On page 1-5 Bands 1 and 4 were compared, with notable tonal differences between them. On this page, Bands 2 and 3, which are close together in the spectral intervals sensed by the TM, are shown side by side. At first glance, they appear almost the same in tonal balance and intensity. But small differences can be discerned, which become important when color composites are made and/or special computer processing (described later in this Section) is applied. In Band 5, is shown for the first time, you should note any differences in its tones compared to Bands 2 and 3 and to Band 7. The page ends with an "aside", namely, a brief discussion of what has been called "linears" (these will be discussed further in Section 2).
Returning for a moment to the two other visible bands, Landsat TM 2 and 3, which monitor reflected light broadly distributed in the green and red bands, respectively. First look at Band 2 (left) and then Band 3
(right) and, if you wish, compare each of them to Bands 1 and 4. On first glance, there seems to be little difference in tonal variations for features imaged in Bands 1 through 3. The most perceptible change is the extent to which the ocean sediment tonal patterns diminish in going from Band 2 to 3. The sun-facing slopes in the hill country seem a bit brighter in 3 than in the other two bands, perhaps influenced by the red component in light reflected from a surface underlain by brown soils and also by the brown color of the grass.
Because of the apparent similarities in tone for major features seen in the first 3 bands, we might wonder how they combine so well into a color composite, which we will inspect shortly. The answer to this question is subtle and rather complex but here is the essential reason: when images for the individual bands are generated, they have appropriate contrasts. The images have similar tonal variations for the ground classes, because most of the features have approximately the same relative differences in reflectance in each band. But, their actual reflectance values (partially embodied in their DN's or digital numbers [see page I-5a for a brief synopsis of what DNs are]) vary more from one band to the next. Thus, a green reflector may have a higher average DN value in Band 2 than in 1 or 3 and that will contribute more effective brightness to the green component of a color composite.
Now, we will finish our look at the black and white renditions of the individual bands by displaying TM Band 5. This band, centered around 1.65 µm, most closely resembles Band 7 but by flipping back and forth into 7 and 5 you should be able to pinpoint several moderate differences. In the beach bar (c), Band 5 shows a moderately darker tonal variation, with some structure in the patterns within the sand other than that caused by vegetation. As with Band 7, the breakers have lost most of their reflectance output and appear therefore as mottled medium-dark grays. The mysterious feature (s) has good internal contrast highlighting the distinctive patterns that give away its identity (to the imaginative). The urban areas' structure, defined by the crisscrossing street patterns, (within towns labeled in yellow on the overlay map), is rather washed out in the Band 5 version. The two rows of bright dots representing storage tanks near the power plant (t) almost disappear in 5 (and 7).
1-9: There is something in Bands 5 and 7 that is almost absent from view by its lack of contrast; but, this feature stands out in Bands 2 and 3. What is it? (Hint: it's in the lower half of the image.) Which band seems lighter overall, 5 or 7? ANSWER
Before leaving this seven-band review, we comment on a peripheral item but one that illustrates a typical danger and misuse of space imagery. Look at any of the images (particularly Bands 5 and 7) at two points marked w. You should see two thin dark features that are almost straight. These "linears," as some call them, are common phenomena, observed in satellite images and in aerial- or astronaut-camera photographs. Peruse the rest of the image you chose and you should find more linears (although some are actually sharp boundaries between two features or classes). In the early days of Landsat applications, many geologists reported that the MSS (and, later, TM) sensor was especially adept at highlighting such linear features. They chose to identify a significant fraction of these features as geological in nature, such faults or fracture systems. Maps showing numerous linears of presumed faults or fractures were produced and often published but too frequently without appropriate field checking. When several exacting studies discredited this interpretation of many such features (although some, and sometimes a majority, were verified), this use of Landsat led to widespread skepticism and negative criticism. Today, we know to be careful, and properly use field inspectionsto verify our work. Incidentally, we haven't established the nature of the two linears at w but one is probably vegetation lining a narrow gully. The fact that the two line up may be a coincidence and does not prove any commonality.
1-10: Make a list of other features besides faults and fractures that have linear expressions. ANSWER
