Geospatial Data Analytics and Urban Applications

Big data refers to the large volumes of data collected by several sensors to monitor natural or anthropogenic activities. These data are temporal in nature and come in various forms, volume, and frequencies. Most of these have a geographical connotation and hence are essentially considered as spatial big data. Treating big data statistically can derive several insights. But spatial analytics add another dimension to the insights which make it critical for decision making. In this chapter, the fundamentals of spatial big data and analytics are discussed from a data science perspective. It also discusses the urban applications of spatial big data analytics.

Accessibility to amenities is an essential consideration for renting or purchasing a home. Therefore, ease of access and proximity to amenities is something Singapore’s Housing and Development Board (HDB) emphasises on for the planning and development of housing towns in Singapore. Most amenities can be categorised as transport amenities, healthcare amenities or education amenities. Improving the availability and access to such amenities is an effective way to meet the housing needs of Singaporeans and HDB tries to ensure that the needs of Singaporeans are taken care of. This study presents a geospatial evaluation of access to various amenities from HDB blocks and housing towns. It features a proximity analysis (walking time and distance) to and from each amenity to HDB blocks. Factors like personal preference of individuals are considered in querying a suitable located HDB to assist home search using the Closest Facility Analysis tool in ArcGIS. Weights are assigned to distance to amenities to personalise home search and this provides a potential buyer with an additional tool to spatially query on a list of homes to choose from. The study shall provide potential buyers with a list of suitable options using which a multiple criteria search in terms of nearby amenities can be used to look for a suitable home. The study also aims to identify housing towns in Singapore which lack ease of accessibility to amenities in terms of time and distance and this can be used for planning of setting up amenities near these HDB blocks to make them more saleable. This study sheds light on housing towns which require improvements in the provision of amenities for residents.

The outbreak of novel coronavirus pneumonia was the most serious global issue in 2020, that caused enormous impacts on various aspects of human society from public health to economic well-being. Our ecological environment also experienced transformation due to restricted human activities during the epidemic. The implementation of ‘Lockdowns’ and ‘Stay-at-Home’ policies reduced the pollution emissions from transport and commercial activities altering the urban environment. In this study, the spatial and temporal distribution and changes of the air quality and thermal environment in Wuhan City and New York City in 2020 were conducted and analyzed. Spatial data from these two cities were acquired, processed, interpolated and analyzed to identify hot spots and cold spots which were reasoned. The dynamics of air pollution and Urban Heat Island (UHI) effect was the prime object of investigation. The study discovered interesting patterns of changes in those two cities which we the early epicenters of the pandemic.

Mass Rapid Transit (MRT) is the backbone of Singapore’s transport system. It ferries its population from their homes to work and back. On weekends, people use the MRT to visit places of leisure. This travel ritual of Singapore residents was disrupted during the onset of Covid-19 where strict measures were enforced to contain the infection chain. These measures, called circuit breakers, were enforced and later relaxed to allow people to carry out some activities which involved MRT travel. From 2nd of June 2020 the country entered ‘Phase One’ of safe re-opening during which certain economic activities, which do not pose high risk of transmission, were allowed and the remaining ones, with higher risk of disease transmission remain closed. Ridership of MRTs partly recovered during this period. ‘Phase Two’ of re-opening followed from 19th June which allowed selective recreational activities with safe distancing measures still in place. The current study was aimed to understand the Spatio-temporal patterns of MRT ridership during Phase One and Two. The study reflects on the MRT stations which were popular and brings forth insights which may help city planning authorities for the future.

Since 19 July 2020, Singapore entered Phase 2 of re-opening after one and half month’s “Circuit Breaker” measures to curb the spread of COVID 19. Although most businesses and public places have resumed operation at a reduced capacity, individuals were strongly advised to practice social distancing and avoid crowds. Both implicit and explicit measures to prevent overcrowding had impacted on how people visit places in Singapore. The current study used geotagged Twitter data between September to October in 2020 to examine the spatial and temporal patterns of residents’ locations in Singapore and explored the service amenities which remain “attractive” to residents. Random Forest Supervised Machine Learning Model was used to train and predict spatial distribution of activities during off-work recreational hours using service amenities point of interests (POIs) and land use merge. Five explanatory variables used were parks, public links between parks and malls, taxi stands, residential areas, and shopping malls which had the strongest influence in driving the model prediction of spatial distribution of activities in off-work recreational hours. While distinct temporal patterns of tweets were expected during office hour, this analysis revealed no such statistically significant clusters. The regression analysis showed that distances to service amenities did not provide strong explanations for tweeting patterns.

