Urban Water Demand - Report Example

Urban water demand Name Institution Date Course Urban water demand Introduction Today, the world’s population has reached 7 billion, and from a few decades ago, more and more people have moved to live in the cities. Such rapid and sustained increase of people in the cities has been a major concern for water system managers and planners as cities continue to increase their water demand. Water is an important natural resource for the growing urban areas. Residential, industrial and commercial users continue to increase demand for this resource, which will usually require treatment, and may be located far away from the cities. In urban areas, there is increased competition with industry and agriculture a cities increase in physical size and political influence. The demand for this resource is still expected to double by 2050 and the competition among rural, peri-urban and urban areas expected to worsen (UNDP, 2006). This paper will discuss water and its demand in the cities around the world. The importance of understanding the water requirements, the factors that influence the demand and the methods for effective forecast and planning of the demand have been discussed in detail. Water demand models continue to give planners insight into future water demands in the cities, and as has been discussed later in the paper, water managers now explore different models to achieve specific demand requirements in their service to clients. The paper first discusses demands for various cities around the world before exploring a number of water demand models that have been successfully implemented in cities around the world. A. Comparison of urban water demand in different countries Singapore Over the pasts few decades, Singapore has made significant efforts towards creation of a comprehensive management system of its environment with organised water supply, controlled pollution of the rivers, proper planning of industrial estates and a state of the art sanitation system. Sustainable water supply has been at the fore front of the management effort and the country has managed to achieve great progress in the management of its water system (World Bank, 2006). With a population of about 4 million in 2006, the country boasted of highly developed industrial, financial and business services, but suffered serious shortage of water resources. The country’s water demands were growing and by 2006, the government’s main concern was how to ensure that the citizens access clean water, with the demand standing at about 1.36 billion litres every day. The country could only supply 50% of this demand at the time. The challenges of water supply were not because of lack of rainfall (rainfall was 2400mm/ year), but due to the scarce land and space to store the rain water (Tortajada, 2006). The country imported much of its water from Johor state in Malaysia. By 2010, the county’s GDP had increased from USD 428 in 1960 to USD 43,867 and there was increased demand for resources. The country had only two national taps and the supply depended on catchment and the imported water from Malaysia. While still relying on the bucket system, several homes were not sewered yet and there were high volumes of unaccounted for water. Today, however, 4 national taps are operational: water from local catchments, NEWater, imported water from Malaysia and desalination (Tortajada & Joshi, 2006), supplying the domestic daily water consumption that stands at 155 litters per day per capita. Investments and proper management of the water in this country has led to 100% service by modern sanitation system and only about 5% unaccounted for water with integrated urban and water planning aimed at achieving enhancement of quality of the environment. Brazil cities Brazil has been known as the country of abundant waters. It has the highest total renewable fresh water supply in the world with an average availability well over 1700m3 per person per year. However, the country’s estimated figure of 6950km3 per year of fresh water is not equally distributed. In fact, about 70% of this water is in Amazon Basin, an area with about 7% of the population. Water remains critical in the nation that greatly relies in agriculture and industrial processing. The highest water demand is experienced in the agriculturally active areas and about 54% of the country’s demand is used for irrigation. Urban demand accounts for 22% of the country’s total water demand. Great demand is experienced in the major cities around the country. The metropolitan region of Sao Paulo, for example, has a demand of about 57m3 / second of drinking water for its 16 million inhabitants. This demand is supplied from six large production plants in the region, with nine reservoirs, pipelines and tunnels. Due to the rising populations around this city, there has been a growing need re-organize the planning and development of the supply system and management of the rising pollution in the natural sources (Porto, 2001). Jordan Valley This region indicates two seasons of demand and supply. The summer season begins at the start