Influence of cities population size on their energy consumption and CO2 emissions: the case of Spain

RESEARCH ARTICLE Influence of cities population size on their energy consumption and CO 2 emissions: the case of Spain Irene M. Zarco-Soto 1 &Pedro J. Zarco-Periñán 1 &Rafael Sánchez-Durán 2 Received: 2 June 2020 /Accepted: 18 January 2021 #The Author(s), under exclusive licence to Springer-Verlag GmbH, DE part of Springer Nature 2021 Abstract Half of the world population live in the cities. Cities energy consumption, environmental impact, and the opportunities they provide for our planet’s sustainability make them attractive for governmental authorities. Any action taken in the cities has immediate repercussions. For this reason, many statistical data are published every year. This paper makes the best use of these data to calculate cities CO 2 emissions and their thermal and electric energy consumption. The methodology applied takes into consideration each city size by number of inhabitants and gets results per inhabitant and household. This will makepossible to put into practice the right actions to reduce CO 2 emissions and to use alternative energy. This paper also defines an index to facilitate and simplify the analysis of results. This study was applied to the case of Spain to show the methodology here proposed. In fact, this type of study has never been carried out in Spain before. With this purpose, the 145 Spanish cities with more than 50,000 people were considered. Results show that cities with larger populations present higher consumptions per inhabitant and household. The smallest the population of a city is, the less energy the city consumes. However, electric energy consumption remains constant regardless of the population size. With regard to the CO2 emissions, results bring to light that the biggest cities produce the highest emissions. Furthermore, comparing emissions produced by electrical sources to the total emissions, it was concluded that the smallest cities produce the highest electrical emissions. Keywords Population size .CO 2 emissions .Energy consumption .Cities .Buildings .Spain Introduction In 2007, the world had more than 6600 million people. From then on, more than a half of this population live in the cities. Prospects estimate that this percentage will reach 60% in 2030. Even more people live in specific areas, such as Oceania, North America, or the Caribbean. And in Europe, the number of urban residents reaches 70% at the moment. This percentage will turn into 84% in 2050. In addition, 70% of the world emissions from coal are produced in the cities, where 9 out of 10 urban residents breathe polluted air. Cities consume 60% of resources (Department of Economic and Social Affairs 2015). In Europe, this strong demographic increase of cities entails that almost 80% of the energy is consumed in them (Committed to local sustainable energy 2019). Therefore, buildings are key elements due to their energy use and CO 2 emissions. For this reason, this study focuses on cities. Buildings’energy consumption is electric and thermal (natural gas) (Shahrokni et al. 2014). Energy consumption, environmental impact, and opportunities for our planet’s sustainability are the main elements that make cities so attractive for governmental authorities. Actions taken in the cities bring quick results. The Sustainable Development Goal 11 of the United Nations (Fig. 1)(United Nations Educational 2019) refers to the cities: make cities and human settlements inclusive, safe, resilient, and sustainable. Two of its objectives are the following: by 2020, substantially increase the number of cities and human settlement adopting and implementing integrated policies and plans toward inclusion, resource efficiency, mitigation, and adaptation to climate change and, by 2030, reduce the adverse per capita environmental impact of cities (Department of Economic and Social Responsible Editor: Philippe Garrigues *Pedro J. Zarco-Periñán [email protected] 1 Departamento de Ingeniería Eléctrica, Escuela Técnica Superior de Ingeniería, Universidad de Sevilla, Camino de los Descubrimientos, s/n, 41092 Seville, Spain 2 Endesa, Avenida de la Borbolla 5, 41004 Seville, Spain https://doi.org/10.1007/s11356-021-12624-3 / Published online: 2 February 2021 Environmental Science and Pollution Research (2021) 28:28146–28167 Affairs 2016). And cities are involved in other goals, such as the following: Goal 7, with the use of renewable energies that ensures access to energy for everyone; Goal 12, with a sustainable and responsible consumption; and Goal 13, with a reduction of CO 2 emissions. The European Union launched the Covenant of Mayors in 2008 after the adoption of the 2020 European Union Climate and Energy Package. The Covenant of Mayors changed its name in 2015 into Climate & Energy after the addition of the goals set for 2030. This was agreed by 53 different countries from around the world (not only from Europe). It involved almost 8000 cities and had an impact on more than 250 million people. Covenant signatories commit to adopting an integrated approach to climate change mitigation and adaptation. The objectives are to reduce CO 2 emissions by at least 40% by 2030 and to increase resilience to climate change (Covenant of Mayors for Climate and Energy 2019). It is well-known that cities are one of the main key points for mitigating climate change and reducing greenhouse gases. However, studies carried out to date base on developing formulas to forecast energy consumption; analyzing the factors that have a major (or minor) influence on consumption; and analyzing the factors that affect the most the CO 2 emissions production. The scope of these studies is a whole country or a huge area, although they do not focus on cities as very important cores. Unlike these existing studies, this paper centers on cities as cores that consume energy (in their buildings) and produce CO2 as derivative wastes. Moreover, this paper focuses on each city in particular, since their consumption and emissions can vary depending on the size of their population. Lastly, this methodology will be used for the case of Spain, considering all Spanish cities with populations greater than 50,000 people. To the best of these authors’knowledge, a similar study to this one has never been done before: analyzing cities energy consumption and their buildings emissions considering the size of their population and analyzing each and every city of a country separately and Spain in particular. Only a recent study analyzes how cities energy consumption and CO2 emissions are related to their climate (Zarco-Soto et al. 2020). Therefore, cities are so important. And so are buildings, since they are CO2 producers and energy consumers. For this reason, this paper focuses on analyzing cities and investigating the impact that their size has on them. Note that the word size used in this paper will refer from now on to the population size, or in other words, to the number of inhabitants. The information used in this paper is public. Consequently, it was not necessary to invest economical resources or conduct specialized surveys according to the data here needed. A large amount of information is published in this regard. However, this information needed to be