Semi-parametric Geographically Weighted Regression Research Articles

Exploring the Spatially Heterogeneous Effects of the Built Environment on Bike-Sharing Usage during the COVID-19 Pandemic

Bike-sharing holds promise for available and healthy mobility services during COVID-19 where bike sharing users can make trips with lower health concerns due to social distancing compared to the restricted transportation modes such as public transit and ridesharing services. Leveraging the trip data of the Divvy bike-sharing system in Chicago, this study exploresspatially heterogeneous effects of built environment on bike-sharing usage under the pandemic. Results show that the average weekly ridership declined by 52.04%. To account for the spatially heterogeneous relationship between the built environment and the ridership, the geographically weighted regression (GWR) model and the semiparametric GWR (S-GWR) model are constructed. We find that the S-GWR model outperforms the GWR and the multiple linear regression models. The results of the S-GWR model indicate that education employment density, distance to subway, COVID-19 cases, and ridership before COVID-19 are global variables. The effects between ridership and the built environment factros (i.e., household density, office employment density, and the ridership) vary across space. The results of this study could provide a useful reference to transportation planners and bike-sharing operators to determine the high bike-sharing demand area under the pandemic,thus adjusting station locations, capacity, and rebalancing schemes accordingly.

Voter Turnout in Portugal: A Geographical Perspective

The decline of voter turnout in Portugal was confirmed in the legislative election of 2015. The unquestionable democratic value associated with the act of voting, leads to the discussion of this issue, and emphasizes the need for additional investigation. Particularly, it is crucial to identify the characteristics of citizens who vote, to better understand the phenomenon and think about solutions. This work identified the most significant sociodemographic variables in explaining voter turnout in continental Portugal and described the relationship between those variables and voter turnout, including the geographical variation existing across the municipalities. A Semiparametric Geographically Weighted Regression (SGWR) model enabled the investigation of local variations in turnout values, simultaneously considering that its relationship with some variables might vary over space. Results show that turnout is influenced by a set of sociodemographic variables. While some variables affect turnout differently over the country (percentage of family cores with children aged less than 15, and percentage of owner-occupied houses), others affect it uniformly (percentage of graduated residents, percentage of classic families, and distance to Lisbon or Oporto – the nearest). These results suggest the use of a semiparametric approach to better understand turnout and for further research on voting issues.

PEMODELAN SEMIPARAMETRIC GEOGRAPHICALLY WEIGHTED REGRESSION PADA KASUS PNEUMONIA BALITA PROVINSI JAWA TENGAH

Geographical and inter-regional differences have contributed to the diversity of child pneumonia cases in Central Java, so a spatial regression modelling is formed that is called Geographically Weighted Regression (GWR). GWR is a development of linear regression by involving diverse factors geographical location, so that local parameters are produced. Sometimes, there are non-local GWR parameters. To overcome some non-local parameters, Semiparametric Geographically Weighted Regression (SGWR) is formed to develop a GWR model with local and global influences simultaneously. SGWR Model is used to estimate the model of percentage of children with pneumonia in Central Java with population density, average temperature, percentage of children with severe malnutrition, percentage of children with under the red line weight, percentage of households behave in clean and healthy lives, and percentage of children who measles immunized. SGWR models on percentage of children with pneumonia in Central Java produce locally significant variables that is population density, average temperature, and percentage of households behave in clean and healthy lives. Variable that globally significant is percentage of children with severe malnutrition. Based on Akaike Information Criterion (AIC), SGWR is a better model to analize percentage of children with pneumonia in Central Java because of smallest AIC. Keywords: Akaike Information Criterion, Geographically Weighted Regression, Semiparametric Geographically Weighted Regression

Spatial hedonic analysis to support tourism-sensitive tsunami mitigation planning

