Satoshi Miyazawa

This study develops a globally adaptable and scalable methodology for high-resolution, building-level population mapping, integrating Earth observation techniques, geospatial data acquisition, and machine learning to enhance population estimation in rapidly urbanizing cities, particularly in developing countries. Using Bangkok, Thailand, as a case study, this research presents a problem-driven approach that leverages open geospatial data, including Overture Maps and OpenStreetMap (OSM), alongside Digital Elevation Models, to overcome limitations in data availability, granularity, and quality. This study integrates morphological terrain analysis and machine learning-based classification models to estimate building ancillary attributes such as footprint, height, and usage, applying micro-dasymetric mapping techniques to refine population distribution estimates. The findings reveal a notable degree of accuracy within residential zones, whereas performance in commercial and cultural areas indicates room for improvement. Challenges identified in mixed-use and townhouse building types are attributed to issues of misclassification and constraints in input data. The research underscores the importance of geospatial AI and remote sensing in resolving urban data scarcity challenges. By addressing critical gaps in geospatial data acquisition and processing, this study provides scalable, cost-effective solutions in the integration of multi-source remote sensing data and machine learning that contribute to sustainable urban development, disaster resilience, and resource planning. The findings reinforce the transformative role of open-access geospatial data in Earth observation applications, supporting real-time decision-making and enhanced urban resilience strategies in rapidly evolving environments.

Urban planning and management increasingly depend on accurate building and population data. However, many regions lack sufficient resources to acquire and maintain these data, creating challenges in data availability. Our methodology integrates multiple data sources, including aerial imagery, Points of Interest (POIs), and digital elevation models, employing Light Gradient Boosting Machine (LightGBM) and Gradient Boosting Decision Tree (GBDT) to classify building uses and morphological filtration to estimate heights. This research contributes to bridging the gap between data needs and availability in resource-constrained urban environments, offering a scalable solution for global application in urban planning and population mapping.

Monitoring the crowd in urban hot spot has been an important research topic in the field of urban management and has high social impact. It can allow more flexible allocation of public resources such as public transportation schedule adjustment and arrangement of police force. After 2020, because of the epidemic of COVID-19 virus, the public mobility pattern is deeply affected by the situation of epidemic as the physical close contact is the dominant way of infection. In this study, we propose a confirmed case-driven time-series prediction of crowd in urban hot spot named MobCovid. The model is a deviation of Informer, a popular time-serial prediction model proposed in 2021. The model takes both the number of nighttime staying people in downtown and confirmed cases of COVID-19 as input and predicts both the targets. In the current period of COVID, many areas and countries have relaxed the lockdown measures on public mobility. The outdoor travel of public is based on individual decision. Report of large amount of confirmed cases would restrict the public visitation of crowded downtown. But, still, government would publish some policies to try to intervene in the public mobility and control the spread of virus. For example, in Japan, there are no compulsory measures to force people to stay at home, but measures to persuade people to stay away from downtown area. Therefore, we also merge the encoding of policies on measures of mobility restriction made by government in the model to improve the precision. We use historical data of nighttime staying people in crowded downtown and confirmed cases of Tokyo and Osaka area as study case. Multiple times of comparison with other baselines including the original Informer model prove the effectiveness of our proposed method. We believe our work can make contribution to the current knowledge on forecasting the number of crowd in urban downtown during the Covid epidemic.

The Japanese government adopted policies to control human mobility in 2020 to prevent the spread of severe acute respiratory syndrome coronavirus 2, which causes coronavirus disease 2019 (COVID-19). The present study examined the impact of human mobility on COVID-19 cases at the prefectural level in Japan by devising an indicator to have a relationship between the number of infected people and on human mobility. We calculated origin–destination travel mobility within prefectures in Japan from March 1st to December 31st, 2020, using mobile phone data. A cross-correlation function (CCF) was used to examine the relationship between human mobility and a COVID-19 infection acceleration indicator (IAI), which represents the rate of change in the speed of COVID-19 infection. The CCF of intraprefectural human mobility and the IAI in Tokyo showed a maximum value of 0.440 at lag day 12, and the IAI could be used as an indicator to predict COVID-19 cases. Therefore, the IAI and human mobility during the COVID-19 pandemic were useful for predicting infection status. The number of COVID-19 cases was associated with human mobility at the prefectural level in Japan in 2020. Controlling human mobility could help control infectious diseases in a pandemic, especially prior to starting vaccination.

Detailed demographics play an important role in the development of smart cities. However, especially in developing countries, the maintenance and management of this data is incomplete, which hinders the promotion of smart cities. The objective of this study is to develop a method to create detailed building distribution maps from satellite images, which will serve as a basis for developing detailed demographic data to support the promotion of smart cities around the world. The target area is several areas of Tokyo where validation data is available. We first developed a method for extracting buildings from satellite images and then estimating the building use to determine the buildings where residents are distributed. Both methods use deep learning. As a result, it was possible to extract buildings with an extraction rate (the number of buildings in the automatically extracted building data divided by the number of buildings in the data for verification) of up to 60.3% for the entire target area. In addition, in the estimation of building use, our method was able to classify detached and non-detached buildings with an average accuracy of 78.7% for the entire target area.