of April until the end of September while the winter season lasts from beginning of October to end of December and again the period from start of January to the end of March of the following year. Two main demands are dominant here: the urban demand in Amman city as well as the agricultural demand that is separated into North JV, Middle JV and South JV. In Jordan, agricultural activities consume about 600 million cubic metres (MCM) of water every year. A third of this water is supplied to Jordan Valley. Of the total supply to Jordan valley, 50% is reclaimed water while about 20% comes from freshwater sources in the valley. The municipal consumed about 290 MCM in 2006, with 42.6% of this being pumped to Amman Governorate while about 1.27% supplied to Ajloun. Out of the total supply, 111 MCM was treated in wastewater treatment plants. Close to 40% of the population was not connected to the sewer at the time so that much of the affluent was lost without reuse or recycling. Amman city has witnessed significant growth of population which was about 1.6 million in 1994 to about 2 million in 2004. By 2012, official estimates indicated that the annual water demand in the city stood at 51m3 per person per year but there are suggestions that it could be more. Current estimates show that the city uses around 102 Mm3 every year (equivalent to 280,000m3/day) of water. This demand is supplied from wells (60%) and the Zai WTP (40%) (Alfarra et al, 2012). Other cities Several urban areas consume greater volumes of water as compared to their available water sources and usually depend on imported water from neighbouring areas. The suburb Stenlose Syd in Denmark, for example consists of about 750 households with the city’s annual demand standing at 90, 000 m3. Of this demand, only about 23, 000 m3 was supplied from the rains through the rooftop collection and the rest of the demand satisfied through importing from the public water supply. For a city like this, the self-sufficiency ratio is calculated as 23, 000/ 90, 000. Berlin, on the other hand meets about 70% of its water demand through re-use and this remains the major source of water within the city. The remaining 30% demand is imported from external sources. Over the years, cities have strived to reduce dependence on importation of water and thereby increasing their self-sufficiency significantly by employing new technologies in water reclamation, rainwater and desalination. Sufficiency ratios therefore vary greatly in major cities from 15% in Orange County California to over 80% in Pimpama-Coomera in Australia (Rygaard, Binning & Albrechtsen, 2011). B. Identify factors that can influence water demand 1. Urbanization The 20th century has witnessed the most dramatic urbanization of population both in the industrialized and developing world. Globally, according to the authors, cities have added 2 billion and more people from 1950 to 2000 and a further 2 billion will be added to the cities by the year 2025. The rate of urban growth has been overwhelming such that African countries that were only 34% urbanized are expected to be 54-55% urbanized by the year 2025. These growing urban populations, increased activity, and increased industrialization have created greater demand for natural resources in these cities, water being one of the most demanded resources. The concentration of people in the cities is one major cause of the increased water demand for household use as has been discussed in the next section, but also significantly important has been the increased urban economic activity (McDonald et al, 2014). Industrialization creates great demand for water since these establishments require large quantities of clean water for industrial production. In China, for example, urban water demand was projected to grow 60% over a period of ten years from 50 billion cubic meters to 80 billion cubic meters while India’s demand is expected to rise to 52 billion cubic meters by 2025 for domestic use. Industrial production and energy generation water demand in India that is about 67 billion cubic meters is projected to increase to 228 billion cubic meters by the same year (United Nations Population Division, 1996). Also growing along with urbanization has been agricultural activities around the major cities. Global figures indicate that agriculture is that largest water consumer, especially in the developing countries. Irrigation has increased and continues to serve the food demands of the burgeoning city populations. Estimates indicate that about 17% more water will be required by 2015 for irrigation purposes to meet the global food demand (World Bank, 1998). 