handled so as to get results regarding energy consumption and CO 2 emissions. The methodology here used is similar to the one of those studies that intend to expose the relation between different variables and Fig. 1 Sustainable Development Goals (United Nations Educational 2019) 28147Environ Sci Pollut Res (2021) 28:28146–28167 to get values, rather than developing prediction formulas (Zarco-Soto et al. 2020; Urquizo et al. 2017; Hekkenberg et al. 2009). In the case of this paper, this research addresses the dependence of cities population size on other variables. This study’s findings will allow governments and administrations for taking actions to promote energy saving and efficiency. It will allow companies as well for investing in renewable energy, which helps reducing CO 2 emissions. And consequently, it will help to achieve the sustainable development goals. Nevertheless, it is crucial to identify the starting point to achieve these aims. In summary, this article shows a methodology similar to that applied in other types of studies but adapted to the analysis of the influence that the size of the city has on energy consumption and CO2 emissions. It is applicable to any geographical area and can be as large or as small as desired. The analysis is made for each city based on its size. With the vision provided by the results obtained, conclusions can be drawn on how to act in the future with regard to the energy needs of cities and the mitigation of CO2 emissions. In particular, and as an example, the case of the application of the methodology to all Spanish cities with more than 50,000 inhabitants is shown, obtaining a series of conclusions with the results obtained. And based on the sample used, it can be expected that the results obtained will be valid when considering other geographical areas. To the knowledge of the authors, this type of study, and with this specific scope, has not been carried out previously. This paper is structured as follows: the “Literature review” section presents studies that relate cities size to their energy consumption and CO 2 emissions; the “Methodology”section describes the proposed methodology, which can be applied to any country or region using the available public information; the “Application of this study to the case of Spain”section details the way that the proposed methodology is applied for the case of Spain; the “Results and discussion”section presents findings and proceeds with their discussion; and lastly, the “Conclusions”section explains conclusions of this study. Literature review This study analyzes electric and thermal (natural gas) consumption and CO 2 emissions considering cities sizes. Taking this into account, the following literature review will address these three mentioned contents block. Later, this study is applied to the case of Spain, reason why the literaturerelated to this country is as well reviewed at the end of this section. There is not much literature about this subject. Therefore, literature corresponding to the next level (not only in terms of cities) will also be reviewed. Most existing studies regarding energy consumption base on developing consumption forecasts. They take into consideration different variables that may affect consumption, such as population. They always analyze it in a global way and not in terms of a city. In general terms, Wang et al. (2019) review all studies that assess urban energy performance. It makes an exhaustive study of published works and brings to light the rising interest this topic has in the whole world. The number of published papers went from 9 in 2007 to 82 in 2017. Even so, the mentioned paper points out that the number of studies is small. It includes energy performance studies with different geographical scopes: from a whole country to just one building. It also presents different points of view, such as the welfare that comes along with energy or safe energy supply. Regarding electric energy consumption in national terms, most studies that relate electric energy use to population are about Turkey. Kankal et al. (2011) make consumption forecasts using the following variables: gross domestic product (GDP), population, employment, importation, and exportations. Moreover, its literature review section presents two tables that summarize the studies to date about energy forecasts in terms of country. One of the tables refers just to Turkey and the other one to the rest of countries. These tables present the type of energy forecasted, as well as the methodology and variables used to make calculations. Forecasts using population as a variable were made on the following countries: Taiwan, Italy, South Korea, and New Zealand. From that date on, more studies considering population as a variable were carried out to estimate energy consumption. Continuing with studies about Turkey, Günay (2016)makes use of six variables for its model. Population and GDP per capita are the variables that had a major impact. Aydin (2014) forecasts primary energy in Turkey based on linear regressions. It uses population and GDP. It makes an energy consumption prediction for 2025. Regarding other countries, Bianco et al. (2013)estimatethe electric energy consumption in Italy using GDP, GDP per capita, and historical electricity consumption as variables. It uses a simpler regression analysis and it leads to quite good estimations. For the case of Pakistan, Zaman et al. (2012) analyze the impact of population growth, GDP per capita growth, and foreign investments on electricity use. Population growth is the variable that has a major impact on electricity consumption growth. In terms of no national electricity consumption forecasts, Hor et al. (2005) focus on England and Wales and base on weather variables, GDP, and population growth. They use different models and changes in the weather variables are considered. In terms of city, Wangpattarapong et al. (2008) analyze the factors that affect residential electric consumption in Bangkok Metropolis. They base on climatology, income, number of houses, and shipment of air-conditioner. And Ranjan and Jain (1999) get four equations to model electricity consumption in Delhi for different seasons. 