Matching Items (7)
Filtering by
- All Subjects: Geography
- All Subjects: spatial statistics
- Genre: Academic theses
- Creators: Salon, Deborah
Description
It has been identified in the literature that there exists a link between the built environment and non-motorized transport. This study aims to contribute to existing literature on the effects of the built environment on cycling, examining the case of the whole State of California. Physical built environment features are classified into six groups as: 1) local density, 2) diversity of land use, 3) road connectivity, 4) bike route length, 5) green space, 6) job accessibility. Cycling trips in one week for all children, school children, adults and employed-adults are investigated separately. The regression analysis shows that cycling trips is significantly associated with some features of built environment when many socio-demographic factors are taken into account. Street intersections, bike route length tend to increase the use of bicycle. These effects are well-aligned with literature. Moreover, both local and regional job accessibility variables are statistically significant in two adults' models. However, residential density always has a significant negatively effect on cycling trips, which is still need further research to confirm. Also, there is a gap in literature on how green space affects cycling, but the results of this study is still too unclear to make it up. By elasticity analysis, this study concludes that street intersections is the most powerful predictor on cycling trips. From another perspective, the effects of built environment on cycling at workplace (or school) are distinguished from at home. This study implies that a wide range of measures are available for planners to control vehicle travel by improving cycling-level in California.
ContributorsWang, Kailai, M.U.E.P (Author) / Salon, Deborah (Thesis advisor) / Rey, Sergio (Committee member) / Li, Wenwen (Committee member) / Arizona State University (Publisher)
Created2015
Description
Bicycle sharing systems (BSS) operate on five continents, and they change quickly with technological innovations. The newest “dockless” systems eliminate both docks and stations, and have become popular in China since their launch in 2016. The rapid increase in dockless system use has exposed its drawbacks. Without the order imposed by docks and stations, bike parking has become problematic. In the areas of densest use, the central business districts of large cities, dockless systems have resulted in chaotic piling of bikes and need for frequent rebalancing of bikes to other locations. In low-density zones, on the other hand, it may be difficult for customers to find a bike, and bikes may go unused for long periods. Using big data from the Mobike BSS in Beijing, I analyzed the relationship between building density and the efficiency of dockless BSS. Density is negatively correlated with bicycle idle time, and positively correlated with rebalancing. Understanding the effects of density on BSS efficiency can help BSS operators and municipalities improve the operating efficiency of BSS, increase regional cycling volume, and solve the bicycle rebalancing problem in dockless systems. It can also be useful to cities considering what kind of BSS to adopt.
ContributorsCui, Wencong (Author) / Kuby, Michael (Thesis advisor) / Salon, Deborah (Committee member) / Thigpen, Calvin (Committee member) / Arizona State University (Publisher)
Created2018
Description
Transportation infrastructure in urban areas has significant impacts on socio-economic activities, land use, and real property values. This dissertation proposes a more comprehensive theory of the positive and negative relationships between property values and transportation investments that distinguishes different effects by mode (rail vs. road), by network component (nodes vs. links), and by distance from them. It hypothesizes that transportation investment generates improvement in accessibility that accrue only to the nodes such as highway exits and light rail stations. Simultaneously, it tests the hypothesis that both transport nodes and links emanate short-distance negative nuisance effects due to disamenities such as traffic and noise. It also tests the hypothesis that nodes of both modes generate a net effect combining accessibility and disamenities. For highways, the configuration at grade or above/below ground is also tested. In addition, this dissertation hypothesizes that the condition of road pavement may have an impact on residential property values adjacent to the road segments. As pavement condition improves, value of properties adjacent to a road are hypothesized to increase as well. A multiple-distance-bands approach is used to capture distance decay of amenities and disamenities from nodes and links; and pavement condition index (PCI) is used to test the relationship between road condition and residential property values. The hypotheses are tested using spatial hedonic models that are specific to each of residential and commercial property market. Results confirm that proximity to transport nodes are associated positively with both residential and commercial property values. As a function of distance from highway exits and light rail transit (LRT) stations, the distance-band coefficients form a conventional distance decay curve. However, contrary to our hypotheses, no net effect is evident. The accessibility effect for highway exits extends farther than for LRT stations in residential model as expected. The highway configuration effect on residential home values confirms that below-grade highways have relatively positive impacts on nearby houses compared to those at ground level or above. Lastly, results for the relationship between pavement condition and residential home values show that there is no significant effect between them.