2. Population Population is one of the major factors that will influence the demand of water any city or country. Marella (1992) highlights three elements of population that had greatest impact on the water demand in Florida: magnitude of resident population, percentage of households getting the service of the public supply water system and magnitude of the nonresident population. Magnitude of Resident population After a study to determine factors influencing water demand in Florida, Marella (1992) stated that since residential use represents the largest use sector for public-supply, it can be expected that resident population has the most significant influence on water demand in the city. The author revealed that over the years 1950-1987, the demand of water in the city and the city population increased concurrently with a correlation coefficient of 0.9993. With this result, the author maintained that an evaluation of the residential population patterns is among one of the most critical factors in determination of public water demand and its supply Population served by public-supply systems As the population rises, more public supply systems also came up in Florida, increasing from 1, 400 to 2, 300 in ten years. In response to this increase, several public and private plants increased their production and expanded sewer and water services into those areas initially unincorporated. In 1970, for example, about 40% of the city’s population did not live within incorporated areas while 76% of the population was served by the public system. When in 1987 about 49% lived in unincorporated areas with 86% of the population served by public system, there was direct increase in demand with the facilities having to increase capacity (Marella, 1992). Nonresident population This is another factor for concern in big cities. Florida’s nonresident population rose from about 18.8 million visitors to about 34.1 million visitors between 1977 and 1987. This increase was closely followed by the opening of more amusement and recreational facilities. These visitors may be those staying for a short period or they may stay for a while. The influence of these visiting populations is majorly experienced in commercial deliveries and they may cause extremely high demand depending on the season or period of the year. The result could be high demand during certain months and a drop in demand during other months. Those visitors staying for long may cause a longer demand on the water supply system (Marella, 1992). 3. Climate Ada & Carl-Erick (2006) argue that climate and weather patterns have a significant influence in the consumption of water, hence its demand. According to the authors, more water is likely to be consumed when the weather in hot and less will be consumed during the rainy seasons. Marella (1992) also mentions precipitation and temperature as the two major variables with the strongest influence on water demands mainly because of the quantity of residential water used in lawns and garden. In dry weather, great amount of water is used to water plants in domestic gardens and lawns as well as out in the fields. Irrigation has remained a major exercise during dry weather and consumes enormous amounts of water as farmers try to keep their crops hydrated. During wet weather, not irrigation is done and therefore the demand on water is significantly reduced. Increased populations have led to increased need for agricultural production due to high demand for food stuffs. Such increase of agricultural farming land has a direct influence on the water needed for irrigation. This has been the trend in and around major cities as farmers step up their efforts to feed the growing masses. Another study revealed that intraurban differences, as far as affluence and vegetation was concerned, greatly affected water consumption by the residents. Hillsboro, Oregon and Phoenix, Arizona exhibit significantly different climates and urban form. The cities have neighbourhoods that are characterized by high irrigated land scale proportions, big housing lots and high incomes. However, the neighborhoods display great sensitivity to weather conditions and climate, which has been found to contribute to their variations in household water consumption (House-Peters et al, 2010). 4. Household water amenity and facility In a study to determine causes of difference in water demand between Beijing and Tianjin cities in China, Marella (1992) found that facilities available greatly contributed to the difference in demand. The urban population in Beijing and Tianjin is served by tap water, but the households in Beijing enjoyed diversity for their sources of drinking water. Although tap water was majorly used as drinking water in both cities, other alternatives like bottled water were gaining more popularity in Beijing. More households in Beijing also had other facilities like shower, washing machines, dishwashers, hand basins, flushing toilets and many others as compared to Tianjin. This factor was considered a major contributor to the different water demands in both cities. While this was considered true generally, households with comparable facilities from the two cities still revealed more consumption (about 40% more) from households in Beijing. This difference was attributed to the different water use habits in the two cities. 