The construction of protective infrastructure such as seawalls and the designation of land-use and building regulation zones are commonly recommended strategies for tsunami and sea-level rise oriented mitigation in coastal areas. While the effectiveness of these strategies are undoubted, in regions where coastal tourism is the primary industry, the implementation of such strategies have been low due to fear of negative economic impact related to loss of coastal view and accessibility. Therefore, this paper examines the influence of coastal amenities to hotel room rates alongside other attributes through hedonic analysis. Specifically, it investigates whether rooms with coastal views, accessibility to beaches, and those located on higher elevations are priced higher than other rooms, in order to quantify the associated values of Japanese coastal areas where tourism is a key economic driver. Subsequently, it suggests the geographical market boundaries to guide the management and risk-mitigation of coastal areas. Findings reveal that: Semi-parametric Geographically Weighted Regression (S-GWR) results can accurately identify both stationary and non-stationary relationships present between the dependent and explanatory variables; view of the sea and other environmental attributes have significant influence on hotel room pricings; and tsunami mitigation strategies which can have long-term implications should be adopted in a manner sensitive to the tourism industry.

Spatial distribution of carbon storage in natural secondary forest based on geographically weighted regression expansion model.

To accurately assess carbon storage and its spatial distribution in natural secondary forest at the regional scale, we constructed seven expansion models by modifying the geographically weighted regression (GWR) in aspects of spatial dimension, parameter heterogeneity and residual spatial autocorrelation, based on data collected from 165 bureau level permanent plots in Langxi Forest Farm of Wangqing Forestry Bureau in Jilin Province. Stand factor, topography factor, and soil factor were selected as the influencing factors. The expansion models included geographically and altitudinal weighted regression (GAWR), semiparametric geographically weighted regression (SGWR), semiparametric geographically and altitudinal weighted regression (SGAWR), geographically weighted regression Kriging (GWRK), geographically and altitudinal weighted regression Kriging (GAWRK), semiparametric geographically weighted regression Kriging (SGWRK), and semiparametric geographically and altitudinal weighted regression Kriging (SGAWRK). Coefficient of determination (R2), mean square error (MSE) and Akaike's Information Criterion (AIC) were used to evaluate the fitness of these models. Finally, the spatial distribution diagram of forest carbon storage was drawn with the fitting results of the optimal regression model, and the distribution pattern of forest carbon storage in the research area was analyzed. The stand factor and topographic factor had strong influence on carbon storage of natural secondary forests, among which the average diameter at breast height (DBH) of stands was the dominant variable. There was positive correlation between stand factor and topographic factor. SGWR and SGAWR model could reduce the spatial autocorrelation of the GWR model residual. The geographically regression expansion model could improve the fitting effect of GWR model. Among them, the SGWRK model had the highest R2 and the lowest MSE and AIC. The method with altitude as the spatial weight did not effectively improve the fitting effect of the model. The total forest carbon storage of Langxi Forest Farm was 205×104 t, and the carbon density ranged from 8.56 to 145.74 t·hm-2, with a mean value of 57.98 t·hm-2. Overall, the distribution pattern of carbon storage was high in the northwest and low in the southeast, while high in the edge and low in the interior. By improving the parameter heterogeneity and residual spatial autocorrelation in the GWR model, we can accurately assess the spatial relationship between forest carbon storage and relevant variables in the study area, and improve the estimation accuracy of the forest carbon storage and its spatial distribution at the regional scale.

The Diversified Impacts of Urban Morphology on Land Surface Temperature among Urban Functional Zones.

