decision makers should utilize long term urban planning in a timely manner in order to prevent detrimental consequences on the various sectors and demographic groups.
[54] Additionally, while mortality in urban areas generally increases with rising level intense heat during the summer months, certain groups such as elderly people or children may be especially affected. [55] In conclusion, adaptation of urban areas to climate change has to be worked out individually in accordance with the local conditions and should start rather sooner than later. This undertaking is profoundly important, as cities house a high number of people that are reliant on the associated supply infrastructure and may be especially vulnerable to the failure thereof or otherwise dramatically altered conditions. This study will provide a detailed case study for understanding the inherent environmental connections between water and energy provision by examining the current system in Rhode Island. Additionally, a sustainability rating scheme on a municipal level will be worked out in order to determine potentials for further sustainable development. This will involve establishing the current sustainability status of Rhode Island communities by collecting data and creating indicators to rank them according to one another. Furthermore, possible wide ranging benefits, through a joint evaluation of water and energy provision in urbanized areas, will be examined by exploring spatially relevant implications. The study will conclude with evaluating the sustainable development potentials in Rhode Island by identifying major areas for improvement.

Hypotheses
This paper will evaluate the following two hypotheses, which will be elaborated in chapters 3 and 4, respectively, and reviewed in chapter 5 subsequently to all necessary assessments.
The primary hypothesis of this study is that a monitoring approach, involving social, environmental and economic indicators, will be useful to assess the sustainability of Rhode Island's communities. It is believed that this rating of individual municipalities against one another will provide possible knowledge gains in fostering sustainable development in the state.
As mentioned previously, nexus thinking aims at creating more efficient resource management by accounting for the interactions between individual sectors.
Accordingly, the secondary hypothesis of this study is that such an approach for the water and energy provisions, with distinct focus on the urbanized areas of Rhode Island, will allow for the identification of sectors and areas where more efficient practices and resource management can be ensued.

Thesis Structure
Overall, this thesis has two main research objectives, which are as follows: • Compilation of a comprehensive municipality ranking areas, which will be discussed in more detail in section 2.6, with the metropolitan agglomeration around Providence clearly being the most significant one. All of the aforementioned geographical features, except the allocation to counties, can be seen in figure 2.1. Additionally, it also displays the population distribution, which is clearly centered around Providence or largely along the Narragansett Bay.  As mentioned previously, counties mainly work as geographic reference in Rhode Island and municipalities function as local government instead. However, counties remain significant at times, as federal agencies and census reports still commonly refer to them in their publications.

Climate
Rhode Island's climate is best described as humid continental and as such has a rather high annual temperature of 50 • F and an average yearly precipitation of 46 inches. There may be extreme weather events such as hurricanes, blizzards, heavy snowfall and periods of uncharacteristic seasonal behavior, but the precipitation pattern is generally well distributed over the course of the year and snowfall is a commonly occurring phenomenon. Furthermore, there are clearly distinguishable differences between coastal and inland locations with the overall main wind direction originating from the west. [62] [66]

Climate Change Projections
The particular challenges, projections and vulnerabilities of Rhode Island regarding climate change will be discussed in the following section. Global projections and trends, which include, for instance, changing weather patterns, rising sea levels and temperatures have been discussed in section 1.1.4.
Overall, climate change is a significant challenge for Rhode Island and an average increase of 3 • F, which is projected to accelerate even faster under a high emissions pathway, has been observed over the course of the 20th century.
Accordingly, the intensity of heat waves is projected to increase while cold weather periods are decreasing. Furthermore, precipitation has increased as well as the occurrence of associated extreme events, which may lead to a higher frequency of flooding as this trend is projected to continue. Additionally, the town of Newport has experienced a sea level rise of 9 inches since 1930. This trend is above the global average and may culminate in an additional four feet by 2100. [73] Rhode Island is judged to be particularly susceptible to sea level rise out of the aforementioned developments. This is evident just by the fact that it has the second highest ratio of shoreline, which includes offshore islands, bays and certain sections of streams and creeks, to total area of all states. Its shoreline of 384 miles is heavily shaped by the Narragansett Bay, which reaches about 30 miles inland till Providence as can be seen in figure 2.1. Furthermore, the states most significant urban area, and with it about 88% of its population, is centered around this coastal estuary, resulting in a close proximity of a majority of its citizens and infrastructural assets being located close to the shore.
Accordingly, Rhode Island is generally considered to be severely threatened by   Overall, the lower emissions scenario ranges from a temperature increase roughly between 2 • F and 8 • F. On the other hand, the higher emissions scenario may result in up to 14 • F higher temperatures and is projected to stay above 7 • F at best.
[73] Additionally, Rhode Island has experienced a slightly higher increase of mean temperature, which has risen 1.

Urbanized Areas
In general, urban areas are associated with high population densities as well as a heavily built up environment ranging from industrial or commercial sites, infrastructural assets such as bridges or railroads and residential quarters. Furthermore, they are often centered around specific cities and also encompass the surrounding areas such as suburbs and attached towns. However, the particular definition and methodology for delineation differs by country. For instance, while communities with 2,500 or more inhabitants are considered urban in the United States, this threshold is set at 30,000 inhabitants in Japan. [88] In the United States of America the most basic form of a comprehensive community is known as an incorporated place, which requires a local government,  Overall, Rhode Island's population is highly urbanized with more than 90%   inhabitants. However, it is only the second most densely populated municipality as Central Falls houses almost 6,000 more people per square mile than Providence.
Overall, a declining population density seems to go along with a higher share of people living in a rural environment, which is quite an expected relationship. There are several communities, whose entire populace or at least a high portion thereof lives in urban environments, while there are just about thirteen municipalities with 50% or more of its people living in rural areas. This observation conforms to section 2.6 as more than 90% of the entire state's citizens live in urban areas. The state's unique situation is concisely described by it being both the smallest in regards to area and second most densely populated states. Its biggest community is Providence with about 178,000 residents, which is a relatively low figure in context of major US metropolitan areas. However, it poses as the center of the overall 39th largest urban area of the entire USA, which emphasizes its significance both on a regional and national scale. The municipalities will be thoroughly analyzed and evaluated by using the parameters of tables 2.5 and 2.6 and the results of the sustainability ranking in chapter 3.

Utility Provision and Important Sectors
As both main assessments of this paper, namely the sustainability ranking and Urban-Water-Energy-Nexus evaluations, heavily relate to specific features of Rhode Island's environment and service provision, the most significant areas will be highlighted in the following section. This includes energy provision, public water, waste water, solid waste, economic aspects, transportation infrastructure, air quality monitoring and land use.

Energy
This section will provide an overview for the energy sector of Rhode Island, ranging from spatial distribution of major infrastructural assets and generation capacity to time series data for power generation. Furthermore, the state's demand and consumption pattern will briefly be put into a national context.   Furthermore, Block Island houses one of only two petroleum fueled generators, which was also the only generating facility of the island when the EIA power plant dataset was updated early in 2017. However, the first offshore wind farm of the United States commenced operation in May 2017 and consists of five turbines with 6 MW capacity each. Next to increasing the state's renewable generation as a whole, the wind farm will likely benefit the local residents by reducing the electricity rate significantly from over 60.00 cents kW h , which was driven up due to fuel prices and the transportation thereof. [99] [100]   Overall, while Rhode Island has a relatively low energy demand in comparison to other U.S. states, the costs for residential consumption are the second highest overall. Its electricity is predominantly produced in natural gas power plants, which account for over 90% of the overall generation capacity and for almost 96% of the total generated power from 2013 to 2015.

Water and Waste Water
This section will discuss the current infrastructure for both water supply and waste water management, as well as overall usage figures. First of all, water suppliers are a highly diverse sector as there are up to 490 individual supply companies. They range from major suppliers, such as the Providence Water Supply Board (PWSB) which manages about 68 MGD on a daily basis to individual systems responsible One of the few nation-wide reports to compare water usage is carried by the can be further differentiated into withdrawals and consumptive use, which describes processes that alter the availability of water for instance via evaporation or temporal embodiment in products e.g. crops or livestock. However, limited data availability does not allow for more detailed nation-wide reporting on this matter, rendering appropriate assessments for the individual sectors even more important. The composition between the usage categories is displayed in figure 2.14, which reveals Thermoelectric Power as the most significant sector as it accounts for almost two thirds ov the overall demand. Due to its small size and population, Rhode Island has the second lowest overall water demand of all states next to the District of Measures including total amount of waste generated and the processing such as landfilling or recycling are paramount for assessing the performance and long term planning or designing of associated facilities. Accordingly, table 2.10 shows commonly used specifications which will be used for subsequent evaluations. There are several processing rates, which largely improve upon another. For instance, while the MRF recycling rate incorporates recyclable materials, the mandatory recycling rate goes one step further and also includes other salvageable items such as leaf and yard debris. The diversion rate is the most comprehensive measure in terms of waste disposal, as it describes the ratio of total waste that is not disposed of in a landfill but otherwise processed. Therefore, a higher diversion rate allows to operate the corresponding landfills longer and thus indicates a higher useful facility lifetime and return of investment.