5. Household water uses and behavior In further analogy between the two cities, Zhang (2005) mentioned the critical role played by habits and behaviours in determining city water demands. Much of the water supplied in Beijing and Tianjin is used for drinking, aquariums laundering, cooking, and bathing while outdoor uses like car washing and gardening are largely minimal. The study revealed that Beijing households used more water in cooking (5% more), toilet flushing (3.5% more), bathing (6% more), and car washing (3.6% more) as compared to Tianjin households. The result, as found by the study, revealed that on average, Beijing families used 60% more water that Tianjin families. Beijing residents were found to take more showers that their Tianjin counterparts. In the summer, Beijing residents spend even more time in the shower than Tianjin residents while in winter they take shower more frequently. The use of washing machines is also becoming more common in both cities. Although more Tianjin residents still use their hands to wash clothes than in Beijing, a mixture of both hand wash and machine wash is becoming popular. Other household factors linked with demand of water at the household level include income, sex and educational level (Ayanshola, Sule & Salami, 2010). These water use differences are also noticed in water conservation practices. Generally, in every category, about 6% more families in Tianjin reported putting efforts to conserve water than in Beijing. Tianjin households engaged in re-use of water for other requirements like using washing or kitchen water for watering the gardens (58.6% of the families reported doing this) and for flushing the toilets (55.7%). On the other hand, in Beijing, the most accepted water-saving method (54.3%) was the installation of water-saving tap (Zhang, 2005). A study revealed that an increase of 1 unit in the percentage of families with pools could result to a 1% increase in the mean consumption of water in Phoenix, Arizona, census tract (Wentz & Gober, 2007). Other natural processes in the form of soil types, vegetation types, prevailing rates of evapotranspiration and the such is likely to interact with human preferences in a way that will influence water demand to maintain healthy vegetation. C. Water demand modelling Greater numbers of people now reside in urban areas, more than any other time in history. It is therefore becoming a challenge to predict and manage water demand in the urban areas due to the close relationships between human and natural systems within the cities. These relationships come as a result of multiple interactions between macroscale processes and patterns (regional or municipal) and those at the micro-level (individual or household level). Complex systems, for example, exhibit local interactions among individuals that cumulate over time and space and generate macroscale and mesoscale variables which once again find their way back to affect individual choices. Such circumstances of ecological and social systems in the management of natural resources offers great challenges to water managers in system analysis and accounting for the complex and changing reactions to policies, interventions and shocks. Modelling has continued to play a critical in helping managers complete the analysis and forecast of the urban water demand while they strive to meet the demands of the city residents. Peak water demands forecasts is a major design consideration for the supply system and greatly influences strategies for expansion of existing structures. Making sure that a least cost and up to the task facility expansion strategy means that there must be done an accurate estimation of the required size and operation of pipe system, pumping stations and reservoirs (House-Peters & Chang, 2011). As Bougadis et al (2005) points out, there exists two types of demand forecasting: short-term forecasts and long-term forecasts. Short-term forecasts are generally used for operation and management while the long –term forecasts are required for extensive planning and design of infrastructure. Water managers today use long-time climate trends to come up with demand estimates. They have also widely used the principle of stationarity – which is based on the presumption that systems in nature continue to fluctuate within unchanging envelop of variability. These methods, however, have not remained reliable. Climatic changes have introduced uncertainties that limit the accuracy of such methods since historical trends have not been reliable in the prediction of future climate-sensitive demand of water. Modelling Cities receive their water supplies from different sources like rainfall, underground and surface water. Researchers have developed models from application of different water systems and proposed efficient interventions to help manager predict and manage water supplies efficiently. House-Peter & Chang’s (2011) model first considered urban water demand as a coupled human and natural system with human behaviours and the demand for resources acting as drivers and at the same time constraints of the natural ecosystem function. As Billings & Agthe (1998) put it, within the context of management of the water resource, human and natural mismatches have the potential of imposing significant unanticipated costs to the water utilities if correct demand