Local warming induced by rapid urbanization has been threatening residents’ health, raising significant concerns among urban planners. Local climate zone (LCZ), a widely accepted approach to reclassify the urban area, which is helpful to propose planning strategies for mitigating local warming, has been well documented in recent years. Based on the LCZ framework, many scholars have carried out diversified extensions in urban zoning research in recent years, in which urban functional zone (UFZ) is a typical perspective because it directly takes into account the impacts of human activities. UFZs, widely used in urban planning and management, were chosen as the basic unit of this study to explore the spatial heterogeneity in the relationship between landscape composition, urban morphology, urban functions, and land surface temperature (LST). Global regression including ordinary least square regression (OLS) and random forest regression (RF) were used to model the landscape-LST correlations to screen indicators to participate in following spatial regression. The spatial regression including semi-parametric geographically weighted regression (SGWR) and multiscale geographically weighted regression (MGWR) were applied to investigate the spatial heterogeneity in landscape-LST among different types of UFZ and within each UFZ. Urban two-dimensional (2D) morphology indicators including building density (BD); three-dimensional (3D) morphology indicators including building height (BH), building volume density (BVD), and sky view factor (SVF); and other indicators including albedo and normalized difference vegetation index (NDVI) and impervious surface fraction (ISF) were used as potential landscape drivers for LST. The results show significant spatial heterogeneity in the Landscape-LST relationship across UFZs, but the spatial heterogeneity is not obvious within specific UFZs. The significant impact of urban morphology on LST was observed in six types of UFZs representing urban built up areas including Residential (R), Urban village (UV), Administration and Public Services (APS), Commercial and Business Facilities (CBF), Industrial and Manufacturing (IM), and Logistics and Warehouse (LW). Specifically, a significant correlation between urban 3D morphology indicators and LST in CBF was discovered. Based on the results, we propose different planning strategies to settle the local warming problems for each UFZ. In general, this research reveals UFZs to be an appropriate operational scale for analyzing LST on an urban scale.

Global and localized neighborhood effects on public transit ridership in Baton Rouge, Louisiana

This study examines what socio-demographic and spatial factors explain the variation of public transit ridership in a medium-size city in southern U.S. – Baton Rouge, Louisiana. In order to gain a sharper spatial resolution in the analysis, the ecological inference method is used to disaggregate socio-demographic data from the census block group level to the census block level. Monte Carlo simulation and transit schedule data are used to improve the estimation of travel time by private vehicle and public transit, respectively, also at the census block level. The semi-parametric geographically-weighted regression (SGWR) is used to identify, among significant variables, what are global factors and what are local factors. The results indicate that neighborhoods with higher concentrations of non-White minorities, recent immigrants and carless households have positive global effects on public transit ridership. The effects by neighborhood median income, public transit to private vehicle commuting time ratio, and accessibility to employment via public transit are localized or vary across the study area, and some of these variables even show opposite effects in specific pockets in contrast to their area-wide average effects.

A Semi-Parametric Geographically Weighted Regression Approach to Exploring Driving Factors of Fractional Vegetation Cover: A Case Study of Guangdong

Ecological degradation caused by rapid urbanisation has presented great challenges in southern China. Fractional vegetation cover (FVC) has long been the most common and sensitive index to describe vegetation growth and to monitor vegetation degradation. However, most of the studies have failed to adequately explore the complexity of the relationship between fractional vegetation cover (FVC) and impact factors. In this research, we first constructed a Semi-parametric Geographically Weighted Regression (SGWR) model to analyse both the stationary and nonstationary spatial relationships between FVC and driving factors in Guangdong province in southern China on a county level. Then, climate, topographic, land cover, and socio-economic factors were introduced into the model to distinguish impacts on FVC from 2000–2015. Results suggest that the positive and negative effects of rainfall and elevation coefficients alternated, and local urban land and population estimates indicated a negative association between FVC and the modelled factors in each period. The SGWR FVC make significantly improves performance of the geographically weighted regression and ordinary least squares models, with adjusted R2 higher than 0.78. The findings of this research demonstrated that, although urbanisation in the Pearl River Delta in Guangdong has encroached on the regional vegetation cover, the total vegetation area remained unchanged with the implementation of protection policies and regulations.

Spatiotemporal Patterns of Population Mobility and Its Determinants in Chinese Cities Based on Travel Big Data

