Other Color Combinations - Lecture Material - Completely Remote Sensing tutorial, GPS, and GIS - facegis.com
Other Color Combinations

Reluctantly, we leave these attractive false and true color composites to examine several others made by different three band/filter combinations. The first we will construct renders Band 1 = red, Band 3 = green, Band 2 = blue:

Another false color image, with TM Band 1 = Red; 3 = Green; 2 = Blue.

Features of special interest in the resulting image are, first, the intensification of the expression of sediment in the ocean, standing out sharply in red, and two, areas of more active vegetative growth (which would have lighter gray tones in the green band) appearing as distinct blues.

Let us now experiment further by using Bands 5, 6, and 7 in some combinations. The first casts 5 as blue, 6 as red, and 7 as green.

Non-standard false color composite; bands shown on right.
  • TM Band 6 = red
  • TM Band 7 = green
  • TM Band 5 = blue

This combines the thermal band, sensitive to emitted radiation, (with its lower resolution 120 m. pixels), with two infrared bands, 5 and 7 (with their 30 m. pixels). In this color composite generated with IDRISI, the thermal band image (check TM 6 again) dominates. In the IDRISI mapping, reds and blues greatly outweigh any greens contributed by light tones in Band 7. The reds represent the warmer surfaces, and the blues the cooler surfaces, as recorded in Band 6. Note that the blues extend over a broader fraction of the hill slopes than we might expect from the shadow effects seen in other bands. Some areas of blue that lie on the sunlit foreslopes are covered by the uplands vegetation at g and the grasslands at v, which offset the direct heating effect through cooling by evapotranspiration. Other areas supporting vegetation (as at y), and also along streams have dark red tones, not because they are hotter but because their tonal expressions in 5 and 7 are darker (almost no blue or green inputs).

Both urban communities show several shades of red, with the street patterns showing through because of their sharper definition in bands 5 and 7. This is, in part, due to the "urban island" thermal effect, which is the tendency of populated areas to be warmer because of heat-absorbing materials (darker streets, tar-covered roofs, etc.), reduced surface areas with vegetation, and heat emitted from furnaces, air-conditioners, and other human activities. Fully perceptible is the thermal plume (whitish patch) emanating from near the power plant (t), but the ocean sediments introduce no noticeable effect. Note that the breakers now completely disappear (are black) since they provide no reflectance input in these bands.

Some areas we see as light greenish-blue in this composite include the beach bar and several of the extraction pits (u). These surfaces reflect highly in most bands, hence, their light tones in Bands 5 and 7 combine the blues and greens assigned to these bands whereas their darkness in Band 6 (reflective materials are cooler) leaves red out.

Let's now experiment briefly with combining one of the reflective longer-wave IR bands (7, assigned to green) with the vegetation band, TM 4 (blue), and the ocean water/sediment band, TM 1 (expressed in red), as produced by the IDRISI Composite function. Remember it converts three 8-bit bands into a single 8-bit image that mimics the results of using the three bands to make a true 24-bit color image. The result is a colorful rendition that, in some respects, shows certain features best.

Non-standard false color composite; bands shown on right.
  • TM Band 7 = green
  • TM Band 4 = blue
  • TM Band 1 = red

Again, blues and reds are the prevailing colors, whereas greens are subordinate. The blues correlate closely with vegetation, which stands out in sharp contrast to most areas that either lack this cover or now support dormant grasses, etc. (note the active grasslands in lighter blue). The reds tie into three principal surface classes: the waves (note the bluish-purple tones within them, representing the Band 4 contribution), which now sharply separate from the beach sands; the sediments; and the towns, with the red streets speckled by blues from local vegetation. The greens show some specific features that are relatively brighter in Band 7 and darker in Bands 1 and 4, including areas in the hills around Los Osos and some uncultivated areas (m) in valleys. Note scattered greens along the otherwise bright slopes.

In this last image, you no doubt noticed that there are some obvious similarities between this Band 7-4-1 combination (when the bands are stated this way, they refer implicitly to the color sequence BGR or blue-green-red; sometimes they are also written 741) and the Band 2-3-1 composite at the top of the page. But several distinctions occur: In the 2-3-1 version the breakers are white whereas they are reddish in the 7-4-1 version. This is because the breakers have about the same gray levels in Bands 1, 2, and 3 but are darker gray in 5 and almost black in 7. The blues in this last version are more intense and widespread since Band 4, which shows vegetation as very light tones, is rendered in blue and thus effectively emphasizes the locations of active growth. Also, note several areas of red in the 7-4-1 composite that are found within parts of the built-up or urban class. The color patterns are not the same in the 2-3-1 rendition, suggesting that equivalent areas are suppressed tonally in Band 7 and/or 4. If curious, retrace to page 1-3, which has all 7 bands together, to try to deduce whether it is one or the other or both.

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