This course is being offered in collaboration between the Odum Institute and the Center for Urban & Regional Studies.

Urbanization has been a fundamental trend of the past two centuries and a key force shaping the development of the modern world. While urbanization in rapidly growing nations is helping lift hundreds of millions of people out of poverty, it is also creating immense societal challenges by increasing greenhouse-gas emissions, destabilizing fragile ecosystems and creating new demands on public services and infrastructure that pose significant challenges on the environment. Despite the importance of understanding the drivers of urban growth, we are still unable to quantify the magnitude and pace of urbanization in a consistent manner at a high resolution and global scale.

The revolution in geospatial data has transformed how we study cities. Since the 1970s, terrestrial Earth observation data have been continuously collected in various spectral, spatial and temporal resolutions. As improved satellite imagery becomes available, new remote-sensing methods and machine-learning approaches have been developed to convert terrestrial Earth-observation data into meaningful information about the nature and pace of change of urban landscapes and human settlements. But until recently, most remote sensing studies focused on local settings. Mapping land cover at a national or regional scale is challenging because of the lack of high-resolution global imagery, the heterogeneous and complex spectral characteristics of land, the small and fragmented spatial configuration of many cities, and importantly, computational constrains (for storage and processing). Emerging cloud-based computational platforms now allow for scaling analysis across space and time. Google Earth Engine (GEE) is one platform that leverages cloud-computing services to achieve planetary-scale utility. GEE leverages cloud-computational services for planetary-scale analysis and consists of petabytes of geospatial and tabular data, including a full archive of Landsat, Sentinel-2, Sentinel-1, and MODIS, together with a JavaScript, Python based API (GEE API), and algorithms for supervised image classification.

This hands-on course will focus on the use of Google Earth Engine (GEE) for urban research applications. It will demonstrate how free and open-source satellite imagery – including electro-optical (EO) and Synthetic Aperture Radar (SAR) imagery – can be utilized to map urban areas and urbanization trends and patterns, across space and time, and to perform a qualitative analysis of the impacts of urban expansion on the landscape. In addition to analyzing existing classification schemes of urban areas to understand how cities expand and evolve, the course will provide a brief introduction to concepts in Remote Sensing Machine Learning, with a focus on supervised pixel-based image classification. Students will learn how to automatically map built-up land cover based on publicly available satellite imagery (e.g., Landsat and Sentinel). In addition, the course will demonstrate recent innovations in the use of remotely sensed nighttime light observations to understand variations in economic activity within and between cities – all utilizing data and tools that are available in GEE. The course will include PowerPoint slides, group hands-on coding (in JavaScript) and short exercises. Prior coding knowledge is not required.