Large-scale population mobility has an important impact on the spatial layout of China’s urban systems. Compared with traditional census data, mobile-phone-based travel big data can describe the mobility patterns of a population in a timely, dynamic, complete, and accurate manner. With the travel big dataset supported by Tencent’s location big data, combined with social network analysis (SNA) and a semiparametric geographically weighted regression (SGWR) model, this paper first analyzed the spatiotemporal patterns and characteristics of mobile-data-based population mobility (MBPM), and then revealed the socioeconomic factors related to population mobility during the Spring Festival of 2019, which is the most important festival in China, equivalent to Thanksgiving Day in United States. During this period, the volume of population mobility exceeded 200 million, which became the largest time node of short-term population mobility in the world. The results showed that population mobility presents a spatial structure dominated by two east–west main axes formed by Chengdu, Nanjing, Wuhan, Shanghai; and three north–south main axes formed by Guangzhou, Shenzhen, Shanghai, Wuhan, and Chengdu. The major cities in the four urban agglomerations in China occupy an absolute core position in the population mobility network hierarchy, and the population mobility network presents typical “small world” features and forms 11 closely related groups. Semiparametric geographically weighted regression model results showed that mobile-data-based population mobility variation is significantly related to the value-added of secondary and tertiary industries, foreign capital, average wage, urbanization rate, and value-added of primary industries. When the spatial heterogeneity and nonstationarity was considered, the socioeconomic factors that affect population mobility showed differences between different regions and cities. The patterns of population mobility and determinants explored in this paper can provide a new reference for the balanced development of regional economy.

Exploring spatial variation of bike sharing trip production and attraction: A study based on Chicago’s Divvy system

Bike sharing systems are adopted by many cities due to its contribution to energy saving and mitigation of traffic congestion. Understanding factors that influence bike sharing ridership and accurate estimation of ridership play an important role in designing the system. Previous studies assume the relationship between predicting variables and the response variable are the same across the study area. However, this assumption may not be true, since the study area is usually wide and thus the relationship between predicting variabels and the response variable may change across space. As a result, semi-parametric geographically weighted regression (S-GWR) model is used to explore the spatially varying relationship. S-GWR is an extension of the GWR model. While in GWR model, all predicting variables are local variables with spatially varying relationship with the response variable, S-GWR model allows predicting variables to be either global or local, which is closer to reality. We also extend previous studies by differenciating members and 24-h pass users, as well as data related to trip production and trip attraction. Results show that S-GWR models fit the data better and the relationship between some predicting variables and response variable are local while other relationships are global. Ridership of both members and 24-h users are positively related to number of employed residents nearby and capacity of the station, and negatively related to distance to central business area and percent of low-income workers living nearby. Number of employments is only significantly associated with trip attraction. Among them, the variable capacity is always a global variable, with higher capacity associated with higher ridership. As a result, S-GWR model could be used to estimate the ridership of stations for accurate prediction and spatially varying relationship between ridership and influencing factors should be considered when designing bike sharing system.

Spatiotemporal Differences in Determinants of City Shrinkage Based on Semiparametric Geographically Weighted Regression

City shrinkage, as an ongoing worldwide phenomenon, is an issue for urban planning and regional development. City shrinkage is remarkable in Japan, with over 85% of municipalities experiencing population loss from 2005 to 2015. As Japan’s society ages and with its low fertility rate, city shrinkage has had a tremendous negative effect on economic development and urban planning. Understanding the spatial dependence and spatial heterogeneity of city shrinkage and its determinants is essential for ensuring the sustainable development of a city or region. In this study, a semiparametric geographically weighted regression (SGWR) model was adopted to explore the spatiotemporal differences in determinants of city shrinkage. The results reveal that the SGWR model incorporating the global and local variables is more interpretive compared to ordinary least squares and geographically weighted regression models in exploring the correlates of city shrinkage. We found the spatial dependence and heterogeneity of shrinking cities resulted from demographic, economy, and social factors, and revealed low fertility, the ageing population, and enterprise change ratio influenced city shrinkage in different regions at different times in Japan, whereas foreign population ratio, industry structure, and social welfare had global impacts. The findings provide useful information for understanding city shrinkage at global and local scales.

Estimating Urban Shared-Bike Trips with Location-Based Social Networking Data

Dockless shared-bikes have become a new transportation mode in major urban cities in China. Excessive number of shared-bikes can occupy a significant amount of roadway surface and cause trouble for pedestrians and auto vehicle drivers. Understanding the trip pattern of shared-bikes is essential in estimating the reasonable size of shared-bike fleet. This paper proposed a methodology to estimate the shared-bike trip using location-based social network data and conducted a case study in Nanjing, China. The ordinary least square, geographically weighted regression (GWR) and semiparametric geographically weighted regression (SGWR) methods are used to establish the relationship among shared-bike trip, distance to the subway station and check ins in different categories of the point of interest (POI). This method could be applied to determine the reasonable number of shared-bikes to be launched in new places and economically benefit in shared-bike management.

