Guofeng Cao

Explore spatiotemporal and demographic characteristics of human mobility via Twitter: A case study of Chicago

Abstract Characterizing human mobility patterns is essential for understanding human behaviors and the interactions with socioeconomic and natural environment, and plays a critical role in public health, urban planning, transportation engineering and related fields. With the widespread of location-aware mobile devices and continuing advancement of Web 2.0 technologies, location-based social media (LBSM) have been gaining widespread popularity in the past few years. With an access to locations of hundreds of million users, profiles and the contents of the social media posts, the LBSM data provided a novel modality of data source for human mobility study. By exploiting the explicit location footprints and mining the latent demographic information implied in the LBSM data, the purpose of this paper is to investigate the spatiotemporal characteristics of human mobility with a particular focus on the impact of demography. To serve this purpose, we first collect geo-tagged Twitter feeds posted in the conterminous United States area, and organize the collection of feeds using the concept of space-time trajectory corresponding to each Twitter user. Commonly human mobility measures, including detected home and activity centers, are derived for each user trajectory. We then select a subset of Twitter users that have detected home locations in the city of Chicago as a case study, and apply name analysis to the names provided in user profiles to learn the implicit demographic information of Twitter users, including race/ethnicity, gender and age. Finally we explore the spatiotemporal distribution and mobility characteristics of Chicago Twitter users, and investigate the demographic impact by comparing the differences across three demographic dimensions (race/ethnicity, gender and age). We found that, although the human mobility measures of different demographic groups generally follow the generic laws (e.g., power law distribution), the demographic information, particular the race/ethnicity group, significantly affects the urban human mobility patterns.

These lit areas are undeveloped: Delimiting China’s urban extents from thresholded nighttime light imagery

Abstract Nighttime light imagery is a powerful tool to study urbanization because it can provide a uniform metric, lit area, to delimit urban extents. However, lit area is much larger than actual urban area, so thresholds of digital number (DN) values are usually needed to reduce the lit area. The threshold varies greatly among different regions, but at present it is still not very clear what factors impact the changes of the threshold. In this study, urban extent by province for China is mapped using official statistical data and four intercalibrated and geometrically corrected nighttime light images between 2004 and 2010. Lit area in the imagery for most provinces is at least 94% greater than the official amount of urban area. Regression analyses show a significant correlation between optimal thresholds and GDP per capita, and larger thresholds more commonly indicate higher economic level. Size and environmental condition may explain a province’s threshold that is disproportionate to GDP. Findings indicate one threshold DN is not appropriate for multiple (adjacent) province urban extent mapping, and optimal thresholds for one year may be notably different than the next. Province-level derived thresholds are not appropriate for other geographic levels. Brightness of nighttime lights is an advantage over imagery that relies on daylight reflection, and decreases in brightness indicate faster growth in the horizontal direction than the vertical. A province’s optimal threshold does not always maintain an increase with population and economic growth. In the economically developed eastern provinces, urban population densities decreased (and this is seen in the brightness data), while urban population increased.

Improving accuracy of economic estimations with VIIRS DNB image products

ABSTRACT A new-generation of night-time lights (NTL) image products, the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) monthly composites, have been produced and released by the National Oceanic and Atmospheric Administration’s National Centers for Environmental Information. Compared with the last generation NTL image products, the Defense Meteorological Satellite Program’s Operational Linescan System stable light composites, the new NTL image products have finer spatial resolution with compatible radiance values across different month/year images. However, the current defects in VIIRS DNB monthly composites show ephemeral lights, relatively high radiance in winter months, and missing data over the high-latitude regions of the northern hemisphere in summer months. This study presents a method to improve the accuracy of the new NTL image products by statistically modelling the time series VIIRS NTL images and uses the improved imagery to estimate socio-economic factors. In this method, we first estimate radiance for each pixel with ‘no data’ in May, June, July, and August images and then exponentially smooth the monthly time series images to produce a 2014 annual VIIRS DNB image for the contiguous USA. Sum radiance derived from the smoothed annual image shows stronger correlations with gross domestic product at the state level and smaller standard errors of the estimate at the metropolitan and county levels compared with that extracted from the annual image produced by simply averaging the original monthly DNB composites. Such results infer that exponential smoothing effectively improves the quality of the VIIRS DNB images for annual economic estimation.

Forecasting China’s GDP at the pixel level using nighttime lights time series and population images

China’s rapid economic development greatly affected not only the global economy but also the entire environment of the Earth. Forecasting China’s economic growth has become a popular and essential issue but at present, such forecasts are nearly all conducted at the national scale. In this study, we use nighttime light images and the gridded Landscan population dataset to disaggregate gross domestic product (GDP) reported at the province scale on a per pixel level for 2000–2013. Using the disaggregated GDP time series data and the statistical tool of Holt–Winters smoothing, we predict changes of GDP at each 1 km × 1 km grid area from 2014 to 2020 and then aggregate the pixel-level GDP to forecast economic growth in 23 major urban agglomerations of China. We elaborate and demonstrate that lit population (brightness of nighttime lights × population) is a better indicator than brightness of nighttime lights to estimate and disaggregate GDP. We also show that our forecast GDP has high agreement with the National Bureau of Statistics of China’s demographic data and the International Monetary Fund’s predictions. Finally, we display uncertainties and analyze potential errors of this disaggregation and forecast method.

