Geographically Weighted Regression Model Research Articles

PENERAPAN METODE GEOGRAPHICALLY WEIGHTED REGRESSION (GWR) PADA KASUS PENYAKIT COVID-19 DI PROVINSI BALI

COVID-19 has spread widely to all parts of the world, including Indonesia. Bali is included in the top ten with the highest number of COVID-19 cases in Indonesia. The spread of COVID-19 is thought to be influenced by various factors in each location that cause spatial heterogeneity. The method that can be used to analyze if there is spatial heterogeneity is Geographically Weighted Regression (GWR). This study aims to model the factors that influence the number of COVID-19 cases in Bali. The results showed that the GWR model with the adaptive kernel bisquare weighting function was more suitable to be used to model the number of COVID-19 cases in Bali because it had the largest value of 96.36%. The factors that influence the number of COVID-19 cases in Bali are population density and the population aged 20-44 years old.

Temporal and Spatial Changes of Habitat Quality and Their Potential Driving Factors in Southwest China

Climate change and human activities have considerably changed the spatial patterns and functional elements of regional habitats. Understanding spatiotemporal changes in habitat quality (HQ) and their potential driving factors is essential for maintaining ecosystem health and protecting biodiversity. To explore the effect of physical and human factors on HQ changes in Southwest China, we firstly analyzed the land-use change intensity (LCI). We then evaluated spatiotemporal changes in HQ based on the InVEST model and explored the spatial heterogeneity of the main driving factors of HQ changes based on a geographical detector and a geographical weighted regression model. The results showed that LCI had obvious spatiotemporal differences, and LCI from low-quality habitat to high-quality habitat (LCI1) was significantly higher than that from high-quality habitat to low-quality habitat (LCI2). The HQ improved steadily in Southwest China in 1990–2015, showing a trend of low–high–low from southeast to northwest. Moreover, there were twelve factors, including aboveground biomass, ecological land area ratio, population density, slope, etc., which had a significant impact on the spatial differences in HQ, and the effects of different factors on HQ had observable spatial heterogeneity. The effect of LCI2 on the spatial difference of HQ was greater than that of LCI1. These results suggested that the current ecosystem protection and management policy had a positive effect on improving HQ. Our study provides an important decision-making reference for sustainable land development and utilization and regional ecological protection and restoration.

Spatial patterns in sociodemographic factors explain to a large extent the prevalence of hypertension and diabetes in Aragon (Spain).

The global burden of multi-morbidity has become a major public health challenge due to the multi stakeholder action required to its prevention and control. The Social Determinants of Health approach is the basis for the establishment of health as a cross-cutting element of public policies toward enhanced and more efficient decision making for prevention and management. To identify spatially varying relationships between the multi-morbidity of hypertension and diabetes and the sociodemographic settings (2015-2019) in Aragon (a mediterranean region of Northeastern Spain) from an ecological perspective. First, we compiled data on the prevalence of hypertension, diabetes, and sociodemographic variables to build a spatial geodatabase. Then, a Principal Component Analysis (PCA) was performed to derive regression variables, i.e., aggregating prevalence rates into a multi-morbidity component (stratified by sex) and sociodemographic covariate into a reduced but meaningful number of factors. Finally, we applied Geographically Weighted Regression (GWR) and cartographic design techniques to investigate the spatial variability of the relationships between multi-morbidity and sociodemographic variables. The GWR models revealed spatial explicit relationships with large heterogeneity. The sociodemographic environment participates in the explanation of the spatial behavior of multi-morbidity, reaching maximum local explained variance (R2) of 0.76 in men and 0.91 in women. The spatial gradient in the strength of the observed relationships was sharper in models addressing men's prevalence, while women's models attained more consistent and higher explanatory performance. Modeling the prevalence of chronic diseases using GWR enables to identify specific areas in which the sociodemographic environment is explicitly manifested as a driving factor of multi-morbidity. This is step forward in supporting decision making as it highlights multi-scale contexts of vulnerability, hence allowing specific action suitable to the setting to be taken.