Some differences in the effect of infrastructure on property values emerge between residential and commercial markets. In the commercial models, the accessibility effect for highway exits extends less than for LRT stations. Though coefficients for short distances (within 300m) from highways and LRT links were expected to be negative in both residential and commercial models, only commercial models show a significant negative relationship. Different effects by mode, network component, and distance on commercial submarkets (i.e., industrial, office, retail and service properties) are tested as well and the results vary based on types of submarket.
Consequently, findings of three individual paper confirm that transportation investments mostly have significant impacts on real-estate properties either in a positive or negative direction in accordance with the transport mode, network component, and distance, though effects for some conditions (e.g., proximity to links of highway and light rail, and pavement quality) do not significantly change home values. Results can be used for city authorities and planners for funding mechanisms of transport infrastructure or validity of investments as well as private developers for maximizing development profits or for locating developments.
Some differences in the effect of infrastructure on property values emerge between residential and commercial markets. In the commercial models, the accessibility effect for highway exits extends less than for LRT stations. Though coefficients for short distances (within 300m) from highways and LRT links were expected to be negative in both residential and commercial models, only commercial models show a significant negative relationship. Different effects by mode, network component, and distance on commercial submarkets (i.e., industrial, office, retail and service properties) are tested as well and the results vary based on types of submarket.
Consequently, findings of three individual paper confirm that transportation investments mostly have significant impacts on real-estate properties either in a positive or negative direction in accordance with the transport mode, network component, and distance, though effects for some conditions (e.g., proximity to links of highway and light rail, and pavement quality) do not significantly change home values. Results can be used for city authorities and planners for funding mechanisms of transport infrastructure or validity of investments as well as private developers for maximizing development profits or for locating developments.
ContributorsSeo, Kihwan (Author) / Michael, Kuby (Thesis advisor) / Golub, Aaron (Committee member) / Salon, Deborah (Committee member) / Arizona State University (Publisher)
Created2016
Description
Walking and bicycling bring many merits to people, both physically and mentally.
However, not everyone has an opportunity to enjoy healthy and safe bicycling and
walking. Many studies suggested that access to healthy walking and bicycling is heavily
related to socio-economic status. Low income population and racial minorities have
poorer transportation that results in less walking and bicycling, as well as less access to
public transportation. They are also under higher risks of being hit by vehicles while
walking and bicycling. This research quantifies the relationship between socioeconomic
factors and bicyclist and pedestrian involved traffic crash rates in order to establish an
understanding of how equitable access to safe bicycling and walking is in Phoenix. The
crash rates involving both bicyclists and pedestrians were categorized into two groups,
minor crashes and severe crashes. Then, the OLS model was used to analyze minor and
severe bicycle crash rates, and minor and severe pedestrian crash rates, respectively.
There are four main results, (1) The median income of an area is always negatively
related to the crash rates of bicyclists and pedestrians. The reason behind the negative
correlation is that there is a very small proportion of people choosing to walk or ride
bicycles as their commuting methods in the high-income areas. Consequently, there are
low crash rates of pedestrians and bicyclists. (2) The minor bicycle crash rates are more
related to socio-economic determinants than the severe crash rates. (3) A higher
population density reduces both the minor and the severe crash rates of bicyclists and
pedestrians in Phoenix. (4) A higher pedestrian commuting ratio does not reduce bicyclist
and pedestrian crash rates in Phoenix. The findings from this study can provide a
reference value for the government and other researchers and encourage better future
decisions from policy makers.
However, not everyone has an opportunity to enjoy healthy and safe bicycling and
walking. Many studies suggested that access to healthy walking and bicycling is heavily
related to socio-economic status. Low income population and racial minorities have
poorer transportation that results in less walking and bicycling, as well as less access to
public transportation. They are also under higher risks of being hit by vehicles while
walking and bicycling. This research quantifies the relationship between socioeconomic
factors and bicyclist and pedestrian involved traffic crash rates in order to establish an
understanding of how equitable access to safe bicycling and walking is in Phoenix. The
crash rates involving both bicyclists and pedestrians were categorized into two groups,
minor crashes and severe crashes. Then, the OLS model was used to analyze minor and
severe bicycle crash rates, and minor and severe pedestrian crash rates, respectively.
There are four main results, (1) The median income of an area is always negatively
related to the crash rates of bicyclists and pedestrians. The reason behind the negative
correlation is that there is a very small proportion of people choosing to walk or ride
bicycles as their commuting methods in the high-income areas. Consequently, there are
low crash rates of pedestrians and bicyclists. (2) The minor bicycle crash rates are more
related to socio-economic determinants than the severe crash rates. (3) A higher
population density reduces both the minor and the severe crash rates of bicyclists and
pedestrians in Phoenix. (4) A higher pedestrian commuting ratio does not reduce bicyclist
and pedestrian crash rates in Phoenix. The findings from this study can provide a
reference value for the government and other researchers and encourage better future
decisions from policy makers.