Improve ground-level PM2.5 concentration mapping using a random forests-based geostatistical approach

Accurate measurements of ground-level PM2.5 (particulate matter with aerodynamic diameters equal to or less than 2.5 μm) concentrations are critically important to human and environmental health studies. In this regard, satellite-derived gridded PM2.5 datasets, particularly those datasets derived from chemical transport models (CTM), have demonstrated unique attractiveness in terms of their geographic and temporal coverage. The CTM-based approaches, however, often yield results with a coarse spatial resolution (typically at 0.1° of spatial resolution) and tend to ignore or simplify the impact of geographic and socioeconomic factors on PM2.5 concentrations. In this study, with a focus on the long-term PM2.5 distribution in the contiguous United States, we adopt a random forests-based geostatistical (regression kriging) approach to improve one of the most commonly used satellite-derived, gridded PM2.5 datasets with a refined spatial resolution (0.01°) and enhanced accuracy. By combining the random forests machine learning method and the kriging family of methods, the geostatistical approach effectively integrates ground-based PM2.5 measurements and related geographic variables while accounting for the non-linear interactions and the complex spatial dependence. The accuracy and advantages of the proposed approach are demonstrated by comparing the results with existing PM2.5 datasets. This manuscript also highlights the effectiveness of the geographical variables in long-term PM2.5 mapping, including brightness of nighttime lights, normalized difference vegetation index and elevation, and discusses the contribution of each of these variables to the spatial distribution of PM2.5 concentrations.

Short communication: emerging technologies for biometeorology

The first decade of the twenty-first century saw remarkable technological advancements for use in biometeorology. These emerging technologies have allowed for the collection of new data and have further emphasized the need for specific and/or changing systems for efficient data management, data processing, and advanced representations of new data through digital information management systems. This short communication provides an overview of new hardware and software technologies that support biometeorologists in representing and understanding the influence of atmospheric processes on living organisms.

An interactive approach for deriving geometric network models in 3D indoor environments

Humans spend most of their life in indoor spaces. As indoor spaces are becoming increasingly complex, there are compelling needs for efficient indoor GIS and navigation systems. For indoor navigations, numerous geometric network models have been proposed as navigable spatial models for 3D indoor environments in the past decade. Most of the existing discussions, however, tend to focus on conceptual representations of geometric networks; not enough attention has been given on the generation processes of navigable networks for 3D indoor environments. It is actually nontrivial, considering accurate and complete floor plans, the conventional data sources for building indoor geometric networks, are oftentimes not available for various reasons (e.g., copyright, public safety concerns). With the continue advances of 3D imaging and scanning technologies, 3D data models with fine geometric structures and high quality textures are increasingly available for indoor spaces, thus provide a novel data source for building indoor geometric networks. In this paper, an interactive approach is presented to derive 3D, navigable, geometric network models from these 3D data models. Specifically, this approach includes three steps: decomposing 3D building models in terms of floors, interactively creating geometric network elements (e.g., nodes and edges) and then automatically generating geometric network models. The presented approach is implemented and its advantages are demonstrated with a real world 3D building data.

Quantifying and visualizing language diversity of Hong Kong using Twitter

The wide penetration of location-aware mobile devices and location-based services renders the location-based social media as a reliable proxy to study the real-world geographic space. Language diversity is an important indicator of a citys internationalization level. People communicate using different languages in the cyberspace of social media as they do in the geographic space. The location-based social media therefore provides an innovative set of lens to map the language diversity and study the internationalization of cities. In the enclosed graphics, based on a collection of geo-tagged Twitter posts, we generated a fine resolution map of language diversity index in the area of Hong Kong to illustrate the potential of location-based social media in city research.

Visualizing changes in nationally averaged PM2.5 concentrations by an alluvial diagram

Ailments related to ambient fine particulate matter (PM2.5) cause 3.22 million deaths per year on average worldwide, which, along with high blood pressure and tobacco smoking, is defined as the leading risk factor for global burden of disease (Lim et al., 2013). The alluvial diagram displayed in Figure 1 demonstrates the temporal changes in the ranks of nationally averaged PM2.5 concentrations of developed areas from 1998 to 2013 (Rosvall and Bergstrom, 2010). A ‘developed area’ within a country was delineated using the Defense Meteorological Satellite Program’s Operational Linescan System annual stable lights image composites with pixels’ digital number values equal to or larger than 10 (Zhao and Samson, 2012). A total of 46 countries with at least 10 identified developed areas in the 1998 stable lights image product are included in the alluvial diagram (Doll et al., 2000). Here each wide line in the alluvial diagram is called an alluvium. PM2.5 concentrations were extracted from a satellite-derived gridded PM2.5 datasets (van Donkelaar et al., 2016). This dataset is considered to be one of the most accurate global PM2.5 concentration image products at the 0.01 0.01 (or 1 km 1 km) spatial resolution and has been used in applied public health studies (van Donkelaar et al., 2015). Figure 1 displays differences in the width of each alluvium, which represents a country’s yearly variations in PM2.5 concentrations. The scaled width of the alluvium corresponding to the largest national average PM2.5 concentration (63mg/m ) of the 46 countries, which is found in Iraq, is marked in the diagram. In the vertical direction, the countries are ranked upward by their average PM2.5 concentrations (from lowest to highest) for each year. The left-hand axis is ordered from the lowest to highest concentrations in 1998, while the righthand axis is for that of 2013. Thus, a country with a drastic change in PM2.5 concentration between the time period, such as Côte d’Ivoire, will change position substantially.