28148 Environ Sci Pollut Res (2021) 28:28146–28167 The same as electricity consumption, studies about thermal energy consumption focus on getting formulas to forecast consumption. These studies address a country. And they have in common GDP in addition to population. Bianco et al. (2014) make consumption estimates in the Italian residential sector based on population, GDP per capita, heating degreedays, and gas price. Wadud et al. (2011) forecast the national energy consumption in Bangladesh based on the price of natural gas, population, and GDP. Consumption estimates in China are made by Li et al. (2011) based on population, GDP, and investment in the natural gas sector. Azadeh et al. (2011) forecast consumptions in Bahrain, Saudi Arabia, Syria, and the United Arab Emirates. They use population and GDP as variables. Behrouznia et al. (2010) make use of the same variables but focus on Argentina, Brazil, Colombia, Venezuela, and Cuba. Sarak and Satman (2003) based on population, degree-days, and natural gas pipelines (the already existing pipelines and the ones planned) to analyze the residential natural gas consumption in Turkey. Zhang (2004) makes a comparison between total energy consumption of the residential sector in China, Japan, Canada, and the USA. It bases on the relation between consumption per household and heating degree-days. The greatest consumption is observed in Japan, followed by the USA, Canada, and China in the last place. In terms of city, Szoplik (2015) reviews natural gas consumption of Szczecin (Poland) considering the following factors: temperature and calendar. With regard to CO 2 emissions, most studies analyze how different factors impact them. Population is included among which. Unlike electricity and natural gas, only a few studies forecast emissions (Pao et al. 2012). As an exception, some studies analyze in detail the emissions produced by a city. Feng and Zhang (2012) is an example of it. This paper centers on Beijing and takes into consideration different situations to forecast energy consumption and CO 2 emissions. Many countries have analyzed the variables that have an impact on the CO 2 emissions. For all of them, the scope was diverse. And it did not only include the emissions related to natural gas. Population is one of the variables they used the most. In all cases, it is observed that population has a huge and important impact on emissions, regardless of the country of study. Wen and Shao (2019) analyze how the following factors impact CO 2 emissions in the Chinese commercial sector: population, urbanization (urban population divided by total population), economic growth, foreign direct investment, energy structure, and energy intensity. Lin and Ahmad (2017) study CO 2 emissions in Pakistan considering population and GDP. Solarin and Lean (2016) evaluate the impact of urbanization, GDP, and natural gas consumption on the CO 2 emissions of India and China. Wang et al. (2016a)examine the countries that are part of the Association of Southeast Asian Nations (ASEAN). Laos and Cambodia are not included though, due to the lack of data for the time period considered. In other words, the countries of study are Singapore, Malaysia, Indonesia, Thailand, the Philippines, Brunei, Vietnam, and Myanmar. This study analyzes the effect of urbanization and energy use on carbon emissions. For the case of Ghana, Asumadu-Sarkodie and Owusu (2016) focus on the relation between population growth, GDP, energy use, and CO 2 emissions. Wang et al. (2016b) evaluate the relation between urbanization and CO 2 emissions for the BRICS countries (Brazil, Russia, India, China, and South Africa). Alam et al. (2016) analyze the impact that population growth, energy consumption, and income have on CO 2 emissions in Brazil, China, India, and Indonesia. Unlike in the previous studies described, in this one the effect of population is only observed in Brazil and India, but not in China and Indonesia. The same happens in Brizga et al. (2013). It analyzes the impact of population, industrialization, carbon intensity, energy intensity, and energy mix on the CO 2 emissions in the former Soviet Union. However, it concludes that population changes barely affect emissions. This is also observed in Sohag et al. (2017). This study centers on analyzing the impact of energy consumption, population and economic growth, trade openness, and sectoral GDP on CO 2 emissions in 83 middle-income countries. Among its findings, population growth does not significantly impact these emissions. Ali et al. (2016) come to the same conclusion for Nigeria when Type of climate of each city Classificaon of cies by populaon size Electricity consumpon of each city buildings Thermal energy consumpon (natural gas) of each city buildings Energy consumpon by populaon size CO 2 emissions by populaon size Climac map Fig. 2 Methodology flowchart 28149Environ Sci Pollut Res (2021) 28:28146–28167 analyzing the effect of urbanization, economic growth, trade openness, and energy consumption on CO 2 emissions. It concludes that urbanization presents no meaningful impact on CO 2 emissions. In terms of regions and not only of countries as in the studies mentioned before, Zhang and Lin (2012)analyze how urbanization affects CO 2 emissions and energy consumption in the different Chinese regions. They also analyze the differences between them. Wang et al. (2017) focus on the impact of seven factors on emissions in 30 Chinese regions. Population is one of these factors. Additionally, in terms of city and referring not only to CO 2 emissions, but also to air pollution in general, Liu et al. (2018) analyze 83 cities in China. They take population size and seasonality into consideration. Results show that the bigger the population size is, the more pollution the city produces. Fig. 3 Most populated cities in Europe Table 1 Classification of Spanish cities by population size Size Cities Group 1: from 50,000 to 125,000 people Alcalá de Guadaíra, Alcobendas, Alcoy/Alcoi, Algeciras, Aranjuez, Arganda del Rey, Arona, Arrecife, Ávila, Avilés, Barakaldo, Benalmádena, Benidorm, Boadilla del Monte, Cáceres, Cádiz, Castelldefels, Cerdanyola del Vallès, Ceuta, Chiclana de la Frontera, Ciudad Real, Collado Villalba, Cornellà de Llobregat, Coslada, Cuenca, Ejido (El), Elda, Estepona, Ferrol, Fuengirola, Gandía, Getxo, Girona, Granollers, Guadalajara, Huesca, Irún, Jaén, Linares, Línea de la Concepción (La), Lorca, Lugo, Majadahonda, Manresa, Melilla, Mérida, Mijas, Molina de Segura, Mollet del Vallès, Motril, Orihuela, Ourense, Palencia, Paterna, Pinto, Ponferrada, Pontevedra, Pozuelo de Alarcón, Prat de Llobregat (El), Puerto de Santa María, Reus, Rivas-Vaciamadrid, Roquetas de Mar, Rozas de Madrid (Las), Rubí, Sagunto/Sagunt, San Bartolomé de Tirajana, San Fernando, San Sebastián de los Reyes, San Vicente del Raspeig, Sanlúcar de Barrameda, Sant Boi de Llobregat, Sant Cugat del Vallès, Santa Coloma de Gramenet, Santa Lucía de Tirajana, Santiago de Compostela, Segovia, Siero, Talavera de la Reina, Telde, Toledo, Torrelavega, Torremolinos, Torrent, Torrevieja, Utrera, Valdemoro, Vélez-Málaga, Viladecans, Vilanova i la Geltrú, Vila-Real, Zamora Group 2: from 125,000 to 250,000 people A Coruña, Albacete, Alcalá de Henares, Alcorcón, Almería, Badajoz, Badalona, Burgos, Cartagena, Castellón de la Plana, Dos Hermanas, Elche/Elx, Fuenlabrada, Getafe, Granada, Huelva, Jerez de la Frontera, Leganés, León, Lleida, Logroño, Marbella, Mataró, Móstoles, Oviedo, Pamplona/Iruña, Parla, Sabadell, Salamanca, San Cristóbal de la Laguna, San Sebastián/Donostia, Santa Cruz de Tenerife, Santander, Tarragona, Terrassa, Torrejón de Ardoz, Vitoria/Gasteiz Group 3: from 250,000 to 500,000 people Alicante/Alacant, Bilbao, Córdoba, Gijón, Las Palmas, L’Hospitalet de Llobregat, Murcia, Palma de Mallorca, Valladolid, Vigo Group 4: from 500,000 to 1,000,000 people Málaga, Seville, Valencia, Zaragoza Group 5: more than 1,000,000 people Barcelona, Madrid 28150 Environ Sci Pollut Res (2021) 28:28146–28167 In Europe, Cárdenas-Rodríguez et al. (2016) evaluate air pollution in 249 cities according to urban structure. This study concludes that the most populated cities present higher pollution. In the USA, Fragkias et al. (2013) analyze 942 metropolitan areas considering the period from 1999 to 2008. The conclusion in this case is the same: the bigger an urban