Management and estimation of thermal comfort, carbon dioxide emission and economic growth by support vector machine

https://doi.org/10.1016/j.rser.2016.06.034Get rights and content

Abstract

Urbanization and climate change are two defining environmental phenomena and these two processes are increasingly interconnected, as rapid urbanization is often accompanied by a change in lifestyle, increasing consumptions and energy uses, which contribute heavily towards climate change and thermal comfort. Success of public urban areas in attraction of residents depends on thermal comfort of the visitors. Thermal comfort of urban open spaces is variable, because it depends on climatic parameters and other influences, which are changeable throughout the year, as well as during the day. Therefore, the prediction of thermal comfort is significant in order to enable planning the time of usage of urban open spaces. This paper presents Support Vector Machine (SVM) to predict thermal comfort of visitors at an open urban area. Results from SVM-FFA were compared with two other soft computing method namely artificial neural network (ANN) and genetic programming (GP). The purpose of this research is also to predict carbon dioxide (CO2) emission based on the urban and rural population growth. Estimating carbon dioxide (CO2) emissions at an urban scale is the first step for adaptation and mitigation of climate change by local governments. The environment that governs the relationships between carbon dioxide (CO2) emissions and gross domestic product (GDP) changes over time due to variations in economic growth, regulatory policy and technology. The relationship between economic growth and carbon dioxide emissions is considered as one of the most important empirical relationships. GDP is also predicted based on CO2 emissions. The reliability of the computational models were accessed based on simulation results and using several statistical indicators.

Introduction

In recent years there is rising interest for linking urban planning, environmental health and the quality of life [1]. Public open spaces are recognized as the opportunity for achieving healthy environment for inhabitants in cities. Pleasant thermal comfort allows adequate conditions for longer duration of stay of users at the urban open space, which can contribute to health, relaxation, socialization and entertainment of children and adults. Staying outdoors can save energy for heating or cooling devices for work. Hence the importance of the thermal comfort of urban open spaces is reflected in the sociological and even ecological and economic sense. The frequency of usage of urban open spaces is strongly affected by the relation between microclimatic conditions of urban open spaces and human comfort, therefore this relation must be taken into consideration during urban design and planning [2]. In order to describe and assess human thermal comfort the concept of thermal index was developed at the beginning of the 20th century by Gagge, Hill and Barnard [3]. The possibility to assess the human thermal condition by one single number which integrates all relevant parameters was first recognized by Fanger in 1972 [4]. Since then more than 40 indices were developed and among most used ones are Predicted Mean Vote (PMV), Physiologically Equivalent Temperature (PET), Standard Effective Temperature (SET) and Perceived Temperature (PT) [5]. PET was introduced by Hoppe and Mayer and represents equivalent to the air temperature at which the heat balance of the human body is maintained with core and skin temperatures equal to those under the conditions being assessed. Numerous research in outdoor thermal comfort assessment which included PET were conducted for different climatic regions and areas like Italy [6], Iran [5], Greece [4], Spain [7] and other.

Thermal comfort of pedestrians is an important factor affecting outdoor activities of visitors at public places such as commercial plazas, parks, streets, etc. Ameliorating outdoors thermal stress encourage inhabitants to use these places more, which in turn benefits the city socially and economically. For instance, the availability of shaded areas outdoor in a hot summer or presence of water features under dry climatic conditions may attract more people and enhance the outdoor activities. In order to modify the outdoor microclimatic conditions with appropriate design strategies [8], designers require a tool that could predict visitor’s comfort level with respect to the changes in various climatic parameters.

Thermal comfort models are numerous; however, they are mostly developed based upon the same principles which might be either the body energy balance model or the adaptive model. This is proven that the energy balance model cannot fully describe the human thermal sensation, confirming the decisive role of thermal adaptation [9]. Indeed, assessment outdoor thermal comfort deals with various environmental, behavioral and psychological parameters [9], [10], [11]. By using these models, a growing number of studies addressed the correlation between the urban canyons geometry and their energy budget. Indeed, the concept of managing the climate dimensions through urban design is distinguished by far [12]. Nevertheless, it is not practically and holistically well-done due to the complexity of the processes.

Previous research works dealing with the visitor's thermal comfort have mostly considered indoor spaces [13]. Studies conducted for indoor spaces follow simple procedures and have limited variables. For example, at the indoor space the subject wears same cloth in all seasons regardless of the outdoor climatic conditions. However, a similar study in outdoor setting involves numerous complex factors and this makes such studies more complicated compared to the indoor studies.