Transportation System
This section will constitute an overview of the states most important traffic systems which include highways, train connections, public transportation and airports. A general overview about the current rail ways, bus routes, bike trails and alternative fueling stations is displayed in figure 2.19. The area of Providence clearly serves as a major node for all traffic systems as the network of routes and tracks seem to heavily concentrate around it. Furthermore, of all of the three interstates which connect between New York and Boston, Interstate 95 is the most significant.
It lead towards the city, thus further emphasizing its role as a transportation hub for the state. The same observation can be made for the alternative fueling stations, which are primarily located near major traffic ways and cluster around Providence and Newport. [111] Even though the interstates only amount to 1.2% of total road mileage in the state, about 35% of all vehicular motorized traffic passes through them, resulting in occurring congestion on a regular basis. The overall road network encompasses more than 6,700 miles with local roads accounting for almost 60% of it. Additionally, 4,400 miles of the road network were built before 1962, which together with salt exposure and unfavorable weather conditions lead to a higher than average deterioration of infrastructure. The same conditions apply to pavement and bridges, of which there are more than 700 in the state, further highlighting the importance of proper maintenance. Additionally, there are over 60 miles of designated bike lanes in the state and further enhancements are already underway to promote cycling. [111] The Rhode Island Public Transit Authority (RIPTA) maintain and operate all public transportation services on a local in the state, which includes both fixed bus route and flex service. As of 2016, the network of bus routes, which is displayed in figure 2.19 and encompasses 1,019 miles, resulting in a line density of 0.95 mi mi 2 in relation to the state's land area. Additionally, there are slightly over 5,000 stops, which lead to an available stopover about every 1,000 feet in average for all lines. [64] Flex service is employed on demand to enhance the accessibility of Providence also serves as a major node for freight conveyance as it houses a majority of trucking terminals and the state's main commercial port where up to 2.7 million tons of cargo are shipped on an annual basis. However, rail infrastructure is highly important on a regional context as it provides many convenient connections within the state and to other areas of New England. Additionally, a functional highway system with reduced congestion has been determined to be crucial for ensuring an adequate movement of freight. [111] Next to taking stock of inventory, a commonly used method to assess mobility patterns is to display the modal share on total or the commuting traffic. The ACS reports commuting characteristics for the seven different modes of single car driver, carpooling, public transportation, bicycling, walking, working from home and other means of transportations such as taxicabs and motorcycles and thus provides an excellent basis for working out modal splits as a means of comparing communities or regions. Additionally, the mean travel time in minutes is also supplied, which further helps to evaluate the local situation. [65] Accordingly, the modal split has been worked out in relation to the entire USA, Rhode Island and four of it's municipalities and is visualized along with the mean travel time in figure 2.20. In general, the state of Rhode Island commuting modes are quite similarly distributed as for the entire country with individual car use accounting for a marginally higher share, while the mean travel is slightly lower. However, the statewide characteristics can not be easily transferred to its individual communities, as the distributions differ vastly from one another. For instance, Providence has both the highest commuting share by walking and public transportation, while both Cranston and Foster rely much more heavily on individual car usage. Even though the commuting shares of these two municipalities are very similar to one another, Foster exhibits a significantly higher mean travel time possibly due to its rather remote location. On the other hand, New Shoreham, which is easily the state's most isolated municipality, shows the lowest mean travel time overall as a high portion of its residents work from home rather than commuting on a daily basis. In conclusion, the travel characteristics of the individual communities may differ greatly from one another with different impacts on sustainability. For instance, a high share of public transportation is considered to be more sustainable as the amount of emissions are significantly reduced in comparison to single car usage. [65] [116] A sustained effort to maintain an invest in transportation infrastructure has been determined as a main agenda to ensure a safely and adequately usable network.
While the minimum scenario has been estimated to require $ 454 million for that purpose, the most favorable case with a high number of beneficial projects such as walkable communities, streetcars and bicycle accommodations may require up to $ 1,150 million. However, both scenarios are plagued by uncertainty due to a rather outdated financing support structure, for which especially stagnant revenues from fuel taxes are problematic. [111] According to the most recent estimates of the ASCE, about a quarter of Rhode Island's 772 bridges are structurally deficient, even though about $ 99 million were spent on related maintenance projects in 2013.
This situation is the worst among all states, highlighting the upkeep of projects and funding to ensure safe usage of traffic ways. [117] Furthermore, about 54% of all roads, which amount up to over 6,700 miles, are in inadequate condition and cause costs up to $ 810 per user and year. Overall, it is strongly recommended to start investing in infrastructure sooner rather than later in order to keep risks to a minimum level and enhance economic competitiveness. [107] [111] In summary, Rhode Island encompasses over 6,700 miles of roads and 1,000 miles of public bus routes and both its ports and rail infrastructure are of high importance for freight movement. Car use still accounts for the highest share of commuter traffic with over 85% statewide, which may differ greatly depending on the location. Lastly, there is a high projected need of investment in infrastructure to ensure a safe and adequate transportation system.

Air Quality Monitoring
This section will feature a concise introduction to air quality monitoring, the related targets and purposes as well as the current status of both air quality and the related monitoring network in Rhode Island. In the past, worsening ambient air conditions have been viewed as a necessary byproduct to industrial progress, largely due to a lack of understanding of the consequences to the natural environment and health and well being of the affected population. [7] However, extreme events such as the great smog in London of 1952, where extreme agglomerations of air pollutants over several days caused the deaths of over 4,000 people, have sparked research incentives. This results in modern monitoring frameworks aiming to prevent increased mortality due to respiratory or cardiovascular illnesses because of high exposure to detrimental air conditions. [118] The main regulation in the USA is the Clean Air Act of 1970, which has undergone major amendments in 1990 with a focus on acid rain, urban air quality, stratospheric ozone depletion and toxic air emissions. It is managed and enforced by the Environmental Protection Agency (EPA) and the locally assigned authorities. Even though there have been major improvements since 1990, there are still significant areas for concern in regards to ground level ozone receiving increasing attention. In 2015 about 127 million people resided in counties with concentrations above the respective national standards. [119] [120] Proper monitoring of air quality is highly important but at the same time requires an extensive amount of resources. In Rhode Island, the DEM is tasked with planning, management and operation of monitoring and therefore is the primary contact point for that matter. Since 1968, its office of air resources has been observing the state's air quality in a joint effort with the Rhode Island Department of Health (RIDOH) and features seven monitoring locations, five of which are used for daily Air Quality Index (AQI) reporting, in total. [121] Its work is focused on the six criteria air pollutants from the National Ambient Air Quality Standards (NAAQS) established by the Clean Air Act, which can be found on page 193 in the appendix. air quality is not a major concern and only certain pollutants have been described as worrisome overall. [122] The aforementioned AQI monitoring is carried out on a daily basis and delivers almost instantaneous results. In 2011, the continuous reporting network, which is displayed in figure 2.22, featured five sites overall of which two are located in Providence, one in East Providence and the remaining two are in the rather rural environments of West Greenwich and Narragansett. The overall monitoring network may differ from figure 2.22 for measuring specific pollutants and is in general designed to provide the best spatial coverage in the most populous areas of the state. For that purpose, the current status of the network is revised every year in order to adapt to recent trends and to achieve the best possible results. [123] Overall, monitoring sites are concentrated around Providence next to two locations in rather rural or background areas of the state, as displayed in figure 2.22.
Such a set-up may be suitable to report on the situation in the state as a whole, but is inadequate to draw conclusions for every single remote location, which would regardless require an unnecessary amount of resources and effort. Accordingly, the AQI is reported and forecasted only for three separate locations in Rhode Island, which may allow to detect trends in nearby communities but is rather unsuitable for in depth comparisons. As a result, there are complimentary actions to assess impacts on air quality such as tacking stock of emission and especially harmful pollutants and regulation of disadvantageous technologies such as diesel engines. On a national scale, toxic emissions are recorded and published in the Toxic Release Inventory (TRI), which is maintained by the EPA and included 88 facilities for Rhode Island in 2015. Overall, on site toxic air releases amounted to 293.2 thousand lb in that year, which is equivalent to almost the entirety of records in that category. [124] Furthermore, this task is also carried out by the DEM on a With about 58.03% most of the state's land surface is covered by forest and the second biggest share is achieved by development such as residential, commercial and industrial sites or infrastructural assets such as roads and power lines. [64]  is still forested and both state and other conservation lands cover roughly a quarter of its land area. Accordingly, Rhode Island has managed to preserve a strong rural character and many of its natural assets such as beaches, bays, forests, farms and rivers even though it is one of the most densely populated states in general.
[128] Therefore, the statewide land use development plan, which is targeted at the year 2025, has identified preservation of rural areas while promoting growth of urban centers as a guideline with acute priority. Additionally, underutilized urban neighborhoods, highway interchange infrastructure and waterfront areas have been identified as major areas of concern moving forward. [129]

Economy
This section will provide a short introduction to Rhode Islands economic assets in order to provide a frame of reference for the economic section of the sustainability ranking. Accordingly, major employment sectors, metrics and other relevant areas will be discussed.  As of 2014, uneven wage development, diminishing middle class, racial gaps in income, health and employment opportunity have been determined as major upcoming challenges. Furthermore, the population is getting considerably older on average, which leaves jobs to be replaced by young professionals. Therefore, the goals of the long range planning effort include promotion of proper education, an inclusive and more diverse workforce and creation of financially competitive and attractive locations. The goals are expected to be completed by 2035 and require investments in economic sectors, housing and transportation. [130] CHAPTER 3

Sustainability Ranking
This chapter will feature the methodology for working out the sustainability ranking and the individual sectors as well as evaluations on a spatial scale and by using chosen municipal parameters. Additionally, noteworthy divergent areas per community will be identified and discussed accordingly. Eventually, the most sustainable municipalities and most unfavorably performing ones will be determined and a recommendation for future assessments will be worked out.