population is, the more CO 2 emissions. Makido et al. (2012)focus on 50 Japanese cities and conclude that the bigger a city is (in terms of population), the more CO 2 emissions per capita. When reviewing studies about electric and thermal energy consumption and CO 2 emissions in Spanish cities, it is observed once again that there are hardly any. Once the existing ones are analyzed, it is noted that they focus on the whole nation. And as a variable they use the country’s population instead of the population of each city in particular. Pérez- García and Moral-Carcedo (2016) forecast demand in Spain. To achieve this aim, it starts taking into consideration different key factors that have an impact on it. One of these factors is residential consumption, which is calculated based on population growth. Blázquez et al. (2013) analyze the factors that affect residential electricity demand considering provinces. Population, household size, degree-days, household income, and homes with gas access are some of these factors. Labandeira et al. (2006) estimate demand for energy from different energy sources. They also analyze the impact that changes in the prices of energetic products have on households depending on their location: rural, urban, or intermediate. Labandeira et al. (2004) make a similar study that includes a simulation of the effects of an energy tax on CO 2 emissions. Regarding natural gas consumption in Spain, there is no specific study to the authors’knowledge, neither in terms of cities nor in terms of population. Studies that relate Spain to CO 2 emissions are neither about cities nor population. Most of these studies do not exclusively focus on Spain, but they include this country as part of a larger geographical scope. Cansino et al. (2015) analyze different factors to establish their influence on CO 2 emissions. Population is one of these factors. It concludes that both economic activity and population growth are responsible for an increase of emissions. Chen et al. (2018) analyze the impact of six different factors on CO 2 emissions in the OECD (Organization for Economic Co-operation and Development). Spain is one of the OECD members. Population size is one of the analyzed factors. It is observed that the impact of population size is lower than the impact of energy intensity and GDP per capita. Dong et al. (2018)addresses 128 countries. Spain is among them. They focus on the connection between CO 2 emissions, economic growth, Fig. 4 Number of cities of each group Fig. 5 Population of each group of cities 28151Environ Sci Pollut Res (2021) 28:28146–28167 Table 2 Statistical data by population and household of the five defined groups of cities Population Number of households Population size Total Mean Std. dev. Median Maximum Minimum Total Mean Std. dev. Median Maximum Minimum Group 1 6,957,110 75,621 17,654 74,785 121,133 50,334 2,588,844 28,140 6739 27,976 45,559 15,434 Group 2 6,636,954 179,377 36,578 178,288 246,976 125,317 2,541,152 68,680 15,707 69,765 104,385 42,006 Group 3 3,350,814 335,081 59,460 327,952 443,243 257,349 1,283,517 128,352 17,663 127,111 154,421 97,044 Group 4 2,711,182 677,796 89,888 677,186 787,808 569,002 1,052,531 263,133 42,245 265,415 312,339 209,362 Group 5 4,803,790 2,401,895 1,104,622 2,401,895 3,182,981 1,620,809 1,928,425 964,213 421,534 964,213 1,262,282 666,143 Table 3 Statistical data of cities consumption by group of cities Total (MWh/year) Thermal (MWh/year) Population size Total Mean Std. dev. Median Maximum Minimum Total Mean Std. dev. Group 1 27,404,178 297,871 107,709 286,756 664,016 140,786 8,549,523 92,930 80,299 Group 2 29,729,294 803,494 284,437 740,029 1,560,678 354,096 11,512,947 311,161 226,943 Group 3 12,952,786 1,295,279 312,517 1,201,674 1,822,658 898,790 3,725,669 372,567 332,690 Group 4 10,065,882 2,516,471 939,424 2,323,573 3,717,939 1,700,796 2,652,327 663,082 678,304 Group 5 26,156,830 13,078,415 7,526,515 13,078,415 18,400,465 7,756,365 12,417,911 6,208,955 3,904,657 Thermal (MWh/year) Electric (MWh/year) Population size Median Maximum Minimum Total Mean Std. dev. Median Maximum Minimum Group 1 69,761 323,698 0 18,854,655 204,942 50,307 197,277 340,318 120,861 Group 2 308,306 928,818 0 18,216,347 492,334 107,846 498,166 714,245 309,446 Group 3 252,449 1,089,772 1 9,227,117 922,712 268,655 840,360 1,368,528 635,937 Group 4 437,697 1,627,614 149,319 7,413,555 1,853,389 317,360 1,849,919 2,162,239 1,551,479 Group 5 6,208,955 8,969,965 3,447,946 13,738,919 6,869,460 3,621,858 6,869,460 9,430,500 4,308,420 28152 Environ Sci Pollut Res (2021) 28:28146–28167 population growth, and renewable energy growth. They conclude that population growth has a meaningful impact on emissions. Spain is also included in Sharma (2011). This study analyzes 69 countries to check the factors that impact their CO 2 emissions: urbanization, per capita GDP, trade openness, and energy consumption. Among other conclusions, urbanization and GDP per capita are two of the main factors responsible for an increase of emissions. Martínez- Zarzoso et al. (2007) establish how population growth affects CO 2 emissions of 23 European countries. It distinguishes between the 15 countries that joined the European Union before 2004 (Spain was one of them) and the other 8 that joined in 2004. Results show that population growth has a major impact on the new countries rather than on the old ones. As observed, there is too little literature on the subject energy consumption and emissions considering cities population size, not to say hardly any. The existing literature focuses on forecasting national energy consumption or on analyzing the factors that mostly affect CO 2 emissions. Methodology The methodology here proposed calculates cities energy consumption using the official statistical data published by governments (population, households, thermal and electric energy consumption). Consumptions are calculated per inhabitant and household to analyze their performance taking into consideration the city they belong to. Using these results, CO 2 emissions can be also calculated per inhabitant and household. This way, it is possible to analyze cities performance depending on their population size. Moreover, this information is very useful to estimate each city need for renewable energy so as to mitigate their emissions. The same analysis can be carried out eliminating the impact of climate. This makes possible to analyze the effect of population size alone on energy consumption, eliminating the impact that weather conditions may have on both consumption and emissions. In order to clarify this concept, Fig. 2presents a flowchart with the methodology to use. This methodology is similar to the one proposed by Zarco-Soto et al. (2020). Classification of cities by population size The first step is to select the cities case study to evaluate their energy consumption and emissions. Cities are ordered by size and they are grouped by number of inhabitants. Depending on the country or zone considered, cities with a specific number of inhabitants are taken into consideration. Cities are grouped based on the services they have, so that cities expected to present similar performances belong to the same group. Depending on the size, both commercial and administration services will be proportional