estimation is not done. Dynamic modeling approaches Dynamic and persistently changing processes are those which generate water demand; the processes basing on multiscale interactions between the natural world and human agents. This has been the major motivating factor for development of dynamic models, and their implementation. Traditional demand functions have been constructed as static, but recent research has revealed that current water use is greatly dependent on past water use. The dynamic models account for this relationship and use it to improve the accuracy and reliability of parameter traditional estimates over traditional methods. As House-Peter & Chang (2011) argues, dynamic models are able to demonstrate how behaviours and decisions regarding water consumption, under plausible future scenarios, are influenced by urban housing and form, price changes, conservation policies and climate change. Two common methods for dynamic modeling used to examine urban water demand are ABMs and System Dynamics Models SDMs. Manson & Evans (2007) agree that ABMs have been widely used in land change science in examination of drivers and effect of land use change. These methods have been widely accepted because they can: incorporate both temporally and spatially explicit data, model bidirectional relations between macro-behaviour of the system and individual human agents, capture patterns that emerge at higher scales due to interactions at lower levels, and can blend quantitative and qualitative approaches (Manson & Evan, 2007). Using ABMs, modelers can allow for positive reinforcement as well as feedback which could be integrated back into the system since changes in behavior of the agent (water user) happen iteratively over time. Again, these models can capture the impact of social networks on behavior of the agent because different groups will not react to conservation messages and policies immediately. As consumption behaviours for a given group continues to change over time, this group, by means of social pressure, exerts its influence on other agents groups of the new water demand. These effects are subsequently re-included in the next iterations of the ABM process. Researchers in Beijing, for example, successfully quantified the dynamic patters of the residential consumption behavior in the city by disaggregating the water usage into specific end uses. These end uses were then explored with in consideration of human behavior, choices and attitudes under scenarios of adverse policy (House-Peter & Chang, 2011). Another dynamic modeling approach, SDM, has been used to dynamically complicated problems in water resource management. Through this approach, modelers have been able to link external systems like climate, to investigate the effect of climate changes on the demand over extended periods of time. These methods improve on traditional statistical models and take into account a broader number of components, behavioural responses, feedback mechanisms and time lags that exists in the system. Through these methods, for example, it was possible to reveal the usefulness of stochastic systems analysis by means of technologic, behavioural and geographic variables that affect water use and demand and its conservation efforts in Amman, Jordan (Rosenberg et al, 2007). Unlike ADMs, however, SDMs cannot be used to simulate the behavior of neighbours or the impact of multiagent behavior on the components of the system over some time. Input - output model for water consumption Velazquez (2006) describes the input-output model which he says is based on Leontief input – output model. In this model, the water consumed is considered as the difference between the water used by a given sector minus the return. Leontief’s basic equation determines that an economy’s production is dependent on intersectoral relations and final demand. When expressed in matrix notation taking into account the economy as a whole, the equation becomes: x = (1 – A)-1 y Solving for x from this equation, the total production delivered to final demand becomes: x = (1-A)-1 y Where the component (1-A)-1is referred to as Leontief inverse matrix that represents the total production that must be generated by each sector so as to satisfy the demand of the economy Going further, the equation below: represents water consumption in a way that the quantity of directly consumed water by sector i (wdi) is dependent on relations between sectors established between the sector i and other sectors (wij) and the amount of water consumed by sector i in meeting its own demand (wydi). Further modification of the equation gives: And in matrix form: Solving this equation, we find: Where the part (1-Q)-1 represents Leontief inverse matrix in terms of water, (^) causes the vector to be placed on the matrix’s diagonal, (`) indicates that the vector has been transposed, and u is the unit column vector. Therefore, the matrix (1-Q)-1 represents the change in consumption of water if the demand changes in one unit. Written in this way, the model is able to account for both the direct and indirect water requirements. That