Lab 08 - Thermal & Multispectral Analysis

For Lab 08 we learned about radiant energy and thermal infrared data. We explored the Stefan-Boltzmann constant which states that as temperature of an object increases, radiant energy increases, and radiant energy shifts to shorter wavelengths. We also studied emissivity, thermal capacity, conductivity, and inertia. All these concepts help scientists determine which part of the thermal infrared spectrum to use to sense an object or property of interest. For example, forest fires or soil composition, which have unique thermal properties that can help us identify them from surrounding features.

Lab exercises consisted of exploring band 6 in a multispectral Landsat EMT+ image. By looking at just this thermal layer and adjusting the histogram, I could more easily distinguish forest fires from the surrounding landscape. I also looked at cloud shadows, the overall changes in temperature in the Pensacola bay area during two different seasonal images-winter and summer- with more temperature variation in the bay in winter, and increased temperature of the urban areas in the summer. The next part was looking at a bands 6, 4, 7 to identify "heat islands," or areas of increased temperature in urban areas, in downtown Pensacola.

For my map deliverable, I selected a feature of interest from the NW FL Landsat image and used various band combinations to explore the different properties of the feature. My feature was simply a plot of land that appeared in True Color (3,2,1) to be similar scrubland to the surrounding landscape, however, upon inspection in other band combinations this plot of land was quite different. I selected two band combinations for my final map as both showed interesting qualities of the plot: 4,5,3 in which vegetation is reds and greens, urban features are silvery, and darker green indicates higher moisture content. The plot appeared significantly darker green than the surrounding areas. Also 6,4,7 all the infrared wavelengths, in which green is vegetation, bluish colors are urban features, and pink is soil, with brighter colors indicating increased moisture or certain soil compositions. The plot appeared significantly brighter pink than surrounding areas. (The area was also much brighter looking at only band 6, indicating higher thermal radiance and temperature.) All these spectral signatures suggest unique moisture levels or soil composition properties of this plot.

It was very exciting, albeit a tad confusing, to try to figure out what the different bands were telling me about the feature. Very interesting to see a piece of land that appears similar in true color but is quite different when observing it through other parts of the spectrum. Makes me want to go out and do field testing to see if my remote sensing deductions were correct!

Map using various band combinations to identify and explain unique properties of a selected feature.