to the number of inhabitants. The main statistical data of electric, thermal, and total energy consumptions are presented for each size group: Number of cities: ni¼∑ j 1ð1Þ Mean: Ei¼∑ j Eij ð2Þ Standard deviation: si¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ∑ ij Eij−Ei  2 ni−1ðÞ v u u u tð3Þ Median: Mediani¼niþ1 2  th term if the total number of the elements is an odd number, otherwise Mediani¼ ni 2  th term þni 2þ1  th term 2ð4Þ Maximum: Eimax¼max Eij  ð5Þ Fig. 6 Thermal, electric, and total mean energy consumption by group of cities 28153Environ Sci Pollut Res (2021) 28:28146–28167 Minimum: Eimin¼min Eij  ð6Þ where n i is the number of cities that belong to group i;Eiis themeanenergyconsumedingroupi;E ij is the energy consumption of city j, which is located in group i;s i is the standard deviation of the energy consumed in the cities of group i;the energy consumed will be thermal, electric, or total depending on the case study; and cities consumptions should be listed in ascending order to calculate the median. An index was defined to analyze the energy consumption variations of each group. This index is similar to the one defined in Valor et al. (2001) for monthly electric energy consumption and in Zarco-Soto et al. (2020) for energy consumption by climate. The aim of this index is to evaluate variations of consumptions of each group. The size variation index (SVI) is defined as follows: SVIi¼Ei=Eð7Þ where SVI i is the index of the group of cities that have size i,Eiis the energy consumption mean value of group i,andEis the mean energy consumption of all cities (of all groups). This index makes possible to analyze performance of each cities group. Thermal and electric energy consumption To characterize this study, each city was considered as a core where people carry out their work and daily tasks every day. In addition to homes, cities have different services that are also necessary for people’s lives, such as stores and administrative offices. There are more of these services or less depending on the city size. Energy uses related to cities industrial activity were not considered for this study. Otherwise, cities energy consumption could be distorted by their level of industrialization. Consumption of means of transport was neither considered. Consequently, the energy consumption taken into consideration for each city was the electricity and the thermal energy consumed by homes, stores, and administrative offices. These buildings and services are essential for the urban life. This energy is therefore the one consumed by citizens. If a city is bigger than another, it will have more stores and administrative offices than a smaller one. For this reason, this paper considered that the energy consumed by every household is the one properly consumed at home plus a proportional part of the energy consumed by stores and administrative offices. Although this last energy is indispensable to live in the city, citizens do not directly consume it for their personal use. The available statistical information published by official organizations was used to calculate the thermal and electric energy consumptions per inhabitant and household of each Table 4 Statistical data of household consumptions by group of cities Total (MWh/year) Thermal (MWh/year) Electric (MWh/year) Population size Mean Std. dev. Median Maximum Minimum Mean Std. dev. Median Maximum Minimum Mean Std. dev. Median Maximum Minimum Group 1 10.65 3.53 9.86 18.79 5.66 3.39 2.89 2.20 9.16 0.00 7.26 1.09 7.09 9.64 5.37 Group 2 11.65 3.50 11.82 17.28 6.04 4.49 3.03 5.07 8.80 0.00 7.16 1.00 6.96 8.93 5.42 Group 3 10.02 2.22 9.74 14.47 7.79 2.99 2.72 1.93 8.65 0.00 7.03 1.30 6.96 9.26 5.30 Group 4 9.40 2.97 8.45 13.70 7.00 2.40 2.47 1.46 6.00 0.71 7.00 0.69 7.08 7.70 6.12 Group 5 12.87 2.16 12.87 14.39 11.34 6.03 1.40 6.03 7.02 5.04 6.84 0.76 6.84 7.38 6.30 28154 Environ Sci Pollut Res (2021) 28:28146–28167 electricity use, it presents once again an approximately constant value, which in this case is 2.7 MWh per year for all cities. Therefore, it can be concluded that cities total and thermal energy consumptions per household and per inhabitant rise as the cities size increases, except for the case of the four cities with populations between 500,000 and 1,000,000 people, and the electric energy consumption is approximately the same for all cities regardless of their populations. CO 2 emissions Spanish government set CO 2 emissions depending on the energy sources: 0.252 tCO 2 /MWh for generators based on natural gas and 0.291 tCO 2 /MWh for all types of electricity generators and sources (Ministerio de Industria, Energía y Turismo, and Ministerio de Fomento 2016). For this reason, once energy consumptions are identified depending on the energy source, CO 2 emissions can be calculated for each case. Keeping a similar line of reasoning as in the previous sections, the emissions produced are analyzed and then studied once the influence of climate is eliminated. This makes possible to evaluate performance of each group of cities. Figure 17 shows CO 2 emissions per household. It is easily observed that the biggest cities present higher emissions, followed by cities of groups 2 and 1. Cities of group 4 have the shortest emissions. Groups 5 and 2 are the only ones with emissions over the mean (16 and 6% over the mean, respectively). On the contrary, emissions of the group that produces the shortest emissions are 12% under the mean. If thermal emissions are now analyzed, all groups present similar performances. These performances are more pronounced though. Group 5 emissions are the highest ones. They are 65% over the mean. On the contrary, group 4 emissions are the shortest ones. They are 30% lower than the mean. Evaluating emissions produced by electrical sources, cities of group 5 (the 0,00 0,20 0,40 0,60 0,80 1,00 1,20 1,40 1,60 1,80 GROUP 1 GROUP 2 GROUP 3 GROUP 4 GROUP 5 SVI Populaon size Total Thermal Electric Fig. 13 Variation of the SVI index for energy consumptions per household by group of cities without the influence of climate Table 7 Statistical data of energy consumptions per inhabitant by group of cities without the influence of climate Total (MWh/year) Thermal (MWh/year) Electric (MWh/year) Population size Mean Std. dev. Median Maximum Minimum Mean Std. dev. Median Maximum Minimum Mean Std. dev. Median Maximum Minimum Group 1 4.03 0.98 4.21 5.26 1.84 1.30 0.86 1.33 2.69 0.00 2.73 0.32 2.78 3.72 1.72 Group 2 4.26 0.95 4.83 5.51 2.27 1.53 0.81 1.74 2.69 0.22 2.73 0.30 2.74 3.68 2.25 Group 3 4.29 0.85 4.37 5.26 2.47 1.55 0.78 1.69 2.69 0.22 2.74 0.31 2.69 3.47 2.36 Group 4 3.55 1.05 3.60 4.76 2.27 0.82 0.68 0.68 1.69 0.23 2.73 0.33 2.82 3.03 2.25 Group 5 5.12 0.24 5.09 5.26 4.92 2.32 0.53 2.32 2.69 1.94 2.80 0.08 2.80 2.86 2.74 28161Environ Sci Pollut Res (2021) 28:28146–28167 biggest cities) present the shortest emissions. However, the cities that produce higher emissions are the ones that belong to groups 1 and 2. Nevertheless, variations among emissions of all groups are minimal: 2,10 tCO 2 per year. If now each of the groups is analyzed individually, it is observed that emissions of electrical sources are the highest ones. And they vary from 77% (group 4) and 57% (group 1). Figure 18 shows CO 2 emissions per