Spatial Heterogeneity and Influence Factors of Traditional Villages in the Wuling Mountain Area, Hunan Province, China Based on Multiscale Geographically Weighted Regression

Traditional villages are the spatial carriers of the excellent traditional culture of the Chinese nation, the concentrated expression of thousands of years of farming civilization, and a nonrenewable precious cultural heritage. However, in recent decades, with the rapid development of urbanization and modernization, traditional villages have rapidly declined or even disappeared. Therefore, exploring the spatial distribution characteristics and influencing factors of traditional villages and clarifying their formation and development laws, can provide a new perspective for the continuity protection of traditional villages on a cross-regional scale. This study takes six batches of 462 traditional villages in the Wuling Mountains, Hunan, China as the research object. On the basis of the analysis of the spatial distribution characteristics of overall and different ethnic minority traditional villages, this research proposes the idea of simulating the spatial distribution mechanism of traditional villages by using the multiscale geographical weighted regression (MGWR) model. Results indicate that (1) the spatial distribution of the traditional villages of the overall and different ethnic minorities shows significant agglomeration characteristics (among them, traditional Miao and Tujia villages have formed the main and sub cores of the space gathering of traditional villages in the Wuling Mountains in Hunan); and (2) compared with the least squares (OLS) method and the classic geographical (GWR) weighted regression model, the MGWR model can reflect the impact scales of different factors, and the simulation effect is better. From the perspective of influence factors, natural factors are the basis of affecting the distribution of traditional villages in the Wuling Mountain area and are generally positively correlated. Socioeconomic factors have a dual influence on the spatial distribution of traditional villages and generally have a negative correlation. Cultural factors are the key to the spatial distribution of traditional villages and are generally positively correlated. Spatial factors have no significant impact on the spatial distribution of traditional villages. On the basis of these conclusions, this study discusses the relevant strategies for the protection and development of traditional villages.

L0-Norm Variable Adaptive Selection for Geographically Weighted Regression Model

A geographically weighted regression (GWR) model with fewer explanatory variables and higher prediction accuracy is required in spatial analysis and other practical applications. This article proposes an -norm variable adaptive selection method to enhance performances of a GWR by simultaneously performing model selection and coefficient optimization. Specifically, we formulate a regularized GWR model with an additional -norm constraint to shrink those unimportant regression coefficients toward zero and propose an adaptive variable selection algorithm by iteratively distinguishing the important variables from the variable set. At each location, the best variable subset and optimizing coefficient estimations are simultaneously achieved under the -GWR framework. Moreover, two novel criteria, the modified Bayesian information criterion and the interpretability of coefficient symbol, which specify the variable selection and model interpretation, respectively, are also introduced to improve the performance of the -GWR. Experiments on both simulated and actual data sets demonstrate that the proposed algorithm can significantly improve the estimation accuracy of coefficients and can also enhance the interpretative ability of the established model.

Spatiotemporal pattern evolution and influencing factors of green innovation efficiency: A China’s city level analysis

Based on employing the global super efficiency epsilon-based measure (GSE-EBM) model to evaluation the green innovation efficiency (GIE) of 285 prefecture-level or above cities in China during the period 2004–2018, this paper combines the approaches of kernel density estimation, cold hot spot analysis and standard deviation ellipse to intuitively describe GIE’s spatiotemporal pattern evolution features, and then utilizes the geographical weighted regression (GWR) model to explore the spatial heterogeneity of GIE’s affecting factors. The results show that: (1) China’s urban GIE displayed a fluctuating increasing trend, revealing clearly regional disparities, and gradually decreased from the Eastern coastal region to the Central, the Western and the Northeast region. (2) The spatial difference of China’s urban GIE exhibited the characteristics of expansion, polarization, and spatial agglomeration with the center of gravity gradually shifting to the Southeast region. (3) In the analysis of socio-economic factors of China’s urban GIE, the GWR model effectively identified the spatial heterogeneity, and improved the explanatory ability compared to ordinary least squares (OLS) model. (4) The GWR model analysis indicate that population density, economic development, transportation infrastructure, openness and industrial structure played significant impacts on China’s urban GIE, and there exists significant spatial heterogeneity in the impact of each influencing factor. The findings of this study can provide valuable references for urban green transformation and high-quality development in China.