Gun violence in the USA has caused immense socio-economic implications which has plagued urban areas. In the first half of 2020, shooting incidents have been on the rise in New York City (NYC) and a sharp spike of cases was observed in June 2020. In the current study, a space–time analysis on such incidents was done to investigate the clusters and trends with a view to identify NTAs with high shooting densities. Regression analysis on available data for NYC found that such incidents have a close relationship with black population and with locations with vacant housing units. Thus, these factors could play a key role in the predictive analysis of future incidents. A network analysis was conducted which found that shooting incidents generally increase with distance from police stations and that not all of NYC is serviced by police stations within a 15-min drive time. Such findings could potentially help the authorities of NYC to improve their policing, resource allocation and decision-making to address gun violence in the city.

Emissiveness reflects travelers’ overall ability to travel from a source to a destination. Shanghai, as a popular tourist city in China, was always a topic for tourism-related analysis. In this study, the emissiveness of 31 provinces in Mainland China as Shanghai's travel source markets were analyzed. At the same time, Travel Gravity Model (Cesario in Econ Geogr 52:363–373, 1976, [1]) was used to incorporate destination attractiveness to construct an advanced emissiveness model. The study found that the source market's propensity to travel to Shanghai showed a decreasing trend from east to west in Mainland China. Besides, a similar popular tourist city, Guangzhou, was chosen to compare the differences in travel propensity between the source markets to the two cities. In terms of time, the travel peaked in the two cities in different months. Spatially, we drew preliminary conclusions that spatial distance and economic development had more important influences on emissiveness.

Data from Chicago city in Illinois State of the U.S. was extracted to estimate the effects of the onset of the COVID-19 pandemic on crime. There was a general drop in reported crimes, which appeared to precede the stay-at-home orders, and then there was a sudden increase in non-residential car theft after the stay-at-home order issued in March. This change suggested that criminal activity was “substituted” as most people would be staying at home with minimal surveillance of the vehicles which were parked elsewhere. On the contrary, there was an immediate increase in domestic crime from January onwards. This study aims to figure out the impact of the pandemic on changing trends in crime by running various GIS methods. As a result, monthly change of spatial pattern, change in domestic crime and motor theft, and the relationship between some socioeconomic factors and crimes were analyzed. Evidence showed that there was a need to establish an association between crime rates with social and health data to ensure adequate investment in social safety net and programs to strengthen social resilience given that the pandemic was likely to continue for months if not years.

It has been observed that more and more young people choose to use online car-hailing as the preferred travel mode. Hence a study of the trip patterns of online car-hailing can reveal a lot on the travel needs and patterns of the young generation. This study explores the characteristics of Grab trips, which is the major car hailing service provide in Singapore, order points in Singapore, by modelling them geospatially using their GPS tracks and points of interests in Singapore. The spatial analysis evaluates various regression models, namely the Ordinary Least Squares (OLS), Geographically Weighted Regression (GWR) and Multiscale Geographically Weighted Regression (MGWR). The MGWR results were found to be far superior to OLS and GWR. It was found that points of interest lie retail shopping outlets, clusters of young population, taxi pick points and public transport facilities impact the trip start and end points of car hailing services. Grab ride end points are primarily located at interests such as the city centre, airport areas, retail outlets and industrial areas which displays a spatial heterogeneity.

The Alpine ecological region of Qinghai Tibet Plateau is undergoing significant changes endangering its vulnerability to ecology. Based on Pressure-State-Response (PSR) model, six indexes were are analyzed to evaluate the ecological vulnerability of Nyingchi city in Tibet. The Spatio-temporal ecological vulnerability of Nyingchi city from 2000 to 2015 was quantified based on spatial modeling techniques and observed changes were discussed in the context of space and time. The Ecological Vulnerability Standardized Index (EVSI), which was obtained by accounting different vulnerability levels yielded that the overall vulnerability of Nyingchi is relatively high in the north and gradually decreasing towards the south. The changes observed correlated well with local topography, climate, disaster volume and water conservancy construction. As a result, five levels of ecologically fragility could be defined for Nyingchi city, which shall provide a direction to efforts for ecological restoration and re-construction. From the perspective of ecological restoration and reconstruction planning of restoration and reconstruction can be addressed based on the severity of the vulnerability. Such studies on other vulnerable cities at a closer time periods can establish a continuous monitoring systems to detect changes in vulnerability. Automated systems with continuous feed of satellite data with dynamic Realtime processing can help measure and feed information to information dashboards which monitor ecological health of a region. These systems can ensure sustainability of our ecology.