In fact, pedestrians expect variability in the conditions to which they are exposed such as changes of sun and shade, variation in air velocities, and etc. Moreover, people wear diverse cloths for outdoor spaces based on type of climatic, social and cultural zones. Hence, study focusing on the outdoors cannot make simplified assumption by fixing type of cloth worn by subjects. In addition, the influence of a specific microclimatic parameter may differ in different situations e.g. solar radiation in hot summer and cold winter conditions. Air velocity outdoors is normally higher than indoors and may give comfort to visitors to certain extend during summer. Role of water features are considerable particularly in arid regions. On the other hand, due to physiological and psychological impacts, individual differences play an important role in thermal perception. Meanwhile, these effects, particularly gender, are well known on indoor thermal comfort [14], [15], [16], [17], [18], [19]. Nevertheless, these effects have not been thoroughly investigated for some other demographic factors on outdoor thermal comfort. Therefore, these factors need to be considered for evaluating response of subjects at outdoor spaces.

Urban form is increasingly being recognized by scientists for the potential role it might play in the coordination of sustainable urban development and the reduction of carbon dioxide (CO2) emissions. However, despite increasing interest in the morphology of cities in climate change science, few quantitative estimates have been made of the effects of urban form on CO2 emissions. Over half the world’s population lives in cities and the number of people that live in urban areas is constantly increasing [20]. Currently, it is unanimously recognized that urban form can strongly impact on a fast-growing city’s contribution to global climate change through the production of CO2 emissions, and as such, it is clearly necessary to undertake appropriate strategic spatial planning and urban design measures in order to reduce CO2 emissions and thereby address the anticipated impact of global warming [21]. Despite this urgent imperative, existing literature engaging in the task of quantifying the impacts of urban forms on CO2 emissions is limited [22]. Also carbon emissions due to rural energy consumption in have not yet been sufficiently addressed or quantified.

In study [23] was examined the carbon dioxide emissions and its influential factors through the proposed algorithm and the statistical method, providing a theoretical support for further measures to reduce emissions. The results in study [24] was indicated that the total direct CO2 emissions resulting from rural energy consumption have nearly tripled and quantitatively illustrated the importance of rural energy consumption as a contributor to overall carbon emission. The aim of the study [25] was to develop a method to calculate carbon dioxide storage and sequestration at the streetscapes level using field data, an existing tree inventory and available region-specific algometric equations. Rich households generate more emissions per capita than poor households via both their direct energy consumption and their higher expenditure on goods and services that use energy as an intermediate input [26]. An econometric analysis was confirmed a positive relationship between emissions and income [27]. Rapid urbanization increases carbon dioxide emissions both in the short-run and in the long-run [28]. In study [29] was found that nighttime light imagery is appropriate in CO2 estimations at an urban scale. The rapid population growth and resource consumption as a result of urbanization and industrialization during the 20th century have caused global environmental and economic crises [30]. The result in article [31] was revealed that high CO2 emissions are concentrated in dense road network of urban areas with high population density and low CO2 emissions are distributed in rural areas with low population density, sparse road network.

In the last three decades, the effects of CO2 emissions on economic growth have become a topic very significant both at the national and international level. On the other hand, there are a number of studies considering the inseparable relationship between the CO2 emission and economic growth in recent years. In study [32] was confirmed a long-run relationship between CO2 emissions and economic growth. A quantitative structural modeling perspective and policy analysis from an economic integration framework and system estimation on the growth-CO2 emission causality nexus in general and on a major developing country in Asia was presented in [33]. The results in study [34] were shown that there is a nonlinear relationship among CO2 emissions per capita, energy consumption per capita, and gross domestic product (GDP) per capita. Empirical relationship between economic growth, energy consumption and carbon dioxide emissions was examined in [35], was calculated the trend of decoupling effects and finally analyzes the evolution of inequality in CO2 emissions. In article [36] was investigate the causality links between CO2 emissions, foreign direct investment, and economic growth using dynamic simultaneous-equation panel data models and results provided evidence of bidirectional causality between FDI inflows and economic growth for all the panels and between foreign direct investment and CO2 emissions. A network of causal connections among extent of urbanization, CO2 emissions, and economic growth in the short run was sound in Ref. [37]. Dynamic relationship between economic growth and carbon dioxide (CO2) emissions was investigated in Ref. [38] for 181 countries and it was found that for 49 countries (27%), income growth will reduce emissions in the future. Dynamic impacts of GDP growth, energy consumption and population growth on CO2 emissions using econometric approaches for Malaysia was investigated in Ref. [39]. The impact of energy consumption and the CO2 emissions on economic growth using simultaneous-equation models with panel data for 58 countries over the period 1990–2012 was evaluated in Ref. [40] and the empirical results were shown that energy consumption has a positive impact on economic growth. A bidirectional time-varying causality between energy consumption and CO2 emissions was shown in [41], [42]. To evaluate the dynamic behaviors of the energy consumption and CO2 emissions, a few of interdisciplinary studies have been conducted [43], [44], [45].