Assessment of Influence Mechanisms of Built Environment on Street Vitality Using Multisource Spatial Data: A Case Study in Qingdao, China

Street vitality is a significant indicator of a city’s capacity for sustainable development. Significant progress has been made on the basis of measurements of a single indicator of street vitality, but few studies have used multisource data to measure street vitality in a comprehensive way. In this study, in order to explore the multidimensional vitality characteristics of streets, streets were taken as the analysis unit, and the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) evaluation model with combined weights was used to identify the spatial pattern of streets vitality from social, economic, and cultural dimensions using multisource spatial data such as Baidu heat map, Meituan store rating, and cultural facilities points of interest in the main urban area of Qingdao City, China. Using a Multiscale Geographically Weighted Regression (MGWR) model, the spatial correlations and differences between street built environment components and multidimensional street vitality were examined, to reveal the influence mechanism of street vitality creation in each street. The study found that the comprehensive vitality of the streets in the main urban area of Qingdao City exhibits the spatial differentiation features of “weak east–west, strong central, multicenter, cluster type”. Furthermore, although commercial and public services are essential for enhancing street vitality and attracting crowds, a very high degree of functional mix has not resulted in a high degree of street vitality. Lastly, high spatial heterogeneity between built environment factors and street vitality necessitates considering the functional positioning and development basis of the street, tailoring to local conditions and policies, considering the street’s vitality development status and development needs, complementing strengths, promoting coordinated development, and releasing and enhancing the street’s vitality. Therefore, it is essential to explore street vitality and its influencing mechanisms to improve people’s quality of life and promote sustainable urban development.

Modeling Rice Production in West Java by Means Geographically Weighted Regression

In Indonesia, rice production varies from province to province, resulting in both large and small disparities between provinces. In Indonesia, East Java, Central Java and West Java Provinces have the highest rice production. In contrast to East Java and Central Java, however, the total rice consumption per year in West Java is the highest. In linear regression, the coefficients are the same for all regions, while each region sometimes has different influencing factors, resulting in spatial diversity. Consequently, the Geographically Weighted Regression (GWR) method was used to model the rice production of West Java Provincial regencies/municipalities by accounting for spatial heterogeneity. The GWR model employs the fixed bi-square kernel function as its weighting function. This model includes five explanatory variables, such as number of agricultural labor, number of used rice seed, number of two-wheel tractor, number of water pump, and number of farmer groups, with rice production as the response variable. GWR model has greater coefficient determination (96.8 percent) and smaller AIC values (920.76) than global regression. During the period of 2018-2020, the number of two-wheel tractors and the number of water pumps had the greatest impact on rice production in West Java and the number of two-wheeled tractors and the number of farmer groups variables has an effect on rice production in most regencies/municipalities in West Java. There are 11 groups of areas which has the similarity of significant predictor variables.

Modeling of COVID-19 Cases in Indonesia with the Method of Geographically Weighted Regression

The COVID-19 pandemic has spread to all corners of the world, including Indonesia. Various factors affect the spread of COVID-19 cases in an area so that the government and the community can make prevention and control efforts so that this pandemic does not spread. This study aims to model the number of COVID-19 cases in Indonesia using the Geographically Weighted Regression (GWR) method, which develops a linear regression model. The GWR model uses weights based on the location of each observation so that the model is obtained for that location. Determine the weighting on the bandwidth. Optimum bandwidth selection is obtained by minimizing the value of Cross-Validation (CV). The GWR model using a fixed bisquare kernel weighting function has an optimum bandwidth of 0.999948 with a minimum CV value of 397.076.128 with a coefficient of determination R2 of 85.1 %. The results show that the number of positive cases positively correlates with the number of patients who died from COVID-19. In contrast, the number of recovered patients negatively correlates with the number of patients who died from COVID-19.

Probable nexus between Methane and Air Pollution in Bangladesh using Machine Learning and Geographically Weighted Regression Modeling

This paper investigates the probable nexus between methane (CH4) and air pollutants, a public health hazard in Bangladesh. The hypothesis considers that the concentration of CH4 is dependent on the ten air pollutants found in the five districts in Dhaka Division, a major urban and industrial area in Bangladesh. These pollutants are: Particular matters (PM2.5), Nitrogen dioxide (NO2), Nitrogen oxide (NOx), Aerosol optical thickness (AOT), Sulfur dioxide (SO2), Carbon monoxide (CO), Ozone (O3), Black carbon (BC), Formaldehyde (HCHO) and Dust. The study applies Machine Learning (ML) technique and Geographically Weighted Regression (GWR) Modeling. Temporal CH4 datasets from the Sentinel-5P sensor are classified to estimate the annual CH4 concentration during 2019-2021.Seven supervised classifiers of ML coupled with the GWR model are used to predict the statistical and spatial relationships. CH4 increases gradually during 2018-2021 in Dhaka, Gazipur, and Munshiganj Districts. It relates differently with various air pollutants, e.g., positively with BC, Dust, NO2, PM2.5, O3, and AOT, and negatively with NOx, CO, HCHO, and SO2.This study results that Rational quadratic (RMSE-0.001, MAE-0.001, R2-0.96), Random Forest (RMSE-0.004, MAE-0.003, R2-0.91), and Stepwise regression (RMSE-0.002, MAE-0.002, R2-0.87) are the suitable method in ML. The highest goodness-of-fit (R2) of 82%-96% is found in Dhaka and Narshingdi Districts. The key findings may help formulate the appropriate action plan to mitigate ongoing and future air pollution in Bangladesh. In addition, the methodology of the research may be applicable elsewhere nationally and internationally for air pollution research.