General Approach
As discussed in section 1.1.1, actually measuring sustainability is no small feat due to its ambiguous nature and applicability across various intertwined sectors. However, this endeavor may reap tremendous benefits as creating clear and understandable labels is crucial to reach the people whose actions induce tangible outcomes. Furthermore, establishing a framework with distinct local connections may help to create additional motivation and incentive for the respective populace.
Therefore, this study aims to establish a sustainability rating for the municipalities of Rhode Island in order to give a comprehensive overview of the current status as well as identifying areas for further improvement. As scrutinizing rather small individual municipalities regarding such a rating scheme is a novel approach, this iteration may not be entirely free of flaws. However, there is substantial insight to be gained, which renders research in this area highly important.

Methods
This section will discuss the overall ranking approach and methodology regarding the derivation of certain measures. Additionally, section 3.3 will discuss the data sources, while section 3.4 will state the intent of each individual indicator in detail.

Compilation and Evaluation of the Ranking
Overall, there is an abundance of existing frameworks and set of indicators, which are usually tailored to fit the resources, parameters and conditions of the city or research area of interest. [132] As a result, there is no commonly accepted or generally applicable methodology for assessing sustainability of cities on a global scale without requiring extensive work for data acquisition and processing. For instance, the framework developed to assess progress and provide objectives for the Sustainable Development Goals (SDG) features 232 indicators in total, which may often be too detailed and inappropriate for comparably small settlements. [133] In addition, the choice of indicators and the significance thereof may vary with regional characteristics or challenges, leading to different points of emphasis. For instance, reporting on the number of people living in slums or suffering from malaria may be a much more pressing concern in rapidly evolving Asian mega cities rather than a decently sized and wealthy European city. As a result, the applied approach for this paper was modeled after a few existing examples and suited as best as possible in order to provide a holistic comparison while also accounting for availability of data. However, mainly due to time constraints, the chosen indicators were worked out with already existing data or rather easily elaborated parameters. Therefore, potential future iterations may benefit greatly from improved data acquisition and well planned collaborations with appropriate state offices or otherwise affected parties.
In general, the finalized ranking intends to cover all aspects of sustainability or likewise its three pillars of environment, economic and social. As those categories are thought to be part of a highly intertwined and cohesive system, comprehensive evaluations instead of focused examinations on individual aspects is strongly recommended. [7] Furthermore, differentiating the ranking further into individual segments, such as energy, water and transportation amongst others for the environmental category, allows for the identification of exceptionally performing areas for the communities. As mentioned previously, the methodology has been modeled after a few already existing systems, which are discussed in section 1.1.1.
As a result, the approach of this elaboration does not feature novel methods for assessing sustainability but explores the applicability to rather small communities in great detail for the state of Rhode Island.
Given the lack of definitive values on which to base goals or otherwise related measures for quantification, it was decided to compare the municipalities amongst each other according to the respective best and worst performances for each indicator.
Furthermore, all values scattered in between those two thresholds have been alloted a score according to their placement amongst the range between the minimum and maximum values. Additionally, depending on how the indicators have been interpreted, each score has been distributed from low to high values or vice versa.
This approach allows for the derivation of a ranking in respect to all featured municipalities, while achieving a high level of detail in a regional context and clearly identifying overly positive or negative performances. All indicators are listed in tables in section 3.4, where the methodology and reasoning for each case is also explained in detail along with the assigned score, data source and year of origin.  The general structure of the ranking procedure is displayed in figure 3.1 and is split into the three categories of social, environmental and economic aspects.
Additionally, these categories further consist of themed segments such as education, safety amongst others in order to recognized different points of interest or emphasis.
Next up, the segments are made up by the actual indicators with the featured number varying according to data availability and thus the possible level of detail during the working process. While the number of indicators per segment ranges from three to seven, the awarded score per segment always amounts to ten in total in order to allow comparison on an equal basis within the categories.

Derivation of the Indicators
In comparison to, the required amount of effort to derive the indicators may vary significantly. While some sources, for instance the ACS reports and waste related measures as supplied by RIRRC, publish their data already in relation to the municipalities, most information has to be edited in order to fit to the referenced communities. Furthermore, a few indicators such as water and residential energy demand per capita had to be generated almost entirely from the ground up, resulting in an excessive increase of work for those areas. Accordingly, the next few paragraphs will highlight the used methodology for chosen indicators. was decided to work with census blocks as they allow evaluations on a much more detailed spatial scale. However, as block groups are the finest spatial feature of ACS publications, which is supposed to supplant the decennial census procedure, future elaboration may have to focus on using them instead. This would also enable to better match demographic information to the year of origin of the spatial dataset.
Accordingly, this work compromises for some minor discrepancies as for instance TRI facilities stem from 2015, while the population distribution originates from the 2010 census. [64] [138] [139] As mentioned previously, the indicators related to water and energy demand required a comparably high amount of work and thus their derivation will be described in detail during the next few paragraphs. According to section 2.8.2, there are almost up to 500 different suppliers in Rhode Island rendering comprehensive data gathering an excessively complicated task. As a result, the here derived indicator used for comparing the communities water demand on a per capita basis may benefit greatly from future work regarding data collection. For this thesis, a number of reports from the RISPP and the RIWRB were consulted, for which the most significant specifications can be found in appendix C, to obtain figures on water demand. Providence and Providence, whose water demands amounts to 68.14 MGD in total. Furthermore, normalizing to the number of residents results in the final measure by which the communities will be compared later on. This approach neglects water transferrals between the suppliers and thus may underestimate the overall demand for communities that rely heavily on this procedure. However, as no other figures on municipal water demand were available, the derived measures are going to be used nonetheless. Accordingly, establishing a comprehensive reporting scheme to achieve a higher quality of data regarding water usage of Rhode Island's communities proofs to be an advisable future goal. This procedure was executed for all 39 municipalities and five criteria, resulting in an averaged residential energy demand per capita and year for all communities. Furthermore, the applied methods result in a statewide average residential consumption of 12,309 kW h per capita . Naturally, this estimate can not be as accurate as figures measured by the utility provider, which is in this case National Grid, but still adequately matches the statewide averages as discussed in section 2.8.1 for 2015, especially as the derived figures account for all usage categories instead of just accounting for electricity. Furthermore, this methodology features a linear approach, which may be improved upon by accounting for interactions between the parameters and with the ambient conditions in order to achieve more realistic estimates. [141] Overall, the methodology for deriving the indicators has been profoundly discussed in the prior section and major areas for improvement have been mentioned. This includes a lack of data regarding economic figures and water or energy demand per municipality. This procedure can be seen as a crucial step to enable enhancements in future iterations.

Data Resources
In general the used data comes from a variety of sources from within Rhode Island and federal agencies, which in some cases offer datasets with a fitting resolution to work on a municipal level. The US Census Bureau is tasked with providing in depth information about the United State's population with the decennial census being its most popular and widely used publication. Its first iteration was carried out in 1790, as required even by the very first version of the U.S. Constitution, to establish an informative basis to enhance political decision making in communities. From that point forward the census survey has been used to acquire population counts and detailed demographic information every ten years till 2000, when the latter function was assigned to the newly developed ACS.
The main objective of this alteration was to provide figures on an annual rather than a decennial cycle, which is more suitable in the rapidly changing information  used, as the detailed information of the ACS is not publicly available for that spatial resolution in order to prevent privacy intrusions of individual households or companies. However, from Census Tract level upwards the ACS provides extensive information such as mean travel time to work, median age or household income. As the level of accuracy of the survey estimations have been deemed to be questionable from the tract level downward, county subdivision data will be predominantly used in this paper. [141] Occasionally, federal data sources have been used, which include the Toxic

Social
The social indicators are distributed over the following six categories of education, safety, health, work life balance, housing and voter participation and equality.
All individual segments have been allotted ten points in total but the number of featured indicators ranges from seven for safety and health to two for voter participation and equality.
To begin with, table 3.5 features measures on which to evaluate the status of educational attainment in the municipalities. Four different indicators have been derived from ACS 2015 data and weighed equally within this segment. While the first two indicators intend to report on the current status of education via the percentage of population with at least a high school or a bachelor degree, the latter two compare the communities regarding school eligibility and current enrollment for the two age thresholds between 5 and 17 and between 18 and 24 years of age.
All four measurements have been judged to indicate a more beneficial situation with increasing values, thus ranking them from low to high. [65]     Altogether, the aforementioned indicators intend to cover an array of social issues and could largely have stemmed from the ACS of 2015, but local data sources, for instance HousingWorksRI or RIDOH, have been used preferably if available.
For example, both the ACS and RIDOH report births and the number of law enforcement employees is also reported locally or in census data. Overall, the working process revealed two main difference between Census reports and local sources. While the latter may potentially capture the situation in the state more precisely, the former is available on an annual basis and thus allows the derivation of periodic rating iterations as a next step. However, this principal elaboration has first to establish a current evaluation and thus the more precise data has been used if available.
Unfortunately, some important issues could not be considered for this thesis due to a lack of available data for all municipalities. This is especially true for the segment of health, as detailed information for example obesity rates or other significant health concerns are simply not reported on the required spatial scale.
However, worked out indicators still are deemed to cover most issues in appropriate detail and even allow for identification of exceptional situations as visualized in