ContributorsWu, Feiyi (Author) / Nelson, Trisalyn (Thesis advisor) / Salon, Deborah (Committee member) / Kuby, Michael (Committee member) / Arizona State University (Publisher)
Created2020
Description
Spatial regression is one of the central topics in spatial statistics. Based on the goals, interpretation or prediction, spatial regression models can be classified into two categories, linear mixed regression models and nonlinear regression models. This dissertation explored these models and their real world applications. New methods and models were proposed to overcome the challenges in practice. There are three major parts in the dissertation.
In the first part, nonlinear regression models were embedded into a multistage workflow to predict the spatial abundance of reef fish species in the Gulf of Mexico. There were two challenges, zero-inflated data and out of sample prediction. The methods and models in the workflow could effectively handle the zero-inflated sampling data without strong assumptions. Three strategies were proposed to solve the out of sample prediction problem. The results and discussions showed that the nonlinear prediction had the advantages of high accuracy, low bias and well-performed in multi-resolution.
In the second part, a two-stage spatial regression model was proposed for analyzing soil carbon stock (SOC) data. In the first stage, there was a spatial linear mixed model that captured the linear and stationary effects. In the second stage, a generalized additive model was used to explain the nonlinear and nonstationary effects. The results illustrated that the two-stage model had good interpretability in understanding the effect of covariates, meanwhile, it kept high prediction accuracy which is competitive to the popular machine learning models, like, random forest, xgboost and support vector machine.
A new nonlinear regression model, Gaussian process BART (Bayesian additive regression tree), was proposed in the third part. Combining advantages in both BART and Gaussian process, the model could capture the nonlinear effects of both observed and latent covariates. To develop the model, first, the traditional BART was generalized to accommodate correlated errors. Then, the failure of likelihood based Markov chain Monte Carlo (MCMC) in parameter estimating was discussed. Based on the idea of analysis of variation, back comparing and tuning range, were proposed to tackle this failure. Finally, effectiveness of the new model was examined by experiments on both simulation and real data.
In the first part, nonlinear regression models were embedded into a multistage workflow to predict the spatial abundance of reef fish species in the Gulf of Mexico. There were two challenges, zero-inflated data and out of sample prediction. The methods and models in the workflow could effectively handle the zero-inflated sampling data without strong assumptions. Three strategies were proposed to solve the out of sample prediction problem. The results and discussions showed that the nonlinear prediction had the advantages of high accuracy, low bias and well-performed in multi-resolution.
In the second part, a two-stage spatial regression model was proposed for analyzing soil carbon stock (SOC) data. In the first stage, there was a spatial linear mixed model that captured the linear and stationary effects. In the second stage, a generalized additive model was used to explain the nonlinear and nonstationary effects. The results illustrated that the two-stage model had good interpretability in understanding the effect of covariates, meanwhile, it kept high prediction accuracy which is competitive to the popular machine learning models, like, random forest, xgboost and support vector machine.
A new nonlinear regression model, Gaussian process BART (Bayesian additive regression tree), was proposed in the third part. Combining advantages in both BART and Gaussian process, the model could capture the nonlinear effects of both observed and latent covariates. To develop the model, first, the traditional BART was generalized to accommodate correlated errors. Then, the failure of likelihood based Markov chain Monte Carlo (MCMC) in parameter estimating was discussed. Based on the idea of analysis of variation, back comparing and tuning range, were proposed to tackle this failure. Finally, effectiveness of the new model was examined by experiments on both simulation and real data.
ContributorsLu, Xuetao (Author) / McCulloch, Robert (Thesis advisor) / Hahn, Paul (Committee member) / Lan, Shiwei (Committee member) / Zhou, Shuang (Committee member) / Saul, Steven (Committee member) / Arizona State University (Publisher)
Created2020
Description
Geographically Weighted Regression (GWR) has been broadly used in various fields to
model spatially non-stationary relationships. Classic GWR is considered as a single-scale model that is based on one bandwidth parameter which controls the amount of distance-decay in weighting neighboring data around each location. The single bandwidth in GWR assumes that processes (relationships between the response variable and the predictor variables) all operate at the same scale. However, this posits a limitation in modeling potentially multi-scale processes which are more often seen in the real world. For example, the measured ambient temperature of a location is affected by the built environment, regional weather and global warming, all of which operate at different scales. A recent advancement to GWR termed Multiscale GWR (MGWR) removes the single bandwidth assumption and allows the bandwidths for each covariate to vary. This results in each parameter surface being allowed to have a different degree of spatial variation, reflecting variation across covariate-specific processes. In this way, MGWR has the capability to differentiate local, regional and global processes by using varying bandwidths for covariates. Additionally, bandwidths in MGWR become explicit indicators of the scale at various processes operate. The proposed dissertation covers three perspectives centering on MGWR: Computation; Inference; and Application. The first component focuses on addressing computational issues in MGWR to allow MGWR models to be calibrated more efficiently and to be applied on large datasets. The second component aims to statistically differentiate the spatial scales at which different processes operate by quantifying the uncertainty associated with each bandwidth obtained from MGWR. In the third component, an empirical study will be conducted to model the changing relationships between county-level socio-economic factors and voter preferences in the 2008-2016 United States presidential elections using MGWR.