Regional PM2.5 Estimation for Southern Ontario through Geographically Weighted Regression

In this study, a geographically weighted regression (GWR) approach was adopted to forecast regional concentration of particulate matter 2.5 (PM2.5) for the southern Ontario based on both in situ meteorological measurement and Satellite retrievals of aerosol optical depth (AOD). The correlation between monitored concentration of PM2.5 and Satellite-retrieved AOD would be quantified. The ground-level PM2.5 for South Ontario area was then predicted using GWR with AOD and meteorological variables considered as inputs. The results indicated that performance of GWR was slightly better than the ordinary least squares (OLS) model, indicating spatial variations between independent and dependent variables. Consequently, the GWR model can help us to predict the PM2.5 concentration in terms of time or region with satellite data, and also help improve satellite data inversion.

Dynamic Changes in Habitat Quality and the Driving Mechanism in the Luoxiao Mountains Area from 1995 to 2020

The strengthening of regional habitat quality is crucial to protect biodiversity and fully utilize ecosystem services such as those provided by forestry and aquatic ecosystems. However, the long-term patterns of change in the habitat quality of the Luoxiao Mountains area, which is both an important ecological barrier area and a concentrated poverty-stricken area, and the driving mechanism remain unclear. In this study, the InVEST model was used to assess the habitat quality of the Luoxiao Mountains area in 1995 to 2020, and the spatial autocorrelation model was used to explore the spatial and temporal variation and distribution characteristics of habitat quality. Further, ordinary least squares (OLS) model, geographically weighted regression (GWR) model, and random forest (RF) algorithm were combined with multidimensional datasets to explore the underlying mechanisms driving changes in habitat quality. According to the results, the habitat quality indices of the Luoxiao Mountains area in 1995, 2005, 2015, and 2020 were 0.822, 0.818, 0.817, and 0.813, respectively, with an overall decreasing trend. The RF model was the best fit for habitat quality, better than the GWR and OLS models. Physical geographic factors such as slope and precipitation, as well as socioeconomic factors such as gross domestic product, were key drivers of habitat quality in the Luoxiao Mountains. Precise implementation of ecological protection and restoration measures, improvements in the efficiency of spatial utilization, and exploration of the value of ecological products are key factors in maintaining a balance between habitat quality and economic growth into the future.

The effect of sanitation risk on toddler stunting incident with geographically weighted regression approach in Malang Regency, Indonesia

Stunting is a fundamental problem in Indonesia's human development. The risk factors for stunting are very complex, including inadequate environmental sanitation. This study aims to analyze and map the influence of spatially correlated sanitation risk on the incidence of stunting toddlers in Malang Regency, Indonesia. Data collection was carried out by interviewing respondents using questionnaires, counting stunting toddlers, and observing environmental sanitation objects. Determination of respondents was done purposively with the following stages. First, at the village level, 75 sites were selected based on the following criteria: population density; poverty level; the presence of streams, drainage or irrigation canals; and flood vulnerability. Furthermore, in each village, 40 respondents were selected based on the following criteria: women aged between 18 to 65 years, came from poor families and had a child (or children). When the number of respondents who were willing to participate is less than 40 people in each village, they were replaced by other respondents based on recommendations from village officials. A total of 3000 respondents have participated in this study. The selected respondents include wives, dependent married women or widows. The spatial analysis approach was carried out using the Geographically Weighted Regression (GWR) method. Based on the results of the analysis of the spatial effect of the GWR model, the Sanitation Risk Factors that significantly affected the incidence of stunting in Malang Regency were; Environmental Drainage (greywater), Solid Waste, and Domestic Wastewater (black water). Clean and Healthy Lifestyles and Water Sources have less effect on sanitation risk. This study recommends government and public interventions to improve environmental sanitation including integrated waste management, improvement of drainage infrastructure, construction of standardized toilets, and carrying out health promotion efforts with education to increase public awareness.

