Temporal characteristics of thermal satellite images for urban climate study

Urban climate study requires the understandings of both thermal patterns (space) and processes (time). However thermal satellite images can only capture the surface temperature at an instant in time. As such the images do not necessarily imply they are representative for a period of time. This paper investigates the effectiveness of using ASTER data to model the changes of the thermal environment over time by correlating time series auto weather station data with image-derived air temperatures. Results suggest the daytime and nighttime thermal images to be valid for 4 and 11 hours, respectively. These findings suggest that thermal images can represent the temperature patterns of an area over a period, and are therefore potentially useful for urban climate research at times other than the period immediately surrounding the acquisition time.