Environment
The environmental indicators are split into the six categories of energy, water,   The indicators related to land use, which are displayed in table 3.14, give information about the share of municipal areas covered with impervious materials  The communities were compared regarding generation of solid waste and processing thereof by using the annual metrics for 2015 published by the RIRRC. The publications include the figures of overall produced waste, which were normalized with the 2015 ACS population numbers, as well as recycling and diversion rates.
Details and explanations regarding the metrics can be found in table 2.10. [110] In order to easily evaluate all environmental indicators, the distribution of  The transportation segment features the rather urbanized communities of Central Falls, Providence and Newport as above average performers. They tend to feature a high share of commuters using public transportation, a close proximity to alternative fueling stations and a high accessibility to public transit facilities.
The latter is notably true for Central Falls and Providence as almost all of the respective citizens live within half a mile of a bus stops and the amount of bus routes per area, for instance there are 11.86 mi mi 2 in Providence, is especially high. Overall, the municipalities tend to receive rather high scores in the water segment. This is most likely due to the sole indicator of water demand being apportioned seven points, however there are a communities with significantly heightened figures. For instance, New Shoreham has a demand of 331.13 MGD, while the statewide average has been determined to be around 161.90 MGD as described in detail in section 2.8.2. Accordingly, such a considerably higher demand has also been worked out for Burrillville and Cranston. In comparison, the three remaining indicators regarding plumbing facilities, sewered areas and unaccounted for water have a low influence as they are only attributed with one point each.
As previously described, the land use segment features a variability of scores and no drastically different values were identified. This is largely due to the somewhat mutually excluding set-up of the indicators, as a high share of impervious surfaces is likely to entail a low share of conservation values. However, the highest score of 9.4 points belongs to West Greenwich as it exhibits the third lowest share of impervious surfaces and the highest share regarding 300 to 600 LULC coded areas and state conservation areas. On the other hand, Central Falls receives the lowest score in this segment as its densely structured set-up entails a low share of natural areas and leads to the highest overall share of impervious surfaces.
The indicators related to air lead to a tightly packed distribution of scores with Johnston, Burrillville and New Shoreham as clear below expectations performers.
Those communities had the three highest per capita emission and New Shoreham and Burrillville also had the two highest per capita air emissions from the facilities listed in the TRI. As those two indicators alone account for seven of ten points and the mentioned communities exhibit significantly higher values, the remaining municipalities tend to receive rather high scores overall.
Lastly, the point distribution of waste related indicators is similar to those related to air. The 3,881.9 kg per capita of solid waste generation in New Shoreham nearly dwarfs all other communities, as the average value for all is 555.7 kg per capita . Additionally, it also achieves considerably below average results regarding recycling rates and diversion from landfill, rendering it by far the most unfavorably rated municipality regarding waste. Furthermore, Providence has one of the lowest per capita waste generation but features the lowest overall values for all processing rates. In general, the same issues apply to Johnston but in a marginally more unfavorable magnitude, thus ranking it slightly below Providence.

Economy
The economic indicators are distributed over only the four different segments of income, employment, value and mobility and connectivity. In comparison to the other overarching categories of social and environment, the economic category includes slightly fewer indicators overall and has noteworthy gaps of data. This is mainly due to the fact that no extensive database regarding significant economic aspects such as the GDP, which is seldom determined for individual municipalities, and number of businesses or workplaces was available. As a result, the economy of the individual municipalities may not always be portrayed absolutely accurately and certain refinements in future iteration may prove to be very beneficial. Therefore, the economic aspects have been apportioned with only 20% of the final score, while the other two areas were each accredited with 40% of the finalized results. However, the following indicators still provide a complementary overview over the respective economic situation while the chosen segments highlight different, important areas of interest. According to the previously reviewed categories, the economic indicators will first be discussed by segment in tables 3.17 to 3.20 with an overall comparison via a box plot visualization in figure 3.6, while the finalized scores per municipality will be listed in section 3.5.
To begin with, all income related indicators have been retrieved from ACS 2015 data and feature different references such as per capita, household or family.
Additionally, the share of people with an income below the poverty level has also been included and allocated the most points within this segment as it differs substantially from the other featured categories. In an economic sense, higher income was interpreted to be more sustainable, as it allows more spending and thus a higher economic activity in general. [65]      In accordance to the other categories, the score distribution of the economic segments has also been visualized in a box plot, which is displayed in figure   3.6. Connectivity and mobility shows the highest average value with almost 7 points, while value has the lowest average score with slightly less then three points.
Additionally Lastly, connectivity and mobility features three unfavorably rated outliers. All three communities are located in rather remote places, resulting in low accessibility via highways, airports and harbors. Furthermore, all of them show a distinct lack of fiber service availability, indicating a comparably low communication capability.
However, Newport is ranked most unfavorably as it displays the lowest share of workers without an available vehicle.
In conclusion, the featured indicators evaluate the communities based on income levels, current employment and potential availability of jobs, municipal value and the ease of commencing in business vie telecommunication, proximity of relevant transportation infrastructure and personal mobility. However, a lack

Results
In order to determine the most sustainable municipality the individual scores of the three segments were added up in order to achieve a single score on which to clearly rank the municipalities. While figure 3.7 shows the totally achieved score arranged from best to worst performing community, figures 3.8 to 3.10 display the composition of the social, environmental and economic scores by segment before the weighed sum and are also arranged from best to worst. In order to account for a lack of available data, the economic category was attributed with 20 points, while the other two areas each received 40 points. The scores per community were achieved by converting the score per category, for instance 60 points for social to 40 points in the overall rating, and adding them up afterwards. Additionally, the rank regarding the overall rating or for the individual categories is also given in order to provide a concise overview for each community.    indicators related to income, resulting in zero points for that segment. Additionally, it also performs poorly regarding employment and value, while connectivity and mobility receives an average score. In general, there are only very few communities that perform significantly worse regarding income, such as Pawtucket, Providence and Woonsocket, or connectivity and mobility, which includes for instance New Shoreham and Newport.
This evaluation procedure is deemed to be well suited to compare all 39 communities to each other, while also providing an appropriate level of detail.
However, the individual indicators need to be examined in detail in order to clearly identify areas for improvement. For that purpose, the rating score by category and segment can be found in the appendix beginning on page 195. However, the indicators per municipality are not listed in this thesis, but appendix B includes instructions on how to access them on-line.

Evaluation
This section incorporates an evaluation of the ranking results by analyzing the distribution of the attained scores with a box plot, segmental assessment of score composition per municipality, mapping to determine spatial trends and exploring possible correlations between the ranking and demographic parameters of the communities.
To begin with, Overall, the distribution does not show any drastic developments, thus indicating a steady composition for the majority of communities. However, a few common population based denominator. Accordingly, its waste production and water or energy demand are by far the most unfavorable values on a per capita basis.
Additionally, the potential influence of seasonal tourism is further emphasized by the fact that Little Compton, which is the second least populated town of Rhode Island, performs significantly better in both the environmental category and the total ranking, where it achieves the fourth best score overall.

Spatial Patterns
The ranking results were mapped in order to allow identification of spatial patterns. While figure 3.13 displays the overall results, figure 3.14 additionally shows the performance for the three categories Both representations are scaled according to the best and worst performing communities of the respective category, allowing for an easily transferable overview.      Next, the connection between the two most significant measures of population density and income will be analyzed in more detail. Therefore, the respective development of all municipalities is visualized in figures 3.15 and 3.16. To begin with, figure 3.15 displays the ranking performance of all municipalities over the population density and shows a negative trend for all categories, which emphasizes that more densely settled communities tend to perform poorly in the ranking. Furthermore, the coefficient of performance varies with the categories and environmental aspects seem to be least closely connected to this parameter, while economic performance shows a rather high correlation to the ranking scores. In general, just about one point of the overall score is deducted from the results with 1,000 additional people per square mile. However, as there is only a rather low correlation overall, the most densely populated community is not automatically the worst performing one. indicating that a more comfortable individual financial situation is beneficial to overall sustainability. As expected, this relationship is strongest regarding the economic aspects as it exhibits the highest COD for all featured parameters.
However, social considerations are more significant in terms of total development due to the higher allocated score regarding the finalized ranking. Altogether, about 0.4 points are added with each additional 1,000 score 1,000 $ per capita . As a result, residents of more sustainable communities tend to have a higher income.
The featured correlation examinations have a rather simplified and brief character, but point towards promising future research areas and help to characterize the communities of Rhode Island in more detail. However, the results of this section are hardly transferable to other areas as they are closely linked to the study area. Furthermore, they are inherently influenced by the derivation methods, which have noteworthy gaps as discussed in section 5.2. Nonetheless, population density and income have been identified to have the strongest influence on the sustainability rating.

Ranking Significant Findings
In conclusion, the worked out rankings hold a lot of promise for comparing the municipalities of Rhode Island against one another in regards to sustainability.
However, it can not be stressed enough that this thesis largely concludes on the feasibility and potential benefits, as more detailed work with relevant state agencies is required to ensure adequate data quality and choice of indicators. At the very least, this chapter may be seen as a foundation regarding methodology, evaluation procedure and provisional results for the establishment of an official statewide rating system. Furthermore, a snapshot of the current situation is provided along with identification of areas for improvement for the individual municipalities.
For instance, the box plots of figure 3.11 reveal both New Shoreham and North Providence as significantly below average performer regarding environmental aspects.
This observation may then be further analyzed by consulting the results visualized per categories and segments in figures 3.8, 3.9 and 3.10, which identify energy and land use to be particular weak spots for both communities, while New Shoreham is also rated exceptionally unfavorably regarding waste.

CHAPTER 4 Urban Water-Energy-Nexus Evaluation
This chapter is devoted to evaluating potential approaches regarding UWE-Nexus thinking for Rhode Island. As discussed in sections 1.1.2 and 1.1.3 nexus applications aim to increase overall resource efficiency and enhance management by accounting for interactions between otherwise separately evaluated segments, which when applied to urban environments is referred to as the UWE-Nexus. For instance, energy demand for water conveyance may be reduced by limiting the ratio of unaccounted for water as pumping then works more efficiently. Furthermore, runoff from impervious surfaces, which are significantly more prominent in urban areas, conveys a higher amount of pollutants to water bodies, resulting in, amongst other things, a higher effort for necessary purification. These examples demonstrate how such considerations might create further incentives and benefits compared to only focusing on each resource separately. Accordingly, possible approaches will be assessed with the overall target to derive quantifiable measures.
First of all the applied methodologies will be explained followed by examinations of the water usage for energy generation and vice versa. Next, urban interactions will be assessed by evaluating implications from the UHI effect and point pollution sources. Lastly, section 4.6 will summarize the significant findings of this chapter.

