Authors: Meiqing Li*, University of Pennsylvania
Topics: Remote Sensing, Spatial Analysis & Modeling, Urban Geography
Keywords: Google Earth Engine, urban intensity, remote sensing
Session Type: Poster
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded
Remote sensing imagery can be a good resource to explore the unrevealed social and economic conditions of cities and regions where such data are scarce or with limited access. Both daytime and nighttime remote sensing imagery can shade light on spatial patterns of human activities and settlements. While nighttime light (NTL) imagery is a good proxy of urban intensity, daytime land cover data well indicate the built-up areas. Combining these images of different spatial resolutions can enhance their predictive power in terms of statistical significance and scale. This can be especially useful for studying urban growth patterns of cities in developing countries that are experiencing rapid urban changes, while at the same time lack effective monitoring tools to ensure sustainable growth. Several researches have validated the relationship between NTL and urban activities applying variations of DMSP/OLS, Landsat and MODIS images as well as in different contexts. This paper develops a generalizable tool/method using Google Earth Engine that can be easily applied to different contexts and scales for estimation purpose. With California as the study area, it outputs annual estimates of county-level urban intensity between 2000 and 2013, and validates them with socio-economic factors indicated by census and spatial data. The result shows: there is a steadying growing trend of average urban intensity for each county; counties with large metropolitan areas for example Los Angeles and San Francisco exhibit significantly higher urban intensity than other cities.