A spatial hedonic stumpage analysis of standing timber auctions in Jiangxi Province of China

To understand the local market structure of stumpage and analyze the effects of various determinants of stumpage on timber prices in Jiangxi province of China, this study develops, estimates and compares the results of global and local hedonic models of stumpage. The models are estimated using 628 transactions data of timber auctions organized by formal forest trade centres. Ordinary Least Square is used to estimate global models while semi-parametric geographically weighted regression (SGWR) for local models. The results suggest that the local (spatial) models are better fit; the influence of all (except three) determinants varies spatially, and many determinants influence stumpage positively in some locations while negatively in others. The results show that state-owned forests receive higher prices compared to private forests and neither format of auction (sealed-bid versus open) dominates in all areas. The results highlight the importance of the context in the analysis of stumpage markets. The spatial variation in the effects of determinants across diverse natural landscapes should be incorporated in the auction design and organization across geographical areas.

Linking potential heat source and sink to urban heat island: Heterogeneous effects of landscape pattern on land surface temperature.

Rapid urbanization has significantly contributed to the development of urban heat island (UHI). Regulating landscape composition and configuration would help mitigate the UHI in megacities. Taking Shenzhen, China, as a case study area, we defined heat source and heat sink and identified strong and weak sources as well as strong and weak sinks according to the natural and socioeconomic factors influencing land surface temperature (LST). Thus, the potential thermal contributions of heat source and heat sink patches were differentiated. Then, the heterogeneous effects of landscape pattern on LST were examined by using semiparametric geographically weighted regression (SGWR) models. The results showed that landscape composition has more significant effects on thermal environment than configuration. For a strong source, the percentage of patches has a positive impact on LST. Additionally, when mosaicked with some heat sink, even a small improvement in the degree of dispersion of a strong source helps to alleviate UHI. For a weak source, the percentage and density of patches have positive impacts on LST. For a strong sink, the percentage, density, and degree of aggregation of patches have negative impacts on LST. The effects of edge density and patch shape complexity vary spatially with the fragmentation of a strong sink. Similarly, the impacts of a weak sink are mainly exerted via the characteristics of percent, density, and shape complexity of patches.

Investigating the relationship between school facilities and academic achievements through geographically weighted regression

ABSTRACTIn this study, the authors investigated spatial variations in the effects of school amenities on academic achievements of a cluster of provinces in the Philippines. Semiparametric geographically weighted regression (SGWR) techniques were applied to public school facility data to determine whether the effects of school facilities varied depending on school locations. The analysis results presented significant spatial variations and differing effects of school sizes and utilities on the academic achievements of top performing schools across the study area. SGWR modelling revealed that schools in sparsely populated rural areas with basic facilities have performed better than schools in urban areas with poor facilities, indicating that basic facilities are important in far-flung schools. Since the effect on academic achievement varies depending on the social and economic infrastructure, the provision of school facilities should be based on the unique needs of each community. A decentralized approach is beneficial for the management of school facilities in the Philippines where education resources are limited.

Analysing the local geography of the relationship between residential property prices and its determinants

Abstract This paper analyses the local geography of the relationship between residential property prices and its determinants. A semiparametric geographically weighted regression (S-GWR) technique is employed to explore this relationship. Selling prices, structural and locational attributes data were collected from the database of the Department of Valuation and Services of Malaysia, selected maps and reports. The outcome of this paper shows a strong geographically varying relationship between residential property prices and its determinants in which the residential property price determinants have a positive impact on prices in some areas but negative or no impact on the others. The magnitude of the effect is also found to be geographically varied; the capitalisation in residential property prices is found greater in some areas but less or with no effect in some other parts of the areas. The use of S-GWR technique makes it possible to reveal such geographically varying relationships, thus leading to a better understanding of the relationship between residential property prices and its determinants.