General Approach
As mentioned previously, nexus approaches aim to account for all possible interactions between varying resources, which in turn may lead to a more efficient and comprehensive resource management. Furthermore, the specific relationships between individual resources may be highlighted and put into context to the influential parameters of the respective framework. When focusing on dependencies between water and energy supply, there may be rather concealed aspects next to direct implications, for instance energy demand for conveyance of public water.
Those somewhat allusive consequences vary with the respective surroundings and will be evaluated against the urban context of Rhode Island in this paper.
Overall, four different focus areas, which are displayed in figure 4.1, will be assessed regarding possible UWE-Nexus approaches. First of all, the amount of required energy to procure public water in Rhode Island will be analyzed in section 4.3, while the vice versa approach regarding the water demand for thermoelectric power generation will be discussed in section 4.4. Subsequently, section 4.5 features assessments of possible implications due to the respective urban landscape, which includes research for the magnitude of the UHI and pollution sources. While all attained results are closely tied to the state of Rhode Island, they will be put into a larger context to allow broad scoped comparisons.

Methodology
The following section will concisely describe the used methodology for assessing the aspects of potential UWE-Nexus approaches in Rhode Island and are generally split into derivation of land surface temperature and the validation thereof, water for energy examinations and methods for assessing point pollution sources.

Derivation of Land Surface Temperature
This section will demonstrate the process which was applied to derive the Land Surface Temperature (LST) from remote sensing imagery in order estimate the extent of the associated UHI. As the effect is usually most strongly pronounced during summer days, a scene from 6th August 2013 recorded at 3:29 pm, which has a extremely low land cloud cover of 0.58 %, was selected for further evaluations from USGS Earth Explorer. [154] The applied methodology has been derived for the most part from one study about mapping LST from Landsat 8 data with the majority of processing and evaluations being executed with ArcGIS. [155] The chosen procedure is generally referred to as retrieval with unknown emissivity, which is instead derived based of the Normalized Difference Vegetation Index (NDVI) for the scenery. This method assumes that the surface is largely composed of soil and vegetation, which is certainly applicable to the land area of the state, and that the emissivity is linearly dependent on the fraction of vegetation in each pixel of the imagery as displayed in equation 4.5. This method is easily feasible im comparison to other approaches but lacks accuracy as assumptions are made regarding NDVI thresholds and the associated emissivity values. Furthermore, this procedure is rather inaccurate for areas that are primarily composed of soil or contain a high amount of senescent vegetation and can not be applied to surfaces such as water, ice, snow or rocks. [156] Overall, the LST has a strong negative correlation with the amount of present vegetation, which has traditionally been assessed with the NDVI. However, accounting for other contributing factors, such as solar illumination, atmospheric effects, land use or land cover pattern and topography, has great potentially to enhance the accuracy regarding the influence of urban environments on the spatial distribution of temperatures. [157] Subsequently, the work flow to obtain the LST will be described in detail. First of all, the downloaded image was imported into the program and the NDVI was Next, at-sensor temperature is to be obtained via equation 4.2 with the previously determined spectral radiance as an input. Furthermore, the two thermal constants K 1 and K 2 are required. These values are also listed in the meta data and can be found in  The next step is to convert the at-sensor temperature to the LST by applying Finally, the LST can be obtained for each raster cell with equation 4.6, which requires the respective emissivity, the determined at-sensor temperature and the initially used thermal band are required as inputs. In order to achieve a coherent distribution of the LST, the priorly split up raster surface, which was necessary to assign or determine the emissivity values, was merged to a single raster dataset.
At this point, ε represents the emissivity values of the merged raster, which is now constituted of all four different land cover types.   Next, the LST values will be validated against records from weather stations and further processing applications will be evaluated, which is described in detail in section 4.2. Furthermore, subsequent evaluations regarding UHI implications for the state of Rhode Island will be carried in out in section 4.5.1.

Validation of obtained LST values
The obtained LST values will be validated against actually recorded readings from weather stations in Rhode Island with the same date of origin than the evaluated Landsat imagery. First of all, there are going to be discrepancies between the two values as LST and air temperature describe different phenomena. Overall, validation is deemed to be an important step to enhance the reliability of results from remote sensing evaluations, thus significantly increasing the applicability thereof. [156] The determined values will be compared against recorded temperature readings in Rhode Island to put their magnitude into perspective, which goes in accordance to the validation procedure of the used reference methodology. [155] Therefore, data has been extracted from the five NOAA weather stations with available records   can be used to compute the LST, but band 10 was used due in accordance to the referenced methodology. However, as table 4.4 shows band 11 results are actually much closer to the measurements from the weather stations, but as literature suggests band 11 to have significantly higher calibration error it was not used directly in this thesis. [158] But instead of just using band 10 results, the average for each pixel from both methods was calculated in order to better match the obtained results to the recorded data. Additionally, the averaged raster surface was further edited with a one cell circular mean by applying focal statistics on it in order to embed each raster cell in its surrounding parameters. This step was taken in order to account for influences e.g. wind and air movements without generalizing the values too drastically. For the most part, the calculated LST values are considerably above the recorded data with the highest difference coming from station 3 with 5.68 K. This matches well to the validation procedure of the referenced source, where eleven meteorological stations were used for one study area and the highest difference was 5.8 K. Furthermore, the reference source also experienced exceedingly low temperatures, which were interpreted to be clouds or otherwise exceptional events.
Accordingly, the minimum temperature of -2.4 • C, as determined in this thesis, will be considered to be negligible. [155]  In conclusion, the derived LST performs adequately in comparison to the weather stations and matches the deviation as indicated from the referenced methodology. Therefore, the magnitude of potential impacts arising from this spatially distributed temperature will be analyzed in detail in section 4.5.1.

Water Withdrawals for Thermoelectric Power Plants
The USGS water use reports from 2010 and the form EIA-923 have proven to be the most significant resources for assessing the water for energy requirements of the state. However, those two sources are not sufficient for adequately determining measures of withdrawn or consumed water per generated amount of energy. While the latter only includes detailed data for three of the six biggest power plants in Rhode Island, the USGS data shows significant inconsistencies when comparing reports with different years of origin. Therefore, this renders future data collection even more important. However, the achieved results are still sufficient for estimating the situation in Rhode Island.
In general, the United States Geological Survey has been founded as a state agency in 1879 and is tasked with providing high qualitative data for an abundance of fields related to the natural world and the monitoring or mapping thereof. [161] Accordingly, an extensive amount of information is available regarding water and comprehensive reports on water use have been worked out every five years beginning in 1950. [162] The sources for each iteration include national dataset, state agencies, local authorities and individual surveys, resulting in a certain lack of consistency and varying levels of accuracy. However, the reports still hold invaluable information as there is hardly no other elaboration on a national scale [104] Overall, data from 1990 to 2010 has been retrieved for the evaluations of