model spatially non-stationary relationships. Classic GWR is considered as a single-scale model that is based on one bandwidth parameter which controls the amount of distance-decay in weighting neighboring data around each location. The single bandwidth in GWR assumes that processes (relationships between the response variable and the predictor variables) all operate at the same scale. However, this posits a limitation in modeling potentially multi-scale processes which are more often seen in the real world. For example, the measured ambient temperature of a location is affected by the built environment, regional weather and global warming, all of which operate at different scales. A recent advancement to GWR termed Multiscale GWR (MGWR) removes the single bandwidth assumption and allows the bandwidths for each covariate to vary. This results in each parameter surface being allowed to have a different degree of spatial variation, reflecting variation across covariate-specific processes. In this way, MGWR has the capability to differentiate local, regional and global processes by using varying bandwidths for covariates. Additionally, bandwidths in MGWR become explicit indicators of the scale at various processes operate. The proposed dissertation covers three perspectives centering on MGWR: Computation; Inference; and Application. The first component focuses on addressing computational issues in MGWR to allow MGWR models to be calibrated more efficiently and to be applied on large datasets. The second component aims to statistically differentiate the spatial scales at which different processes operate by quantifying the uncertainty associated with each bandwidth obtained from MGWR. In the third component, an empirical study will be conducted to model the changing relationships between county-level socio-economic factors and voter preferences in the 2008-2016 United States presidential elections using MGWR.
ContributorsLi, Ziqi (Author) / Fotheringham, A. Stewart (Thesis advisor) / Goodchild, Michael F. (Committee member) / Li, Wenwen (Committee member) / Arizona State University (Publisher)
Created2020
Description
Since the mid-2000s, the domestic aviation industry has been influenced by new, rapidly growing ultra low-cost carriers (ULCCs) such as Allegiant Air, Spirit Airlines, and Frontier Airlines. These carriers augment the existing low-cost airline model by operating largely point-to-point routes with a minimum of passenger amenities. Existing literature, however, is limited for North American ULCCs, often lumping them together with mainstream low-cost carriers. The pattern of markets served by ULCCs is incongruous with the models of other airlines and requires further research to examine causal factors. This paper sought to establish conclusions about ULCCs and the relevant market factors used for airport choice decisions.The relationship between ULCC operations and airport choice factors was analyzed using three methods: a collection of 2019 flight data to establish existing conditions and statistics, two regression analyses to evaluate airport market variables, and three case studies examining distinct scenarios through qualitative interviews with airport managers. ULCC enplanement data was assembled for every domestic airport offering scheduled ULCC service in 2019. Independent variable data informed by previous research were collected for every Part 139 airport in the U.S. The first regression analysis estimated a OLS regression model to analyze the log of enplanements. The second model estimated a binary logistic equation for ULCC service as a 0-1 dependent variable. Case studies for Bellingham, Washington, Waco, Texas, and Lincoln, Nebraska were selected based on compelling airport factors and relevant ULCC experience.
Results of the research methods confirm certain theories regarding ULCC airport choice, but left others unanswered. Maps of enplanements and market share revealed concentrations of ULCC operations on the East Coast. Each regression analysis showed a strong and positive relationship between population figures and the existence and quantity of ULCC operations. Tourism employment was only significantly related to enplanements. Other factors including distance and competition variables were significantly associated to ULCC service. Case studies revealed the importance of airport fees and costs in ULCC decision-making; factors that proved difficult to investigate quantitatively in this research. Further research may shed light on this complex and ever- changing subset of the domestic commercial aviation industry.
ContributorsTaplin, Drew (Author) / Kuby, Michael (Thesis advisor) / King, David (Committee member) / Salon, Deborah (Committee member) / Arizona State University (Publisher)
Created2021