Pemodelan Geographically Weighted Regression pada Kasus Stunting di Provinsi Nusa Tenggara Timur Tahun 2020

Stunting or short toddlers is one of the problems that Indonesia is currently facing. Based on data from the Indonesia Health Profile in 2020, the highest prevalence of stunting in Indonesia in 2020 was in the East Nusa Tenggara Province. The occurrence of stunting in each district/city has the possibility of having different characteristics from one region or another which is referred to as regional heterogeneity or spatial heterogeneity. This research was conducted to determine the factors that influence stunting cases in each district/city of East Nusa Tenggara Province. This research uses the Geographically Weighted Regression (GWR) method to model stunting cases in each district/city of East Nusa Tenggara Province. The factors that significantly affect stunting in districts/cities in East Nusa Tenggara Province spatially with α =10% percentage of infants receiving complete immunization, the percentage of poor people, the percentage of infants receiving exclusive breastfeeding, the percentage of women who have graduated from senior high school, and the percentage of women who have ever married underage. The modeling of stunting cases in East Nusa Tenggara Province using the GWR method obtained R2 of 99,25% larger than the OLS model of 52,1%, and the AIC of the GWR model of 59,8105 smaller than OLS model.

KONUT PİYASASINDA MEKANSAL HETEROJENLİK: ANKARA METROPOLİTEN ALANI

Advanced statistical models have been widely used in real estate valuations for various purposes over the last fifty years, and hedonic approaches with their simple and easy interpretable features are still the most popular among these models. However, spatial heterogeneity and spatial autocorrelation are the two major features of the housing markets, and traditional regression cannot reflect these locational effects into the model sufficiently. This study employs a Geographically Weighted Regression (GWR) model to explore the spatial heterogeneity in the metropolitan area housing market in the city of Ankara. By applying a Gaussian kernel weighting function with adaptive bandwidth based on cross-validation approach on a house listing dataset, it is found that the GWR fit the data better than the traditional ordinary least squares regression which mostly ignore the spatial effects, and there is spatial heterogeneity in the housing market. Explanatory power of the GWR model and parameter estimations are non-stationary over the geographical area. The variations in the coefficients of the variables are depicted on the map and is supported with the spatial correlations between the housing prices and attributes as well.

Understanding the shared working spaces: a Geography of Co-working supply in Delhi.

The present paper investigates the location pattern of co-working spaces in Delhi which is absent in the existing body of knowledge. Delhi is a political, administrative, educational, scientific and innovation capital that accommodates many co-working spaces in India. We developed Ordinary least squares (OLS) and geographically weighted regression (GWR) models to understand the associations of co-working spaces of digital labourers with other urban socio-economic, services and lifestyle variables in Delhi using secondary data for 117 coworking locations in 280 municipal wards of NCT-Delhi. Model diagnostic suggested that the GWR model provides additional information regarding geographical distribution of coworking spaces, and density of bars, median house rent, fitness centres, metro train stations, restaurants, cinemas, cafés, and creative enterprises are statistically significant parameters to estimate them. The importance of coworking spaces has increased in the post-disaster period, so this study informs public policies to benefit people and companies who choose coworking routes, and recommends urban planners, developers, and real-estate professionals to consider the proximity of creative industries in planning and developing coworking spaces in the future. Also, in the post COVID-19 period, to increase local jobs and long-term place sustainability, a localised policy intervention for coworking spaces in Delhi is highly recommended.

Geographically Weighted Regression dalam Menganalisis Faktor-Faktor yang Mempengaruhi Kasus Tuberkulosis di Sumatera Utara

Geographically Weighted Regression (GWR) is a method used to analyze a case by considering spatial effects. Spatial effects indicate differences in environmental characteristics between regions. A spatial effect can be seen from the violation of the assumption of the multiple linear regression homoscedasticity test. Research in the health sector often shows that the spread of disease, especially infectious diseases, is influenced by spatial effects, one of which is tuberculosis. In 2020, North Sumatra will become one of the areas with high tuberculosis cases, so it is necessary to study tuberculosis cases and determine the factors that influence it by involving spatial effects. This study aims to model tuberculosis cases and analyze the ten factors thought to influence it using the GWR method with the adaptive kernel bi-square weighting function. The results showed that 23 regencies/cities had local GWR models and different factors influencing tuberculosis cases for each region, and ten other regencies/cities were influenced by other factors outside the variables studied. Based on the variables that affect it, 13 regional groups are formed. Modeling with the GWR method can explain tuberculosis cases in North Sumatra by 99.4%, and other predictor variables influence the rest.