Energy for Water
This section will examine the current state of energy requirements for water procurement in the state of Rhode Island. As mentioned in section 2.8.2, there are almost 500 individual companies involved in the provision of public water. These range significantly regarding the amount of annually provided water and customers.
As a result, any assessments concerning this matter are extremely complex and extensive due to the required effort for appropriate data gathering. Unfortunately, no comprehensive data reports regarding energy demand for water procurement in Rhode Island existed during the research phase for this paper. However, some key aspects will be explored by the means of literature and suggestions for further research objectives will be worked out.
To begin with, figure 4.4 depicts water supply system, end use and wastewater system as the three main components of urban water systems and also displays the associated connections with energy. Literally each component requires an energy input and only the wastewater system may also generate electricity, for instance, via power generation with methane. Next to centralized water supply systems, which require energy for transfer of raw water and treatment and distribution thereof, there may also be other local sources, such as decentralized rain water harvesting, with differing energy requirements. Large scale operations as managed by public utilities may have different energy intensities depending on climate, topography, use patterns and operational efficiency. [33] Next, the water end use, which includes residential, commercial or industrial, also has an influence on energy intensity. For instance, efficiency of heating appliances and personal habits like choice of shower temperature and duration affect the amount of required energy considerably. [166] Lastly, wastewater systems require energy for collection, transfer and treatment, with the efficiency depending on level of purification, pumping and terrain amongst other factors. Additionally, energy performance may be enhanced by including anaerobic digestion to operate biogas generators for electricity. [167] In addition, the category of externalities covers resources and effort for construction and maintenance of facilities and associated infrastructure. Overall, energy for water assessments have so far been hindered by a large number of dependencies, lack of quantifiable data and missing frameworks and analytical tools, which highlights the importance of potential future research areas. [168]  Aspects and parameters regarding energy efficiency of water supply and wastewater systems will be discussed in more detail, as these areas are best suited to describe entire cities or even broader applications as a whole. Overall, the local conditions and frameworks may differ greatly for each city, requiring in depth analysis with extensive full life cycle considerations for the respective situation to achieve the best possible results. [32] This observation is further emphasized by the vast differences between individual communities, as for instance, water procurement in Melbourne requires only 10 kW h person a while this value increases to 372 kW h person a in San Diego. This difference is largely caused by the employed technologies as Melbourne is able to utilize gravity for conveyance while San Diego has to use inter-basin water transfer systems with a high energy demand. [33] However, even though there is a high range between these extreme cases, general trends and developments can be worked out.
Accordingly, figure 4.5 displays the qualitative influence of several parameters on the energy intensity of urban water systems as observed in a study featuring the cities of Nantes, Turin, Oslo and Toronto. [32] Given the relatively small sample size of the study, the reported result will be compared against a study with 30 featured communities, which unfortunately does not explore the connections between energy and water in depth. [33]  Overall, the four different categories of geography, technology, socioeconomics and climate have been determined to influence energy requirements for water procurement in the four previously mentioned cities. While technology has been attributed with the most factors, climate has only received one, rendering it the least complex influential aspect. Climate is deemed to have a significant influence on water availability, which induces increased energy demand if the overall supply is scarce and more distant or alternative sources have to be used. The same context applies to the geographic attribute of distance, as longer conveyance processes require more energy. Additionally, a diversely shaped topography with steep contours leads to laborious traversal, resulting in a higher energy demand. This is especially true when water or waste water has to be pumped uphill, thus preventing the usage of gravity-fed conveyance. In addition, energy demand for pumping also increases if groundwater is the predominantly used water source. Socio-economic aspects, such as water quality, network size and use or consumption, also influence energy intensity. A higher water quality of the respective source, which is heavily influenced by the activity around the intake locations, enables the utilization of less intricate treatment procedures. This reduces the associated amount of energy.
The same approach applies to water usage with industrial consumption generally requiring more effort regarding waste water treatment and consequently more energy.
Additionally, a larger size of the associated network entails a higher energy demand as the conveyance distance can be assumed to increase with it. Technology features four individual parameters that are largely related to treatment processes or age of appliances. While anaerobic digestion releases methane, which may be used as a fuel for electricity generation, the process features energy intensive aeration and thus is less favorable than trickling filters. Furthermore, sanitation of drinking water with chlorine is less energy intensive than the rather modern method of ozonation with UV-lighting technologies. Accordingly, both older treatment and systems conditions have been found to lead to a lessened energy demand. [32] However, increasing operational efficiency, which generally involves the deployment of novel and new appliances or approaches, has been determined as a prime opportunity for reducing the overall energy demand of water supply systems. [33] Thus, a more up-to-date condition of the overall infrastructure should still result in a more efficient operation and a reduced energy intensity.
Accordingly, the prior discussed aspects will be transferred to Rhode Island in order to work out a qualitative analysis regarding the energy efficiency of public water supply and waste water treatment. Therefore, table 4.5 lists the categories, for which an adequate amount of data and information was available, and the influence on energy intensity along with a brief description. First of all, Rhode Island's topography ranges from sea level to 247 m above and lacks difficult to traverse topography, resulting in a potentially low energy demand for water conveyance. [62] Additionally, about 85% of the overall public demand is sourced from surface water bodies, which requires less energy than groundwater procurement. [103] Both the conveyance distance and network size are utterly limited, due to the small land area and maximum extent of 48 miles from North to South, indicating a comparably low energy demand for these aspects. [62] In addition, both the high water availability and quality thereof, which significantly reduces the required treatment of raw water, point towards a low energy intensity. [ Topography sea level to 247 meters above positive [62] Water source 85% of drinking water from surface positive [103] Distance Compact state, small land area positive [62] Water availability Sufficient supply positive [103] Treatment process Majority treated with chlorination positive [105] Water quality High quality drinking water supplies positive [103] Network size Limited extent positive [62] Use & consumption High share of residential use positive [104] System condition One of the oldest systems nationwide negative [103] Overall, as eight of nine aspects indicate a favorable situation, the procurement of public water and the treatment of waste water potentially entail a below average energy demand. Accordingly, the respective facilities and services should exhibit a relatively low energy intensity in comparison to other cities or regions with more unfavorable conditions.

Water for Energy
Contrary to the preceding section, where energy demand for water procurement has been discussed, there are promising data resources available for researching the water demand regarding the generation of electricity in Rhode Island. However, even though the three sectors of electricity, thermal and transportation are virtually equal for the overall energy demand in Rhode Island, generation of electricity and the according water requirements will be the focus of subsequent evaluations.
Additionally, the applied methodology and data issues are discussed in section 4.2.3.
First of all, in order to provide an adequate frame of reference, water usage in Rhode Island and its development since 1990 will be discussed in detail by using  There are clearly discernible regional differences, as for instance all New England states, with the exception of Maine where industrial usage accounts for the highest share, water for thermoelectric power generation is the most significant category with public supply ranking second. This results in a clearly different distribution in New England in comparison to the entire USA, where water for energy generation remains the biggest sector but irrigation, which is heavily tied to agricultural activity, accounts for a much higher share overall. In conclusion, water usage may differ substantially by state or region and thus knowledge about the individual situation is crucial for determining advisable future steps and measures regarding water supply. For instance, researching and implementing more efficient irrigation technologies or procedures is likely to yield more benefits in California or Florida than in any of the states of New England. [104] Furthermore, Rhode Island exhibits the lowest generation efficiency regarding water withdrawals with 283 l kW h . However, this measure can be considered to be unreliable, as literature suggests much lower values in general and about 52.6 l kW h of total withdrawal as a worst case scenario for natural gas power plants. As will be described in more detail later on, all major power plants of Rhode Island utilize natural gas with a combined cycle technology, which performs favorably regarding water requirements per generated electricity in comparison to other technologies. Accordingly, table 4.6 lists the requirements for chosen generation methods differentiated into closed-loop cooling, which generally deploys cooling towers, and open-loop cooling. In comparison, rather conventional fuels, such as nuclear or coal, have a considerably higher water demand than natural gas, which is only inferior to the renewable technologies of solar and wind in that regard. Additionally, the deployed cooling technology is highly influential. This is emphasized by the fact that natural gas power plants with recirculation or closedloop cooling can be estimated to require only 1.6 l kW h in total, while open-loop or once through procedures may amount up to 52.6 l kW h . Furthermore, consumptive usage, which includes water being temporally lost due to evaporation, is highest when deploying closed-loop cooling. [23] [99] In general, consumptive water use of the power sector in the USA amounts to 3,310 MGD or about 2.5% of total withdrawals, indicating consumption to be the less influential category besides the sheer amount of withdrawals. [169]  close to the worst case withdrawal of 52.6 l kW h , as suggested by literature. However, evaluating water requirements for thermoelectric power generation is highly important in Rhode Island, as it has consistently been the major demand sector since 1990 and has never accounted for less than 62% of the total withdrawal. In conclusion, other resources besides USGS data need to be consulted for adequately assessing water demand for energy generation. [104] [23] So, as the USGS figures are not well suited to assess water withdrawals per    subsequent cooling of the steam to retrieve and reuse the inherent water. [170] Overall, natural gas power plants promise a more beneficial operation regarding emissions of greenhouse gases, in comparison to conventional fuels such as coal.
[171] Additionally, they generally require the least amount of water for cooling next to renewable technologies, which often do not have an operational water demand.
However, the deployed cooling technology has been determined to be the decisive factor for adequately assessing the overall water demand of power plants over the fuel type alone. [172]   rate of 3,007 gallons minute . The recirculation along with natural draft cooling towers lead to a significantly reduced maximum intake in comparison to the Manchester Street power plant. This facility lacks recirculation of cooling water and uses cooling ponds, resulting in a maximum intake rate of 181,777 gallons minute . The significant difference between the two technologies becomes even more apparent when the associated energy production is taken into account. Peak generation and thus maximum cooling water intake over a whole hour leads to a water requirement of 1,25 l kW h for the RISEC facility and 154,04 l kW h for the Manchester Street plant. This dramatically higher water demand per amount of generated electricity indicates that the latter facility is more likely influence local water systems more severely.
Furthermore, assessing these two facilities in more detail should lead to a better understanding of the water demand characteristics. Accordingly, more detailed examinations regarding seasonal development have been carried out with the monthly specifications for electricity generation and water withdrawal from form EIA-923. As there are occasional recording gaps, the data was averaged over three years beginning in 2013 to achieve more consistent results and to reduce the influence of outliers.  shows the three year averaged values for generated power, cooling water withdrawal, difference between mean water intake and discharge temperature and the withdrawal per generated electricity. Additionally, the monthly minimum and maximum values are also displayed to provide further reference points. This approach allows to determine coherent or diverging development between the featured datasets and allows for comparison of seasonal variability. Overall, there are clearly discernible seasonal differences, as for instance the power generation and the cooling water withdrawal both tend to peak during the summer months.
However, the two remaining measures show different behavior and are at rather average levels during the summer. Furthermore, while the mean difference between intake and discharge temperature of the cooling water peaks only in the spring, the efficiency of water withdrawal for energy generation is at low levels both during the spring and autumn. This indicates an especially unfavorable usage pattern outside of summer months as more water is required per generated electricity. Additionally it recedes to average values during the winter months. The later is largely due to a bigger relative decrease of generated energy in comparison to withdrawn water, resulting in a higher cooling water demand if normalized to l kW h . to realistic values in l kW h in comparison to literature references. Accordingly, the quality of data has been determined as a major area for improvement to enable comprehensive assessments, which has been deemed to be especially applicable to natural gas fired power plants with combined cycle technologies. [172] However, the evaluation of this section are sufficient to indicate promising areas for future research. For instance, evaluations of the six biggest power plants may serve as statewide characterizations, leading to a less excessive data gathering process. Additionally, both the electricity generation and the associated water withdrawals are at high levels during the summer time, thus further intensifying stress on the respective water sources and the aquatic, natural environment.

