Seeing Informal Settlements : the Policy Implications of Different Techniques to Identify Urban Growth Patterns From Satellite Imagery Using the Case of Informal Construction in Ho Chi Minh City

In the midst of urban policy institutions taking greater advantage of technological advances for generating data, this paper analyzes the current state of using satellite imagery to monitor informal settlement patterns. Using the issue of informal settlements in rapid urbanization, this paper finds that policy-oriented research agendas needs further development as much as the techniques. The paper first presents an example of how informal urbanization can be detected even with older imagery combined with fieldwork by applying thresholding and texture analysis of remote sensing imagery developed by Kim et al (2004) for data from 1994 and 2001 for Ho Chi Min City, Vietnam. Then this study compares our identification of urbanized areas in Ho Chi Minh City with those identified for 2000 by Angel et al. (2005) and the World Bank (2015). We find that while all three converge on classifying urban areas in the city’s core, in the periphery each study’s methods systematically identify different kinds of urban spatial patterns. These differences suggest the importance of customizing algorithms to account for the local context and testing through ground-truthing to establish their accuracy. These observations also point to a need for a discussion between urbanization scholars about developing standards in reporting data, methods, and findings. Our study suggests that analysts need to keep a critical eye for marginalized populations when using automated interpretation of satellite images to establish urban settlement patterns. Contextual knowledge gained through fieldwork and collaborative partnerships is necessary to not systematically overlook informal settlements which could have important implications for disaster and resilience, resettlement, climate change adaptation, and public finance policies and planning