A novel approach to estimating urban land surface temperature by the combination of geographically weighted regression and deep neural network models

Growing concerns on excessive urban heat call for better approaches to modeling urban thermal environment and developing effective mitigation strategies. A hybrid model integrating the geographically weighted regression (GWR) and deep neural network (DNN) was developed to estimate land surface temperature (LST). This model was compared with three other data-driven approaches to predicting LST, including the ordinary least squares (OLS) regression, GWR, and DNN. Sixteen satellite image datasets (a total of 155,728 data points) during a four-year period in Hong Kong were used for model development, validation, and comparison. The datasets cover two distinguishable geographical regions and consist of sixteen explanatory variables from five groups, including (1) land use and land cover (LULC) composition and surface characteristics, (2) LULC configuration, (3) urban form, (4) anthropogenic activities, and (5) location and local climate. The results indicate that the hybrid model performs the best in terms of model fitness and prediction accuracy, with R2 equal to 0.85 and 0.73 and the mean squared error (MSE) equal to 0.52 and 0.70 in the two regions, respectively. Compared to the OLS, DNN, and GWR models, the overall R2 for all the datasets of the hybrid model increases by 97.3%, 16.6%, and 6.9%, respectively. The promising result of the hybrid model is due to its ability to capture both spatial heterogeneity and address possible correlations between explanatory variables. Sensitivity of LST to various explanatory variables is also discussed and strategies to mitigate excessive heat are recommended. This study is anticipated to contribute to model development in urban LST estimation and quantitative evaluation of those factors driving LST variations.

Categorical Variable Problem In Real Estate Submarket Determination With Gwr Model

Abstract Real estate market analysis can involve many aspects. One of them is the study of the influence of various factors on prices and property values. For this type of issues, different kinds of measures and statistical models are often used. Many of them do not give unambiguous results. One of the reasons for this is the fact that the real estate market is characterized by the concept of local markets, which may be affected in different ways by economic, social, technical, environmental and other factors. Incorporating the influence of local markets, otherwise known as submarkets, into models often helps improve the precision of mass real estate valuation results. The delineation of submarket boundaries can be done in several different ways. One tool that is helpful in these types of situations are geographically weighted regression (GWR) models. The problem that may arise when using such models is related to the nature of some market factors, which may be of a qualitative nature. Because neighborhoods of individual properties may lack variability in terms of some variables, estimating GWR models is significantly difficult or impossible. The study will present an approach in which the categorical variables are transformed into a single synthetic variable, and only this variable will constitute the explanatory variable in the model. Areas where the slope parameters of the GWR model are similar were considered a submarket. The purpose of this paper is to determine the boundaries of submarkets in the study area and to compare the results of modeling the value of real estate using models that do not take local markets into account, as well as those that take into account local markets determined by experts and using the GWR model.

Social status and air quality in Barcelona: A socio-ecological approach

• There is a relationship between higher building density and higher air pollution. • There is an inverse relationship between the isolated street trees and less pollution. • The areas with higher income level are related with the worse the air quality • The GWR models show a more complete view that OLS models for the air quality studies. During the last two decades, many studies have described the effects of air pollution in urban areas. This paper analyses the spatial patterns of the main air pollutants in the city of Barcelona, i.e. nitrogen dioxide (NO 2 ), fine particulates of less than 2.5μm (PM 2.5 ) and coarse particulates of 10μm (PM10), and explores the factors associated with air quality in relation to the social and environmental characteristics of each area of the city using 1068 census tracts. The city exhibits a situation of high annual averages of pollutants regarding the World Health Organization (WHO) guidelines and the European Union regulations, with most of the census tracts with amounts of NO 2 >20 µg/m 3 , PM 2.5 >10 µg/m 3 and PM 10 >20 µg/m 3 . To analyse the spatial pattern, we use exploratory spatial data analysis (ESDA) and estimation of spatial econometric models. We test global and local spatial autocorrelation, and we identify spatial clusters of pollutants with the Local Moran cluster map. We use different geographically weighted regression (GWR) approaches to provide local insight and to consider the spatial heterogeneity of the relationships analysed, including standard GWR, multiscale and mixed techniques. Our results for Barcelona reveal a clear relationship, but with opposite signs depending on the census tract, between the pollutants commonly analysed and population density (from -2.17 to 3.69, median 1.3), income (from -7.20 to 20.55, median 6.45) and population aged 0-4 (from -1.96 to 12.38, median 0.74).