History of Remote Sensing: MSS Histograms - Lecture Material - Completely GPS, GIS dan Remote Sensing tutorial - facegis.com
History of Remote Sensing: MSS Histograms

The MSS can simulate rather crude spectral signatures. Bands 4, 5, and 6 each have a bandwidth of 0.1 m; and band 7's width is 0.3 m. We represent band responses by bars in a histogram-like plot, in which the height of the bar signifies the relative reflectances averaged for all wavelengths within the bandwidth interval. Several crude spectral signatures (as bar histograms) are shown here, as derived from this Landsat image of a section of the U.S. West Coast.

This Landsat-1 image of central California contains common classes whose spectral responses as measured by the MSS consist of 4 bars (representing the 4 MSS bands) whose relative heights depict differences in reflectance. Note the similarities of the several vegetation types and the somewhat similar water and urban classes.

The scene is the first color composite made from ERTS-1 digital data. This includes Monterey Bay (lower left corner), the Coast Ranges below San Francisco, the Sacramento (Great) Valley, and the western slopes of the Sierra Nevada. Notice that the pattern of each histogram set, for a particular type of surface cover, differs from the others. Thus, each class of material has a distinctive signature, roughly approximated by the 4 bars, that sets it apart. To simplify the plot, we made the width of band 7 (bar on the right) equal to the other three. Urban is most reflective in bands 4 and 5; suburban is strong in bands 4 and 7 (the latter is tied to the influence of live vegetation in lawn grass and landscaping trees). The signatures for forest and healthy croplands are similar, but the heights of the bars for bands 6 and 7 are greater for the crops. The bar heights for small grains and fallow fields are similar, but the response for bands 4 and 5 is just a bit higher than for 6 and 7. The two very dark areas in the Coast Range and in the Sierra Nevada (not shown as histograms) result from the predominance of conifers.

Next, refer to the table on the next page in which we present some criteria for using combinations of band gray tones or colors and their patterns to identify land-cover categories. Most categories are described by their MSS Bands 5 and 7, and sometimes 4, response. Again, apply these to recognize examples of any such categories in the California scene. Become familiar with this table, because we challenge you to practice this approach whenever you examine various Landsat scenes (and imagery from the other earth-observing satellites) in this Tutorial.

Source: http://rst.gsfc.nasa.gov