is, the model recognizes the total amount of water consumed by all the sectors so as to satisfy any rise in demand, as opposed to Q which could only reflect the direct requirements of water (Velazquez, 2006). With this model, it is possible to further go beyond the input-output relationship which has been defined for total water consumption. This may be achieved by modifying the model and distinguish between indirect and direct consumption, introduce the concepts into the model and enable the formulation of a matrix of inter-sectoral water relationships in analysis of consumption. Total water consumption, as has been discussed so far is the sum of direct plus indirect consumption. Water quality demand modelling There has been growing interest in water quality in the drink water distribution system (DWDS) due to the customers’ expectation that the water company should ensure highest quality through prevention of obvious deficiencies like discoloration and assurance of sufficient residual chlorine levels. The most important element in any water quality model for any DWDS is accuracy in the hydraulic model which means that there must be detailed knowledge of the water demand. A stochastic demand model for every household connection every minute or any finer basis is required for a water quality network model. Currently, there exists two demand types that could be used to satisfy this requirement: the end-use model SIMDEUM and the Poisson Rectangular Pulse model (Blokker et al, 2008). Researchers have shown that residential water demand could be viewed as rectangular pulses with a particular intensity (flow) and duration which arrive at different times in a day. It has also been found that households use water at a frequency that follows a Poisson arrival process with a time dependent rate parameter. Any two pulses that overlap at any given time are summed up as shown in figure below Fig: summing up of overlapping pulses Extensive measurements are required for the estimation of the parameters to build the Poisson Rectangular Pulse (PRP) model. To estimate the duration and intensity of the arrival, various distributions are applicable for different data sets, like exponential, log-normal and Weibull distributions. Researchers have also used an analogous model that was based on a Neyman-Scott stochastic process (NSRP model) for which measurements are made to find the parameters. PRP parameters are not easy to obtain; several measurements must be done, usually on a per second basis for several homes. Correlating the parameters obtained from these measurements with factors like age, installed water using appliances, size etc is not easy either. As a result, the parameters retrieved for the PRP model cannot be easily transferred to other network. The second type of stochastic demand model is based on statistical information of end use. This method generates demand using SIMDEUM (SIMulation of water Demand, End Use Model which is able to simulate the end use in the form of rectangular pulse from probability distribution functions for frequency, intensity and duration as well as a given probability of use over the day. The population distribution functions are developed from information regarding whether households posses water using appliances, their use and the data about the population. The result of the simulation demand accommodates all the end uses. The model uses flow measurements information for purposes of validity only (Blokker et al, 2008). The end use model does not require much measurement demand validation. It, however, requires statistical information regarding water appliances and their users, information which may be related to cultural differences and therefore, nation specific. Since the model relies on statistical data on residents and in-home installations, it is easy to determine the effect of an aging population on water demand or the acquisition of new appliances; the method therefore, is easily transferable to other networks. As mentioned by Blokker et al (2006), SIMDEUM has been applied effectively with great results and has helped managers predict with good accuracy the demand in the Netherlands. This method as well as the PRP method could be described by the following equations: Where D represents the duration of pulse in seconds, I is the pulse intensity and the time during which the tap remains open. The function B(I, D, ) is a block function with the value I from times to+D, and remains zero at other times of the day. All pulses are summed up. The PRP approach takes a lognormal probability distribution for the intensity and duration, and all pulses have equal parameters. The number of pulses will follow Poisson arrival process and will vary every hour. SIMDEUM, on the other hand, utilizes probability distributions of intensity and duration and the number of pulses will be dependent on the end use type, and the parameters will depend on the age or number of residents in every household (Blokker et al, 2008). Conclusion The paper has discussed in detail the main demand movers of water in the cities, finding a close relation between