Urban Landsat. Cities from Space

Mapping and Monitoring Urban Form and Growth


Seattle-Tacoma, via  SEDAC (Socioeconomic Data and Applications Center)





Objectives of Urban Remote Sensing

Remote Sensing generally refers to the science, technology and physical processes involved in the detection, analysis and interpretation of information collected without coming into physical contact with the object of interest. In the context of our research and applications, urban remote sensing focuses primarily on understanding the physical properties and processes of urban environments and on the mapping and monitoring of urban land cover and spatial extent. These two objectives are related in the sense that it is necessary to understand the physical properties of the urban mosaic in order to rigorously define, map and monitor urban areas.



Characterizing the Physical Properties of the Urban Environment

The research presented here is primarily focused on the physical properties of a wide range of urban environments using passive measurement of optical reflectance and thermal emission as well as optical emission of nighttime lights. Comparative analyses of urban reflectance (visible and infrared color) and surface temperature allow us to develop robust criteria for distinguishing urban land cover from non-anthropogenic land covers. These analyses also provide important constraints on the physical properties that control mass and energy fluxes through the urban environment. These constraints are used as inputs to physical models of climatic, hydrologic and ecologic processes.

Mapping and Monitoring Urban Form and Growth

Characterizing the physical properties of urban land cover makes it possible to map the form and spatial extent of urban land use and to quantify changes in form and extent. This provides objective, physically-based metrics for comparative analyses of urban dynamics that cannot generally be obtained from administrative definitions of urban extent. Mapping provides static snapshots of the urban mosaic while monitoring allows us to quantify the spatiotemporal dynamics. Mapping urban extent with nightlights complements the information derived from optical reflectance.

Each city is linked to a jpeg image of a Landsat visible/IR composite collected by Landsat 5 (pre 1999) or Landsat 7. Most of the image areas are 30x30 km but a few are larger. All images are shown at full resolution (30 m pixel) so the scale (on screen) is equivalent. The jpeg compression causes significant loss of fine detail from the original image. Most of the images are about 200K (~40 seconds via 56K modem). A few are significantly larger.

http://sedac.ciesin.columbia.edu/ulandsat/