Urban Interactions
This section will assess potential implications on energy and water provision due to pressures posed by urbanized areas in Rhode Island. This will involve examining the extent of the UHI effect via evaluation of the derived LST values and scrutiny of point pollution sources. Both aspects will be spatially related to public supply infrastructure to assess the magnitude of possible impacts.

Urban Heat Island Implications
As discussed in section 1.1.3 urban environments simultaneously put pressure on the associated environment and alter ambient conditions within their respective confinements. As a result, they form an intricate network of interactions and conditions such as elevated surface and air temperatures via the UHI effect. Assessing its extent and consequences for the state of Rhode Island is the main objective for this chapter. Therefore, the UHI and its spatial distribution has to be determined first.
While section 4.2.1 discusses the applied methodology, the mapped results, which can be seen in figure 4.2 on page 135, and the validation procedure, the following examinations focus on characterizing the effect and the associated implications.
First of fall, the magnitude of the UHI is most concisely described by comparing the average temperatures of corresponding urban and rural areas to one another. This has been done by averaging the computed LST for the urban areas declared by the USCB and the remaining rural territory, which results in a 4.3 K higher temperature for urban areas. Given that the featured scene, from which the LST was recorded, is dated to mid August, thermal complacency is already challenged by generally high temperatures, which is even more relevant to cities and towns due to the UHI. Furthermore, elevated temperatures entail an abundance of other detrimental influences, which will be discussed in more detail in the closing remarks of this section. However, as the featured LST determination procedure is rather basic and lacks the sophistication of advanced and increasingly complex methods, there are some dissonant data points. For instance, the overall minimum temperature is -1.62 • , which is far too low for ambient summertime conditions. The results have been used regardless, as the large scale average values are acceptably close to actually recorded readings of NOAA weather stations and to the expected accuracy of the reference methodology, as discussed in more detail in section 4.2.2.     connection between settlement density and LST seems to weaken eventually, as the communities with less than 10% of the maximum value, which is equivalent to roughly 1,500 people mi 2 , show a variable development in both the respective share of affected people and the mean LST values. However, beginning with the town of Burrillville, the ratio of urbanized inhabitants and the associated temperature thresholds decrease significantly. Furthermore, this area almost without exception incorporates all LST values between 16 and 18 • C, indicating drastically cooler ambient conditions in the communities with a rather rural character.
In conclusion, there seems to be a strong connection between population density and LST, with urban versus rural community set-up being a less significant driver overall. Altogether, elevated temperatures due to the UHI effect may potentially affect the majority of Rhode Island's population, as the state's average LST has been determined to be about 19 • C while over 80% of the inhabitants live in areas with 20 • C or more and about 48% in areas with 25 • C or more. This discrepancy indicates potentially high benefits of future UHI related research for the state of The observed correlation between the share of developed or forested area may to a certain degree stem from the derivation method, which evaluates the spatial distribution of a NDVI raster dataset and thus utilizes differences in distribution  [173] Overall, the priorly discussed findings may benefit greatly from evaluating additional cases while taking differing seasonality into account.
Additionally, there are plenty of incentives to evaluate the potential degree of the UHI effect in Rhode Island even further, as elevated temperatures may entail an abundance of detrimental consequences. Next to rather obvious connections, such as increased energy demand for cooling of indoor spaces or rising heat-stress induced mortality, there are even more far reaching implications like increased smog formation and a subsequently deteriorating air quality. [29] [174] Potentially all areas of public life may be affected and the impacts on water resources have been determined to be highly diverse. The impacts range, for example, from an increased residential demand to impaired quality of water bodies and easier transmission of waterborne pathogens, such as salmonella or cholera. [175] [176] [177] Accordingly, in order to further highlight the potential gains of UHI mitigation, the associated decrease on summertime average water demand will be evaluated.
Therefore, significant findings for the city of Phoenix, for which the associated UHI and its implications have been thoroughly researched, will be transferred to Rhode Island. For instance, it has been determined that each additional increase of 1 • F causes an overall raised water demand by 0.8%, or 1.44% per 1 • C, for the entire city of Phoenix in June. [178] Assuming that this correlation can be These reductions are of a high significance, as management of increasing summer peak water demand has been identified as a major challenge for the water supply of Rhode Island. Additionally, water scarcity and stress on natural water bodies is already a significant influence factor during the summer, rendering further reduction efforts highly important. Furthermore, the PWSB supplies a majority of municipalities around Providence, which may benefit even more from UHI mitigation as they tend to exhibit extremely high temperatures, and rely heavily on the Scituate Reservoir. In fact, the four municipalities that are directly associated with the PWSB, namely Johnston, Cranston, North Providence and Providence account for 90.59 MGD of the summertime demand, which can be reduced to just 80.89 MGD. As a result, the stress on the Scituate Reservoir may be significantly reduced and the associated withdrawal may even fall below the estimated safe yield of 83 MGD. [103]

Point Pollution Sources
Lastly, in this section impacts of point pollution sources in the context of an Urban-Water-Energy-Nexus approach, such as increased effort or energy demand for water treatment due to heightened levels of pollutants and emissions, will be evaluated. For the most part, this will include spatial evaluation of TRI facilities in relation to significant areas of the public water supply infrastructure. Additionally, land use features, e.g. impervious surface and major roads, will be allocated to the aforementioned areas.  Overall, the TRI facilities are predominantly clustered in close vicinity around Providence and a few can also be found in the rather remote rural areas. However, the relatively highest emitters are located outside of the aforementioned cluster.
Regarding air releases, the most significant pollution source, which is responsible for 43.81% of all pollutants and belongs to Ocean State Power, is located the north western area close to the Massachusetts state border. In addition, a facility belonging to Newport Biodiesel is the second highest emitter with 20.86% of all air releases and can be found in the south western Aquidneck Island region. Regarding toxic water releases, which are significantly lower in comparison with only 732 lb in total, the most significant pollution source is a lumber processing plant located south of Providence and accounts for 68.94% of all respective releases. Additionally, a company called Toray Plastics is the second highest emitter with 22.98% regarding surface water pollutants and is located in the same area as the priorly discussed facility. Even though the amount of toxic compounds released in water bodies is relatively low in comparison, the two most significant facilities are located in close proximity to one another, rendering more detailed investigations advisable. Overall, the spread of toxic air releases is significantly influenced by wind direction, which may vary greatly on a daily basis in Rhode Island. Due to this lack of predominant wind direction, trends regarding the dispersion of these pollutants can not be formulated with a high level of certainty. However, as most applicable facilities are clustered around of Providence, a high share of inhabitants may be affected by the released substances, highlighting the importance of ongoing monitoring and research. [124] [70] In addition, figure 4.13 also displays the distribution of major roads, for instance US routes and connectors, and impervious surfaces, which have been identified to cause an abundance of detrimental consequences especially in relation to water quality. In general, about 12.94 % of the state's land area is composed of impervious surface with a distinct agglomeration thereof at the northern end of Narragansett Bay around Providence. However, the concentration of this specific land cover type tends to diminish further inland, where most of the relevant protection areas are located. The same trend can be observed for major traffic ways, which tend to be further dispersed with increasing distance from the coast and specifically the city of Providence. Overall, this should result in decreasing stress or detrimental consequences on the protection areas due to urbanized areas with increasing remoteness. This is further emphasized by the fact, that runoff

CHAPTER 5 Evaluation
As stated earlier, this paper aims to examine sustainability in the state of Rhode Island while focusing on two major objectives, which are as follows: • Compilation of a comprehensive municipality ranking • Examination of potentials regarding UWE-Nexus approaches Accordingly, the study region of Rhode Island has been described in great detail as a prerequisite for the aforementioned assessments with special effort spent on utility infrastructure and other relevant sectors for research objectives. The following three sections will state a summary of the study findings, a discussion regarding areas for improvement or advisable future steps and the finalized conclusion.

Summary
The worked out rating features 75 indicators, which are allocated to sixteen thematic segments and the three categories of social, environmental and economic.
Accordingly, it is closely matched to the initially discussed definition of sustainability.
While Jamestown has been identified to be the most sustainable municipality, Central Falls is rated most unfavorably overall. Furthermore, the latter takes up a particular spot amongst all communities as it attains by far the lowest economic rating. The same observation applies to New Shoreham, which ranks lasts regarding environmental aspects and shows an even bigger margin to the second worst rated town of North Providence. Next, the rating is well suited to reveal specific areas of concern for individual communities, for instance, waste for Johnston, New Shoreham and Providence and education for Central Falls and Pawtucket. Additionally, unfavorably rated municipalities seem to be clustered around Providence, while the southern shore of Rhode Island tends to house above average performing municipalities for all examined categories. Lastly, of the evaluated demographic parameters, population density and income have the most influence on the ranking.
While rating results tend to diminish with increasing population density, they tend to increase with per capita income of the municipalities.
Regarding the UWE-Nexus aspects, a high potential for beneficial research has been determined for water requirements due to electricity generation. This usage category has by far the highest water demand and has accounted at least for 62% of total water usage since 1990. Additionally, the six biggest power plants of the state, which are all fueled by natural gas, have significantly different cooling technologies ranging from maximum intake of 3,007 gallons minute for the RISEC facility to 181,777 gallons minute for the Manchester Street power plant. Furthermore, as energy generation peaks during the summer months, the corresponding cooling water withdrawals rise accordingly, putting additional pressure on the local environment where water is already scarce during the summer time. Unfortunately, no appropriate data records were available to research the energy requirements for water procurement, which may require extensive effort for data gathering as there are nearly 500 individual water supply companies in the state. However, qualitative assessments of parameters related to water supply and waste water treatment indicate a favorable situation regarding energy efficiency for the these two sectors.
Furthermore, potentials impacts of the urban landscape have been evaluated largely by determining the magnitude of the UHI, which amounts to roughly 4 • C during the examined afternoon in August, and by assessing the spatial distribution of pollution sources. Due to the elevated land surface temperatures over 40% of the state's citizens are potentially affected by temperatures above 25 • C, which amounts to more than 80% for areas over 20 • C, even though the statewide mean temperature has been determined to be around 19 • C. This indicates a high level of potential impacts ranging from heat stress, higher energy and water demand and an easier transmission of pathogens. Furthermore, complete mitigation of the UHI promises a water demand reduction to 193.70 MGD in the summer, which may be sufficient to meet the safe yield requirements of the Scituate reservoir.
As a result, research of appropriate mitigation approaches is deemed to be very beneficial for Rhode Island. Additionally, allocation of facilities with toxic release, major traffic-ways and impervious surfaces to protection areas related to public water supply has revealed wellhead protection areas to be under the biggest threat by urban development. These areas exhibit a exceedingly high share impervious surfaces, road mileages and toxic releases per area.