population growth and modernization or industrialization with the water demand in the cities. Economic activities within and around cities have contributed greatly to the city water demands. Other factors include household norms, practices and ways of life as well as prevailing climate and weather conditions. As has been seen in the paper, the growing city water demand creates the need for future forecasts and efficient planning for this demand to be met. Future estimated demands are overwhelming and there is need for expansion of capacity and establishment of facilities that will supply these demands. Several water demand models already exists and could be adopted to achieve desirable results as has been seen in the Netherlands and all around the world. List of reference Velazquez E, 2006, An input–output model of water consumption: Analysing intersectoral water relationships in Andalusia, Ecological Economics, Vol. 56, pp. 226 – 240 Ayanshola AM, Sule BF & Salami AW, 2010, Modelling of Residential Water Demand at Household Level in Ilorin, Nigeria, Journal of Research Information in Civil Engineering, Vol. 7, No. , pp. 59 - 68 House‐Peters, LA, & Chang H, 2011, Urban water demand modelling: Review of concepts, methods, and organizing principles, Water Resour. Res., 47, W05401 UNDP (United Nations Development Program), 2006, Human Development Report 2006. Beyond Scarcity: Power, poverty and the global water crisis. UNDP, New York Rygaard M, Binning PJ, & Albrechtsen H, 2011, Increasing urban water self-sufficiency: New era, new challenges, Journal of Environmental Management, Vol. 92, pp. 185 – 194 Marella RL, 1992, Factors that Affect Public-Supply Water Use in Florida, with a Section on Projected Water Use to the Year 2020, retrieved on 7th August 2014 from < http://fl.water.usgs.gov/PDF_files/wri91_4123_marella.pdf> Blokker EJM, Vreeburg JHG, Buchberger SG & van Dijk JC, 2008, Importance of demand modelling in network water quality models: a review, Drink. Water Eng. Sci., Vol. 1, pp. 27–38 Zhang HH & Brown DF, 2005, Understanding urban residential water use in Beijing and Tianjin, China, Habitat International, Vol. 29, pp. 469–491 Blokker, EJM, Vreeburg, JHG, & Vogelaar, AJ, 2006, Combining the probabilistic demand model SIMDEUM with a network model, Water Distribution System Analysis #8, Cincinnati, Ohio, USA, 27–30 August. Rosenberg, DE, Tarawneh T, Abdel‐Khaleq R, & Lund JR, 2007, Modeling integrated water user decisions in intermittent supply systems, Water Resour. Res., 43, W07425, doi:10.1029/2006WR005340. Alfarra A, Kemp-Benedict E, Hotzl H, Sader N & Sonneveld B, 2012, Modeling Water Supply and Demand for Effective Water Management Allocation in the Jordan Valley, Journal of Agricultural Science and Applications (JASA), Vol. 1, No. 1, pp. 1-7 Porto M, 2001, Sustaining Urban Water Supplies: A Case Study from São Paulo, Brazil, retrieved on 6th August 2014 from < http://www.siwi.org/documents/Resources/Water_Front_Articles/2000-2001/WF1-00_Sustaining_Urban_Water_Supplies.pdf> Tortajada C, 2006, Water Management in Singapore, Water resources development, Vol. 22, No. 2, pp. 227-240 World Bank. (2006). Dealing with Water Scarcity in Singapore: Institutions, Strategies, and Enforcement. World Bank Analytical and Advisory Assistance (AAA) Program China: Addressing Water Scarcity Background Paper No. 4 Zhang, HH, & Brown DF, 2005, Understanding urban residential water use in Beijing and Tianjin, China, Habitat Int., Vol. 29, No. 3, pp. 469–491, doi:10.1016/j.habitatint.2004.04.002. Billings, RB, & Agthe DE, 1998, State‐space versus multiple regression for forecasting urban water demand, J. Water Resour. Plann. Manage.,124 (2), 113–117, doi:10.1061/(ASCE)0733-9496(1998)124:2(113). McDonald et al, 2014, Water on an urban planet: Urbanization and the reach of urban water infrastructure, Global Environmental Change, Vol. 27, pp. 96–105 Tortajada C & Joshi YK, 2006, Water Demand Management in Singapore: Involving the Public, Water Resources Management, Vol 27, No. 5 Bougadis, J, Adamowski K, & Diduch R, 2005, Short‐term municipal water demand forecasting, Hydrol. Processes, 19, 137–148, doi:10.1002/hyp.5763. United Nations Population Division. (1996). World urbanization prospects: The 1994 revision. New York: United Nations. Manson, SM, and Evans T, 2007, Agent‐based modeling of deforestation in southern Yucatán, Mexico and reforestation in the midwest United States, Proc. Natl. Acad. Sci. U. S. A., 104(52), 20,678–20,683, doi:10.1073/pnas.0705802104. Ada J & Carl-Erik S, 2006, Water demand and the urban poor: a study of the factors influencing water consumption among households in cape town, South Africa, Working Paper Series in Economics and Management No. 02/06, January World Bank. (1998). India-water resources management sector review.: Report on intersectoral water allocation, planning and management, Volume 1: Main report. Washington D.C: World Bank Wentz, EA, & Gober P, 2007, Determinants of small‐area water consumption for the city of Phoenix, Arizona, Water Resour. Manage., 21, pp. 1849–1863, doi:10.1007/s11269-006-9133-0. Read More