Thursday, September 19, 2013

Using GIS to Analyze Micro-Urban Heat Islands

The presence and potential mitigation of Micro-Urban Heat Islands (MUHIs), isolated urban areas within a city that produce “hot spots” for higher than normal temperatures as compared to rural areas or areas of vegetation cover, are an important issue for urban planners.  As early as 1995, researchers have attempted to use GIS technology to combine temperature readings with remote imagery to better understand the potential impact of such MUHIs so that such knowledge could be utilized in future plans.  Cathy Aniello and her research collaborators performed such a study on an area of Dallas, Texas that contains both new and old neighborhoods as well as tree and vegetation cover and a cleared area around a lake.

The researchers obtained an image of the area and then used an algorithm to pixelate it based on clusters for the different land uses, including tree cover, roads, the lake, and other urban land use.  Another algorithm was then used to determine the temperature at each pixel using LANDSAT TM thermal detectors.  Using GIS analysis programs they were then able to combine the two into a map of the area that illustrated thermal patterns across the area in a color-coded fashion. 

This allowed the researchers to determine that significantly higher temperatures in MUHIs radiated outward whereas tree cover correlated with lower temperatures, and also worked in a radial fashion cooling surrounding areas.  This study demonstrates the increased analysis ability that GIS systems provide and the importance of their inclusion in aspects of urban planning.
Aniello, C., Morgan, K., Busbey, A., & Newland, L. (1995).  Mapping micro-urban heat islands using LANDSAT TM and a GIS.  Computers & Geosciences, 21(8), 965-969.


  1. It is interesting that tree cover reduced heat island effect , shows that tree planting policy can make a difference in mitigating heat islands. Urban forests can reduce temperatures but also reduce cooling costs from ACs. On a macro level this can have big environmental benefits.

  2. I read somewhere about how Portland is thinking about introducing edible species into their city gardens, and providing online maps for what trees are where. Could kill 2 birds with one stone! In regards to spatial observations of patterns, such as the radial cooling in this article, are there general patterns to be found in GIS or is it restricted to disciplines (urban planning here)? When considering the article I read last week about forest fragmentation in the Amazon and the observed "dendritic" patterns that resembled hydrological flow patterns, it makes me think that there are some intersections? So, in order to benefit from these intersections one has to be up to date with findings in GIS in other fields as well as one's own? Will this become easier as patterns are typified over time, if that isn't already being done? I recall the types of map simplifications that can be done from class, so I guess the pattern recognition comes down to the individual thing one is studying and looking to see if it's an already known pattern or not.

  3. I think the whole idea behind MUHIs is completely fascinating. This data further reinforces the idea that vegetation reduces the heat island effect. It would be interesting to do a quantitative map along with the MUHI map on top so we can see the quantity of vegetation in comparison to the heat island.