Background: During the coronavirus disease 2019 (COVID-19) pandemic in Japan, the state of emergency, as a public health measure to control the spread of COVID-19, and the Go To campaign, which included the Go To Travel and Go To Eat campaigns and was purposed to stimulate economic activities, were implemented. This study investigated the impact of these government policies on COVID-19 spread.Methods: This ecological study included all 47 prefectures in Japan as samples between February 3 and December 27, 2020. We used COVID-19 cases and mobility as variables. Additionally, places where social contacts could accrue, defined as restaurants, companies, transportation, and tourist spots; mean temperature and humidity; the number of inhabitants in their twenties to fifties; and the number of COVID-19 cases in the previous period, which were factors or covariates in the graphical modeling analysis, were divided into five periods according to the timing of the implementation of the state of emergency and Go To campaign.Results: Graphical changes occurred throughout all five periods of COVID-19. During the state of emergency (period 2), a correlation between COVID-19 cases and those before the state of emergency (period 1) was observed, although this correlation was not significant in the period after the state of emergency was lifted (period 3). During the implementation of Go To Travel and the Go To Eat campaigns (period 5), the number of places where social contacts could accrue was correlated with COVID-19 cases, with complex associations and mobility.Conclusions: This study confirms that the state of emergency affected the control of COVID-19 spread and that the Go To campaign led to increased COVID-19 cases due to increased mobility by changing behavior in the social environment where social contacts potentially accrue.

Background: During the second wave of COVID-19 in August 2020, the Tokyo Metropolitan Government implemented public health and social measures to reduce on-site dining. Assessing the associations between human behavior, infection, and social measures is essential to understand achievable reductions in cases and identify the factors driving changes in social dynamics. Objective: The aim of this study was to investigate the association between nighttime population volumes, the COVID-19 epidemic, and the implementation of public health and social measures in Tokyo. Methods: We used mobile phone location data to estimate populations between 10 PM and midnight in seven Tokyo metropolitan areas. Mobile phone trajectories were used to distinguish and extract on-site dining from stay-at-work and stay-at-home behaviors. Numbers of new cases and symptom onsets were obtained. Weekly mobility and infection data from March 1 to November 14, 2020, were analyzed using a vector autoregression model. Results: An increase in the number of symptom onsets was observed 1 week after the nighttime population volume increased (coefficient=0.60, 95% CI 0.28 to 0.92). The effective reproduction number significantly increased 3 weeks after the nighttime population volume increased (coefficient=1.30, 95% CI 0.72 to 1.89). The nighttime population volume increased significantly following reports of decreasing numbers of confirmed cases (coefficient=–0.44, 95% CI –0.73 to –0.15). Implementation of social measures to restaurants and bars was not significantly associated with nighttime population volume (coefficient=0.004, 95% CI –0.07 to 0.08). Conclusions: The nighttime population started to increase after decreasing incidence of COVID-19 was announced. Considering time lags between infection and behavior changes, social measures should be planned in advance of the surge of an epidemic, sufficiently informed by mobility data.

BACKGROUND During the second wave of COVID-19 in August 2020, the Tokyo Metropolitan Government implemented public health and social measures to reduce on-site dining. Assessing the associations between human behavior, infection, and social measures is essential to understand achievable reductions in cases and identify the factors driving changes in social dynamics. OBJECTIVE The aim of this study was to investigate the association between nighttime population volumes, the COVID-19 epidemic, and the implementation of public health and social measures in Tokyo. METHODS We used mobile phone location data to estimate populations between 10 PM and midnight in seven Tokyo metropolitan areas. Mobile phone trajectories were used to distinguish and extract on-site dining from stay-at-work and stay-at-home behaviors. Numbers of new cases and symptom onsets were obtained. Weekly mobility and infection data from March 1 to November 14, 2020, were analyzed using a vector autoregression model. RESULTS An increase in the number of symptom onsets was observed 1 week after the nighttime population volume increased (coefficient=0.60, 95% CI 0.28 to 0.92). The effective reproduction number significantly increased 3 weeks after the nighttime population volume increased (coefficient=1.30, 95% CI 0.72 to 1.89). The nighttime population volume increased significantly following reports of decreasing numbers of confirmed cases (coefficient=–0.44, 95% CI –0.73 to –0.15). Implementation of social measures to restaurants and bars was not significantly associated with nighttime population volume (coefficient=0.004, 95% CI –0.07 to 0.08). CONCLUSIONS The nighttime population started to increase after decreasing incidence of COVID-19 was announced. Considering time lags between infection and behavior changes, social measures should be planned in advance of the surge of an epidemic, sufficiently informed by mobility data.