In this article, we motivate and introduce the prediction model of thermal sensation by using the soft computing approach, namely Support Vector Machines (SVM) coupled with Firefly Algorithm (FFA). The one of the objective is to analyse the CO2 emission forecasting based on the urban and rural population growth in European Union. Finally we analyse the GDP forecasting based on the CO2 emissions from gaseous/liquid/solid fuel consumption.

Nowadays, application of modern computational approach in solving the real problems and determining the optimal values and functions are receiving enormous attention by researchers in different scientific disciplines. Support vector machines (SVMs) have achieved growing usage in different engineering disciplines [46], [47], [48]. The prediction accuracy of an SVM model highly depends on proper selection of model parameters. In addition to organized strategies for selecting parameters, alignment of model parameter also needs focus. In order to select the model parameters researchers have applied different conventional-optimization algorithms, but with limited success [49], [50], [51]. Some of the conventional optimization schemes are the grid search algorithm [52], and gradient decent algorithm [53], [54]. Computational complexity seemed to be the downside of conventional grid search algorithms, which can be applied only to simple cases. Further, grid search technique is prone to local minima. Most optimization problems have multiple local solutions. Therefore, evolutionary algorithms seem to be the best approach, because they are capable of providing global solution to such optimization problems.

Section snippets

Thermal comfort

In this study we used physiological equivalent temperature (PET) which is a thermal index recommended by the German Association of Engineers and officially used by the German Meteorological Servicen. PET index was derived from the human energy balance expressed in °C and it is suitable for urban and regional planners because of the simplicity of interpretation. Additional advantage of PET is that it could be used not only for summer conditions, but through the whole year. The PET index

3.1 Model performance evaluation

To assess the success of the SVM-FFA model and other soft computing technique, the statistical indicators such as root mean square error (RMSE), the coefficient of determination (R2), and the Pearson correlation coefficient (r) were examined by Eqs. (14), (15), (16).RMSE=i=1n(PiOi)2n,r=n(i=1nOiPi)(i=1nOi)(i=1nPi)(ni=1nOi2(i=1nOi)2)(ni=1nPi2(i=1nPi)2)R2=[i=1n(OiOi¯)(PiPi¯)]2i=1n(OiOi¯)i=1n(PiPi¯)where the experimental and forecast values are Pi and Oi respectively, and

Conclusion

The global climate system is being affected by the emissions of greenhouse gases from human development, especially urbanization activities, of which the most significant is carbon dioxide (CO2). Many countries have set long-term goals for CO2 emission reduction. It is widely accepted that human settlements occupy a small proportion of landmass, while playing a significant role in the change of global carbon cycle. Accurate analysis of urban carbon cycle and its interactions with other global

References (89)

  • B. Griefahn et al.

    The effects of gender, age, and fatigue on susceptibility to draft discomfort

    J Thermal Biol

    (2001)
  • S. Karjalainen

    Gender differences in thermal comfort and use of thermostats in everyday thermal environments

    Build Environ

    (2007)
  • M. Indraganti et al.

    Effect of age, gender, economic group and tenure on thermal comfort: a field study in residential buildings in hot and dry climate with seasonal variations

    Energy Build

    (2010)
  • Volker Krey et al.

    Urban and rural energy use and carbon dioxide emissions in Asia

    Energy Econ

    (2012)
  • Yi Chen et al.

    Uncertainty analysis for multi-state weighted behaviours of rural area with carbon dioxide emission estimation

    Appl Soft Comput

    (2012)
  • Chuanglin Fang et al.

    Changing urban forms and carbon dioxide emissions in China: a case study of 30 provincial capital cities

    Appl Energy

    (2015)
  • Wei Sun et al.