inhabitant. These emissions present performances similar to the ones described for households. Group 5, followed by groups 2 and 1, produces the highest emissions. Group 4 presents the shortest emissions. Emissions vary from 25% over the mean (group 5) and 10% under the mean (group 4). With regard to emissions produced by thermal energy sources, their performance follows the same pattern: they oscillate from 77% over the mean in the case of the biggest cities to 30% under the mean in the case of cities of group 4. Regarding electrical sources, variations are minimal among all groups. Therefore, they can be considered as constant per inhabitant regardless of the group. It is now observed the same as for the emissions per household: emissions produced by electrical sources are the highest ones. Group 4 presents the highest emissions (77% over the mean) and group 5 the shortest ones (57%). Once the influence of climate per household is eliminated, CO 2 emissions increase progressively as the city size grows (Fig. 19). The only exception in this case are the four cities with populations between 500,000 and 1,000,000 people (group 4). Cities of group 5 produce 11% more than the mean. Cities of groups 1 and 4 are the only ones whose CO 2 emissions are under the mean. In particular, emissions of group 4 are 16% lower than the mean. This group presents the shortest emissions. In thermal terms, performance is very similar: emissions increase as the cities size grows, except for the case of group 4. Emissions of this group are 45% lower than the mean, whereas group 1 presents emissions 50% over the mean. With regard to the electrical analysis, the pattern is just the opposite: the smallest cities produce higher CO 2 emissions, whereas the shortest emissions are observed in the biggest cities. However, variations between both extremes are only 6%. Regarding the emissions percentage of one or Fig. 14 Thermal, electric, and total energy consumption per inhabitant by climate zone (Zarco-Soto 2020) Fig. 15 Thermal, electric, and total mean energy consumptions per inhabitant by group of cities without the influence of climate 28162 Environ Sci Pollut Res (2021) 28:28146–28167 another type, electrical emissions take precedence. They vary between 80% (case of group 4) and 58% (case of group 5). Analyzing CO 2 emissions per inhabitant once the influence of climate is eliminated, it is observed that the bigger the cities sizes are, the higher the emissions. This is not so for group 4 (Fig. 20). The two cities of group 5 produce 20% more emissions than the mean and group 4, 13% less. It happens the same with emissions from thermal sources, although they are more noticeable: the biggest cities produce 65% over the mean, whereas cities of group 4 produce 40% less. Regarding emissions from electrical sources, variations between their most extreme emissions are barely 2%. With regard to percentages over the total, emissions from electrical sources take precedence. They oscillate between 58% (in the case of the biggest cities) and 79% (in the case of cities of group 4). Therefore, when the influence of climate is eliminated, it is observed that emissions are higher as the city size grows, especially emissions from thermal sources. This does not happen in the case of the four cities with populations between 500,000 and 1,000,000 people. Emissions from electrical sources remain barely constant per inhabitant, although they slightly increase when the city size enlarges. However, it is just the opposite per household. Despite minimal variations, the bigger the city is, the shorter the emissions from electrical sources. If the influence of climate is not eliminated, performance is very similar to the exposed one. In this case, the 10 cities of group 3 (populations between 250,000 and 500,000 people) are also the exception. According to pollution sources, emissions from electrical sources are the most important ones. They oscillate between 55 and 80% over the total percentage approximately. In each group of cities, the proportion of electrical emissions increases as the cities size decreases. Group 4 is the exception in this case, since electrical emissions of this group mean the highest percentage. Conclusions Cities are one of the places where a great impact is observed when taking actions to help fulfilling the sustainable energy aims of the United Nations and the environmental guidelines of each country. This is due to the fact that more than half of the world population lives in the cities. It even reaches 70% in the case of Europe. In particular, over a half of Spain’spopulation lives in the 145 Spanish cities of more than 50,000 people. Therefore, any measure taken oncities is so important. The scope of this paper was to present a methodology to calculate and analyze cities energy consumptions (both thermal and electric) and CO 2 emissions depending on population and only making use of public existing data. This methodology is applicable to any geographical area. The analysis of the results obtained can help government organizations and companies to carry out adequate planning of the infrastructures necessary for the growth of energy consumption in cities and the reduction of their emissions. 0,00 0,20 0,40 0,60 0,80 1,00 1,20 1,40 1,60 1,80 GROUP 1GROUP 2GROUP 3GROUP 4GROUP 5 SVI Populaon size Total Thermal Electric Fig. 16 Variation of the SVI index for energy consumptions per inhabitant by group of cities without the influence of climate Fig. 17 Thermal, electric, and total CO 2 emissions per household by group of cities 28163Environ Sci Pollut Res (2021) 28:28146–28167 An application of this methodology was explained for the 145 cities of Spain that have more than 50,000 people. For that purpose, these cities were identified and classified in five different groups based on their number of inhabitants. Next, the energy consumption of each group was analyzed per household and per inhabitant. Moreover, this study was completed eliminating the influence that climate has on the energy consumption of cities. With this aim, it was necessary to locate each city on the climate map of Spain. An index was also defined to make this analysis easier. Fig. 18 Thermal, electric, and total CO 2 emissions per inhabitant by group of cities Fig. 19 Thermal, electric, and total CO 2 emissions per household by group of cities without the influence of climate Fig. 20 Thermal, electric, and total CO 2 emissions per inhabitant by group of cities without the influence of climate 28164 Environ Sci Pollut Res (2021) 28:28146–28167 After the mentioned analysis, a series of findings have been found. The bigger a city is, the higher total and thermal energy it consumes. However, it is not so for the four cities that have populations between 500,000 and 1,000,000 people, since their energy consumptions are the shortest ones. Furthermore, electric energy consumption remains practically constant regardless of the city population size. Therefore, this consumption is more inelastic. This is due to all those electrical and electronic devices in addition to electrical appliances that always are in every home. They are responsible for a base electricity consumption. Finally, CO 2 emissions were calculated: total and emissions from thermal or electrical