Discussion
This section will assess areas for improvement and state requirements or advice for further work on the examined subject matters. Both the ranking and nexus assessments will be discussed separately, while the overall conclusion in section 5.4 will highlight the most significant findings and consequences thereof.
While the worked out sustainability ranking was constructed to cover all relevant aspects and includes 75 individual indicators, it is best seen as a starting point from which the state offices or interested personnel may draw conclusions and experiences for implementation of a more refined ranking. Furthermore, periodic reassessments with tangible goals and measures may greatly help to improve this snapshot of the current situation. In turn, this approach has a high potential for promoting sustainable development in Rhode Island and for laying out a road map from which the conditions, that may affect all areas of everyday life, can be incrementally improved upon. Accordingly, the worked out ranking already excels at depicting individual areas of concern and providing the respective reasoning by detailed evaluations of the associated indicators.
Chapter 2 aims to give a comprehensive overview for the entire state while focusing on key areas for interest in section 2.8. Environmental aspects and utility infrastructure have been most prominently discussed, while other important areas regarding economic and social aspects are only briefly analyzed. As a result, feature reports shall aim to include all relevant sectors equally and may benefit greatly from the in-depth introduction to the state and the environmental discussions.
Unfortunately, even though extensive effort has been spent on compiling the current version of the ranking, some areas still require further refinement. This is especially true for a lack of data regarding energy and water demand and economic measures such as number of businesses and employment opportunity. Additionally, this study would further benefit from an analysis of the political aspects, which could be carried out by reviewing the comprehensive plans per community. Next, the score per indicator is distributed based on the respective best and worst performing communities, resulting in attributing zero points per measure to some communities.
While this approach works well for comparing the municipalities to one another, it might result in occasionally too unfavorably rated indicators. Overall, future iterations should aim to fill in the gaps of available data, include political aspects as a fourth category and establish independent reference thresholds instead of referring to the minimum and maximum values per indicator.
Additionally, as stated in section 1.1.1, evaluations of sustainable development should aim to be repeated in periodic intervals in order to measure change and evaluate goals or adapt them accordingly. About half of the used indicators stem from USCB resources, which are published on an annual basis as part of the ACS.
Furthermore, a few other figures, such as municipal waste and housing reports, are also revisited each year, indicating a proficient basis to repeat the rating over several years. However, some indicators are likely not changing very much on an annual basis, which applies for instance to public transportation infrastructure, number of LEED projects and renewable energy assets. Accordingly, it is recommended to revisit the rating about every five years, which would also leave enough time for the municipalities to evaluate the results and hopefully implement beneficial measures. UWE-Nexus approaches require an extensive amount of information, which was lacking at times during the research process for this thesis, in order to properly assess the connections between energy and water in the chosen study area. This is applicable to the energy for water measures, which have largely been examined by applying already existing literature findings on which to assess the current situation in Rhode Island. In turn, verification of the worked out findings with actual recordings is a strongly advised future step. On the other hand, there is plenty of data available regarding water for energy measures, such as USGS water use figures and water withdrawal information for the individual power plants from the EIA. As a result, water for energy relationships could be examined in more detail, but lack consistency to literature reference. Accordingly, a close cooperation with power plant operators is advised for future research in this area. Assessments related to the UHI may be improved upon by implementing more sophisticated derivation procedures, thus increasing the accuracy and broadening the applicability of the carried out evaluations. This may be done by increasing temporal resolution and evaluating days with differing seasonality. The latter is especially emphasized as findings from the city of Phoenix, which is exposed to vastly different ambient conditions, have been used. Lastly, determining actual consequences related to the pollution sources may help to create further incentives for the mitigation thereof.

Limitations of this study
This section discusses the appropriate scope of the study and the inherent concerns regarding applicability or otherwise noteworthy issues. Accordingly, controversial aspects and potential future research objectives will be highlighted.
To begin with, rising income levels have been interpreted to be more sustainable regarding economic aspects as higher affluence allows more spending, thus increasing economic activity and attractiveness of the respective area. On the other hand, this may also lead to a higher degree of consumption, which in turn leads to increased stress on the associated supply framework and a more unsustainable situation overall. Accordingly, future research in this area should explore this relationship and correlations between sustainability and demographic parameters in more detail.
Overall, this is thought to be highly beneficial for deriving appropriate indicators and refining the rating approach carried out in this thesis.
As discussed in the introductory segment of this paper, urban areas are centers for cultural activity, as they offer access to many vital services and institutions, and resource consumption. Accordingly, appropriately accounting for both of these aspects within sustainability rating schemes will likely result in a more holistic assessment. In this thesis, the rural areas of Rhode Island have been rated more favorably than the urban centers of the state. At the moment, it is up for debate if this observation stems from the applied procedures or if it accurately depicts differences between rural and urban areas. This also raises the question of how to design assessments that equally account for both the associated benefits as well as the inherent disadvantages of consumption patterns. These may often be allocated to diverging locations, such as commuters that cause emissions at their respective place of employment instead of their community of residency. Resolving this issue may potentially lead towards a more appropriate score allocation by highlighting the contribution of urban areas to the overall assets and activity within the state.
Future iterations should explore other opportunities for reaching appropriate results besides normalization of the indicators.
Additionally, the sustainability rating results are closely tied to Rhode Island.
However, the scoping is deemed to be transferable to other areas within the USA, as the goals and points of emphasis should be quite similar. On the other hand, the focus and featured areas should be adapted when evaluating communities in other countries. Overall, availability and quality of data is the key concern for this kind of assessment. When focusing on the USA, detailed demographic parameters are published by the USCB on an annual basis, providing an outstanding foundation to evaluate most social and some economic aspects. However, the environmental assessments of this thesis are based primarily on publications from state agencies with a few federal resources being used as well. Accordingly, successful transferal to other states is dependent on the level of detail and availability of publications.
Overall, a majority of the states should already have appropriate data records available, as environmental regulations are generally applicable nationwide. Additionally, federal agencies usually publish their work on a county level. Accordingly, rating approaches in other parts of the USA may benefit from referring to counties instead of municipalities, which would also drastically lower the required effort for evaluations in the bigger and more populous states.
Lastly, the sustainability rating may be improved by including more directly applicable scores and indicators, which would help to inform municipalities on how to improve their respective situation. This can be done by reducing the number of indicators, thus decreasing the amount of effort to derive tangible measures for each community, or by compiling separate overviews for each municipality. The latter would provide a central accumulation of data on which to comprehensively assess and evaluate each individual situation.
Overall, the worked out ranking is one of few iterations primarily focused on rather small communities while also accounting for the entirety of a whole state. Accordingly, there are unresolved or unproven issues and areas for improvement, which have been thoroughly discussed in order to provide a strong basis for improvements regarding future research in this field.

Conclusion
In conclusion, the proposed sustainability ranking is deemed to be a beneficial tool to evaluate all communities regarding their social, environmental and economic sustainability, and the respective individual areas of concern in a comprehensive manner. Furthermore, periodically recurring revisions may greatly promote sustainable development and thus improve the sustainability of the state as a whole while focusing on the municipalities. Furthermore, they are deemed to be the best suitable frame of reference in Rhode Island to provide a close link to local government and the affected residents. Overall, this approach can be seen as a foundation to improve upon and attain a higher quality of life for the state's citizens.
As Rhode Island features a highly urbanized composition and densely populated settlement structure, interactions within the urban landscape may have far reaching implications. Elevated temperatures due to the UHI effect have been identified to potentially affect a majority of the state's citizens, indicating high benefits via implementation of appropriate mitigation measures. Additionally, allocation of pollution sources and the development of counteractions may play a major role in increasing the resiliency of public water infrastructure. Furthermore, proper quantifications of water demand for energy generation and energy requirements for water provision are deemed to reveal the interactions between these sectors, thus highlighting the true cost of supplying for the state's population.   To attain the PM10 standard, the 24-hour concentration at each site must not exceed 150 g/m more than once per year, on average over 3 years. F To attain the PM2.5 annual standard, the 3-year average of the weighted annual means of 24-hour concentrations must not exceed 15 g/m3. G To attain the PM2.5 24-hour standard, the 3-year average of the 98th percentile of 24-hour concentrations at each population-based monitor must not exceed 35 g/m3.

APPENDIX B
Ranking Scores per Category and Segment The following tables hold all ranking values for the overall score, the three categories and the twelve segments.  The following figures display the data which was used to derive the water demand per municipality as described in detail in section 3.2.2. Next to the values from figure C.2 public supply data for New Shoreham and the Richmond Water Supply System were retrieved from the strategic plan worked out by RIWRB, which has otherwise only been consulted for self supply estimates. [103] • New Shoreham public supply 0.1 MGD