    Using a back propagation neural network based on improved particle swarm optimization to study the influential factors of carbon dioxide emissions in Hebei Province, China

    J Cleaner Prod

    (2016)
  • L.X. Zhang et al.

    Carbon emissions by rural energy in China

    Renew Energy

    (2014)
  • Alessio Russo et al.

    Transportation carbon dioxide emission offsets by public urban trees: a case study in Bolzano

    (2015)
  • Jane Golley et al.

    Income inequality and carbon dioxide emissions: the case of Chinese urban households

    Energy Econ

    (2012)
  • Bo Huang et al.

    Convergence of per capita carbon dioxide emissions in urban China: a spatio-temporal perspective

    Appl Geogr

    (2013)
  • Pengfei Sheng et al.

    The long-run and short-run impacts of urbanization on carbon dioxide emissions

    Econ Model

    (2016)
  • Yong Un Ban et al.

    Assessing the performance of carbon dioxide emission reduction of commercialized eco-industrial park projects in South Korea

    J Clean Prod

    (2016)
  • Yuqin Shu et al.

    Spatial disaggregation of carbon dioxide emissions from road traffic based on multiple linear regression model

    Atmos Environ

    (2011)
  • Hassan Heidari et al.

    Economic growth, CO2 emissions, and energy consumption in the five ASEAN countries

    Electr Power Energy Syst

    (2015)
  • Mohammad Salahuddin et al.

    Economic growth, energy consumption and CO2 emissions in Gulf Cooperation Council countries

    Energy

    (2014)
  • Anis Omri et al.

    Causal interactions between CO2 emissions, FDI, and economic growth: evidence from dynamic simultaneous-equation models

    Econ Model

    (2014)
  • Mak B. Arvin et al.

    Transportation intensity, urbanization, economic growth, and CO2 emissions in the G-20 countries

    Utilities Policy

    (2015)
  • Kais Saidi et al.

    The impact of energy consumption and CO2emissions on economic growth: fresh evidence from dynamic simultaneous-equations models

    Sustain Cities Soc

    (2015)
  • Ahdi Noomen Ajmi et al.

    On the relationships between CO2 emissions, energy consumption and income: The importance of time variation

    Energy Econ

    (2015)
  • Hsiao-Tien Pao et al.

    Multivariate Granger causality between CO2 emissions, energy consumption, FDI (foreign direct investment) and GDP (gross domestic product): evidence from a panel of BRIC (Brazil, Russian Federation, India, and China) countries

    Energy

    (2011)
  • Volker Krey et al.

    Urban and rural energy use and carbon dioxide emissions in Asia

    Energy Econ

    (2012)
  • Yi Chen et al.

    Uncertainty analysis for multi-state weighted behaviours of rural area with carbon dioxide emission estimation

    Appl Soft Comput

    (2012)
  • Chuanglin Fang et al.

    Changing urban forms and carbon dioxide emissions in China: a case study of 30 provincial capital cities

    Appl Energy

    (2015)
  • L. Ornella et al.

    Supervised machine learning and heterotic classification of maize (< i> Zea mays</i> L.) using molecular marker data

    Comput Electron Agric

    (2010)
  • P. Jain et al.

    Supervised machine learning algorithms for protein structure classification

    Comput Biol Chem

    (2009)
  • Y. Bao et al.

    A PSO and pattern search based memetic algorithm for SVMs parameters optimization

    Neurocomputing

    (2013)
  • F. Friedrichs et al.

    Evolutionary tuning of multiple SVM parameters

    Neurocomputing

    (2005)
  • A.C. Lorena et al.

    Evolutionary tuning of SVM parameter values in multiclass problems

    Neurocomputing

    (2008)
  • W.-Z. Lu et al.

    Potential assessment of the “support vector machine” method in forecasting ambient air pollutant trends

    Chemosphere

    (2005)
  • T. Asefa

    Multi-time scale stream flow predictions: the support vector machines approach

    J Hydrol

    (2006)
  • Y. Ji et al.

    Multitask multiclass support vector machines: model and experiments

    Pattern Recognit

    (2013)
  • Hikmet Esen et al.

    Modeling a ground-coupled heat pump system by a support vector machine

    Renew Energy

    (2008)
  • Hikmet Esen et al.

    Modelling of a new solar air heater through least-squares support vector machines

    Exp Syst Appl

    (2009)
  • Cited by (0)

    View full text