sources. This makes possible to estimate the need for renewable energy to eliminate or mitigate pollutant emissions. The bigger a city is, the more emissions it produces (especially emissions from the thermal sources). It was also observed in this case that the four Spanish cities with populations between 500,000 and 1,000,000 people are an exception. With regard to emissions from electrical sources, it was concluded that they remain barely constant. Finally, emissions from electrical sources are responsible for approximately between 55 and 80% of the total pollution in each group. These percentages enlarge as the city size decreases. However, the cities with populations from 500,000 to 1,000,000 people are the exception. For these cities, it was concluded that emissions from electrical sources take precedence. Taking into account the size of the study sample in Spain, the results obtained suggest that the behavior of cities, each considered as an entity that consumes energy and produces emissions, will be similar in other geographical areas. Acknowledgements This research was funded by the CERVERA research program of the CDTI, the Industrial and Technological Development Centre of Spain, under the research Project HySGrid+ (CER-20191019). Availability of data and materials The data that support the findings of this study are available from indicated references. Author contribution Irene M. Zarco-Soto and Pedro J. Zarco-Periñán wrote the manuscript, established the methodology, and conducted the research and analysis. Irene M. Zarco-Soto, Pedro J. Zarco-Periñán, and Rafael Sánchez-Durán contributed to the study conception and design. All authors have contributed to this manuscript, reviewed, and approved the current form of the manuscript to be submitted. Declarations Ethics approval Not applicable. Consent to participate Not applicable. Consent for publication Not applicable. Conflict of interest The authors declare no competing interests. References Alam MM, Murad MW, Noman AHM, Ozturk I (2016) Relationships among carbon emissions, economic growth, energy consumption and population growth: testing environmental Kuznets curve hypothesis for Brazil, China, India and Indonesia. Ecol Indic 70: 466–479. https://doi.org/10.1016/j.ecolind.2016.06.043 Ali HS, Law SH, Zannah TI (2016) Dynamic impact of urbanization, economic growth, energy consumption, and trade openness on CO 2 emissions in Nigeria. Environ Sci Pollut Res 23:12435– 12443. https://doi.org/10.1007/s11356-016-6437-3 Asumadu-Sarkodie S, Owusu PA (2016) CO 2 dioxide emissions, GDP, energy use, and population growth: a multivariate and causality analysis for Ghana, 1971-2013. Environ Sci Pollut Res 23:13508– 13520. https://doi.org/10.1007/s11356-016-6511-x Aydin G (2014) Modeling of energy consumption based on economic and demographic factors: the case of Turkey with projections. Renew Sust Energ Rev 35:382–389. https://doi.org/10.1016/j.rser. 2014.04.004 Azadeh A,Asadzadeh SM,Saberi M, Nadimi V, Tajvidi A,Sheikalishahi M (2011) A Neuro-fuzzy-stochastic frontier analysis approach for long-term natural gas consumption forecasting and behaviour analysis: the cases of Bahrain, Saudi Arabia, Syria, and UAE. Appl Energy 88:3850–3859. https://doi.org/10.1016/j.apenergy.2011.04. 027 Behrouznia A, Saberi M, Azadeh A, Asadzadeh SM, Pazhoheshfar P (2010) An adaptive network based fuzzy inference system-fuzzy data envelopment analysis for gas consumption forecasting and analysis: the case of South America. International Conference on Intelligent and Advanced Systems 1-6. https://doi.org/10.1109/ ICIAS.2010.5716160 Bianco V, Manca O, Nardini S (2013) Linear regression models to forecast electricity consumption in Italy. Energy Sources Part B 8:86– 93. https://doi.org/10.1080/15567240903289549 Bianco V, Scarpa F, Tagliafico LA (2014) Analysis and future outlook of natural gas consumption in the Italian residential sector. Energy Conv Manag 87:754–764. https://doi.org/10.1016/j.enconman. 2014.07.081 Blázquez L, Boogen N, Filippini M (2013) Residential electricity demand in Spain: new empirical evidence using aggregate data. Energy Econ 36:648–657. https://doi.org/10.1016/j.eneco.2012.11.010 Brizga J, Feng K, Hubacek K (2013) Drivers of CO 2 emissions in the former Soviet Union: a country level IPAT analysis from 1990 to 2010. Energy 59:743–753. https://doi.org/10.1016/j.energy.2013. 07.045 Cansino JM, Sánchez-Braza A, Rodríguez-Arévalo ML (2015) Driving forces of Spain’sCO 2 emissions: a LMDI decomposition approach. Renew Sust Energ Rev 48:749–759. https://doi.org/10.1016/j.rser. 2015.04.011 Cárdenas-Rodríguez M, Dupont-Courtade L, Oueslati W (2016) Air pollution and urban structure linkages: evidence from European cities. Renew Sust Energ Rev 53:1–9. https://doi.org/10.1016/j.rser.2015. 07.190 Chen J, Wang P, Cui L, Huang S, Song M (2018) Decomposition and decoupling analysis of CO 2 emissions in OECD. Appl Energy 231: 937–950. https://doi.org/10.1016/j.apenergy.2018.09.179 Comisión Nacional de los Mercados y la Competencia (2017) Informe de supervisión del mercado de gas natural en España. Madrid: Ministerio de Economía, Industria y Competitividad Committed to local sustainable energy, reducing energy dependence in European cities, Covenant of Mayors, https://www. covenantofmayors.eu/IMG/pdf/CoM_Reducing_Energy_ Dependence_for_web.pdf.Accessed13June2019 Covenant of Mayors for Climate & Energy, https://www. covenantofmayors.eu/en/ Accessed 13 May 2019 28165Environ Sci Pollut Res (2021) 28:28146–28167 Department of Economic and Social Affairs (2015) 2014 Demographic yearbook. (Sixty-fourth ed.). New York: United Nations Department of Economic and Social Affairs (2016) Report of the Inter- Agency and Expert Group on Sustainable - Development Goals Indicators (E/CN.3/2016/2/Rev.1), United Nations, New York Dong K, Hochman G, Zhang Y, Sun R, Li H, Liao H (2018) CO 2 emissions, economic and population growth, and renewable energy: empirical evidence across regions. Procedia Energy Econ 75:180–192. https://doi.org/10.1016/j.eneco.2018.08.017 Eurostat, European Commission https://appsso.eurostat.ec.europa.eu/nui/ show.do?dataset=urb_cpop1&lang=en Accessed 8 Dec 2018 Eurostat Methodologies and Working Papers (2008) Statistical classification of economic activities in the European Community. Luxembourg Feng YY, Zhang LX (2012) Scenario analysis of urban energy saving and carbon abatement policies: a case study of Beijing City, China. Procedia Environ Sci 13:632–644. https://doi.org/10.1016/j. proenv.2012.01.055 Fragkias M, Lobo J, Strumsky D, Seto KC (2013) Does size matter? Scaling of CO 2 emissions and US urban areas. PLoS One 8(6): e64727. https://doi.org/10.1371/journal.pone.0064727 Günay ME (2016) Forecasting annual gross electricity demand by artificial neural networks using predicted values of socio-economic indicators and climatic conditions: case of Turkey. Energy Policy 90: 92–101. https://doi.org/10.1016/j.enpol.2015.12.019 Hekkenberg M, Benders RMJ, Moll HC, Schoot AJM (2009) Indications for a changing electricity demand pattern: the temperature dependence of electricity demand in the Netherlands. Energy Policy 37: 1542–1551. https://doi.org/10.1016/j.enpol.2008.12.030 Hor C, Watson SJ, Majithia S (2005) Analyzing the impact of weather variables on monthly electricity demand. IEEE Trans Power Syst 20(4):2078–2085. https://doi.org/10.1109/TPWRS.2005.857397 Instituto Nacional de Estadística, Cifras de población, Ministerio de Asuntos Económicos y Transformación Digital (2018a) http:// www.ine.es/dyngs/INEbase/es/operacion.htm?c=Estadistica_ C&cid=1254736176951&menu=ultiDatos&idp=1254735572981. Accessed 11 Oct 2018 Instituto Nacional de Estadística, Demografía y población, Ministerio de Asuntos Económicos y Transformación Digital (2018b) http://www. ine.es/ss/Satellite?L=es_ES&c=Page&cid=1254735910183&p= 1254735910183&pagename=INE%2FINELayout. Accessed 10 Oct 2018 Kankal M, Akpinar A, Kömürkü MI, Özsahin TS (2011) Modeling and forecasting of Turkey’s energy consumption using socio-economic and demographic variables. Appl Energy 88:1927–1939. https://doi. org/10.1016/j.apenergy.2010.12.005 Labandeira X, Labeaga JM, Rodríguez M (2004) Microsimulating the effects of household energy price changes in Spain. Nota di Lavoro 161. https://doi.org/10.2139/ssrn.644724 Labandeira X, Labeaga JM, Rodríguez M (2006) A residential energy demand system for Spain. Energy J 27(2):87–111. https://doi.org/ 10.5547/ISSN0195-6574-EJ-Vol27-No2-6 Li J, Dong X, Shangguan J, Hook M (2011) Forecasting the growth of China’s natural gas consumption. Energy 36:1380–1385. https:// doi.org/10.1016/j.energy.2011.01.003 Lin B, Ahmad I (2017) Analysis of energy related carbon dioxide emission and reduction potential in Pakistan. J Clean Prod 143:278–287. https://doi.org/10.1007/jclepro.2016.12.113 Liu Y, Wu J, Yu D, Ma Q (2018) The relationship between urban form and air pollution depends on seasonality and city size. Environ Sci Pollut Res 25:1554–15567. https://doi.org/10.1007/s11356-018- 1743-6 Makido Y, Shobhakar D, Yamagata Y (2012) Relationship between urban form and CO 2 emissions: evidence from fifty Japanese cities. Urban Clim 2:55–67. https://doi.org/10.1016/j.uclim.2012.10.006 Martínez-Zarzoso I, Bengochea-Morancho A, Morales-Lage R (2007) The impact of population on CO 2 emissions: evidence from European countries. Environ Resour Econ 38:497–512. https://doi. org/10.1007/s10640-007-9096-5 Ministerio de Industria, Energía y Turismo, Ministerio de Fomento (2016) Factores de emisión de CO 2 y coeficientes de paso a energía primaria de diferentes fuentes de energía final consumidas en el sector de edificios de España. https://energia.gob.es/desarrollo/ EficienciaEnergetica/RITE/Reconocidos/Paginas/ IndexDocumentosReconocidos.aspx Accessed 21 Oct 2018 Pao H, Fu H, Tseng C (2012) Forecasting of CO 2 emissions, energy consumption and economic growth in China using an improved grey model. Energy 40:400–409. https://doi.org/10.1016/j.energy. 2012.01.037 Pérez-García J, Moral-Carcedo J (2016) Analysis and long term forecasting of electricity demand through a decomposition model: a case study for Spain. Energy 97:127–143. https://doi.org/10.1016/j. energy.2015.11.055 Ranjan M, Jain VK (1999) Modelling of electrical energy consumption in Delhi. Energy 24:351–361. https://doi.org/10.1016/s0360-5442(98) 00087-5 Sarak H, Satman A (2003) The degree-day method to estimate the residential heating natural gas consumption in Turkey: a case study. Energy 28:929–939. https://doi.org/10.1016/s0360-5442(03) 00035-5 Secretaría de Estado de la Energía (2018) Estadísticas y balances energéticos. Madrid: Ministerio para la Transición Ecológica Shahrokni H, Levihn F, Brandt N (2014) Big meter data analysis of the energy efficiency potential in Stockholm’s building stock. Energy Build 78:153–164. https://doi.org/10.1016/j.enbuild.2014.04.017 Sharma AA (2011) Determinants of carbon dioxide emissions: empirical evidence from 69 countries. Appl Energy 88:376–382. https://doi. org/10.1016/j.apenergy.2010.03.022 Sohag K, Mamum MA, Uddin GS, Ahmed AM (2017) Sectoral output, energy use, and CO 2 emission in middle-income countries. Environ Sci Pollut Res 24:9754–9764. https://doi.org/10.1007/s11356-017- 8599-z Solarin SA, Lean HH (2016) Natural gas consumption, income, urbanization, and CO 2 emissions in China and India. Environ Sci Pollut Res 23:18753–18765. https://doi.org/10.1007/s11356-016-7063-9 Szoplik J (2015) Forecasting of natural gas consumption with artificial neural networks. Energy 85:208–220. https://doi.org/10.1016/j. energy.2015.03.084 United Nations Educational, Scientific and Cultural Organization (UNESCO), https://en.unesco.org/sdgs.Accessed29June2019 Urquizo J, Calderón C, James P (2017) Metrics of urban morphology and their impact on energy consumption: a case study in the United Kingdom. Energy Res Soc Sci 32:193–206. https://doi.org/10. 1016/j.erss.2017.03.011 Valor E, Meneu V, Caselles V (2001) Daily air temperature and electricity load in Spain. Appl Meteorol 40:1413–1421. https://doi.org/10. 1175/1520-0450(2001)040<1413:DATAEL>2.0.CO;2 Wadud Z, Dey HS, Kabir MA, Khan SI (2011) Modeling and forecasting natural gas demand in Bangladesh. Energy Policy 39:7372–7380. https://doi.org/10.1016/j.enpol.2011.08.066 Wang Y, Chen L, Kubota J (2016a) The relationship between urbanization, energy use and carbon emissions: evidence from a panel of Association of Southeast Asian Nations (ASEAN) countries. J Clean Prod 112:1368–1374. https://doi.org/10.1007/jclepro.2015. 06.041 Wang Y, Li L, Kubota J, Han R, Zhu X, Lu G (2016b) Does urbanization lead to more carbon emission? Evidence from a panel of BRICS countries. Appl Energy 168:375–380. https://doi.org/10.1016/j. apenergy.2016.01.105 Wang Y, Kang Y, Wang J, Xu L (2017) Panel estimation for the impacts of population-related factors on CO 2 emissions: a regional analysis 28166 Environ Sci Pollut Res (2021) 28:28146–28167 in China. Ecol Indic 78:322–330. https://doi.org/10.1016/j.ecolind. 2017.03.032 Wang L, Long R, Chen H, Li W, Yang J (2019) A review of studies on urban energy performance evaluation. Environ Sci Pollut Res 26: 3243–3261. https://doi.org/10.1007/s11356-018-3915-9 Wangpattarapong K, Maneewan S, Ketjoy N, Rakwichian W (2008) The impacts of climatic and economic factors on residential electricity consumption of Bangkok Metropolis. Energy Build 40:1419–1425. https://doi.org/10.1016/j.enbuil.2008.01.006 Wen L, Shao H (2019) Analysis of influencing factors of the carbon dioxide emissions in China’s commercial department based on the STIRPAT model and ridge regression. Environ Sci Pollut Res 26: 27138–27147. https://doi.org/10.1007/s11356-019-05929-x Zaman K, Khan MM, Ahmad M, Rustam R (2012) Determinants of electricity consumption function in Pakistan: old wine in a new bottle. Energy Policy 50:623–634. https://doi.org/10.1016/j.enpol. 2012.08.003 Zarco-Soto IM, Zarco-Periñán PJ, Sánchez-Durán R (2020) Influence of climate on energy consumption and CO 2 emissions: the case of Spain. Environ Sci Pollut Res. https://doi.org/10.1007/s11356- 020-08079-7 Zhang Q (2004) Residential energy consumption in China and its comparison with Japan,Canada, and USA. EnergyBuild 36:1217–1225. https://doi.org/10.1016/j.enbuild.2003.08.002 Zhang C, Lin Y (2012) Panel estimation for urbanization, energy consumption and CO 2 emissions: a regional analysis in China. Energy Policy 49:488–498. https://doi.org/10.1016/j.enpol.2012.06.048 Publisher’snote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 28167Environ Sci Pollut Res (2021) 28:28146–28167