The Klamaths from Space - Remote Sensing Application - Completely Remote Sensing, GPS, and GPS Tutorial
The Klamaths from Space

For this geomorphology research project, I selected a segment of the Oregon Klamath Mountains for prime attention, although for some tasks I also looked at adjacent areas. The following map displays the area, on which I drew terrane boundaries, labeled the terranes, and assigned stratigraphic symbols to each (the first letter(s) denote(s) their ages in terms of geologic periods, with Tr = Triassic; J = Jurassic; K = Cretaceous; Tel = Tertiary; Q = Quaternary). I digitized the boundaries on the map to allow "cookie-cutting" of geographic data in other maps, so that he could group the data into categories for terrane analysis.

Terrane names and symbols map of the Oregon Klamath Mountains.

As a visual aid to realizing that there are distinctly different terranes, we reproduce here a simple computer-generated map (unlabelled) that assigns a different color to each terrane. Match with the map above. The easternmost terrane shown is Jrv, which stands for Rogue Valley (Western Triassic-Paleozoic may not be a terrane). As a guide, the Brown stands for Rogue Valley, the Dark Blue, Medium Blue and Blue Gray denote Dry Butte, Smith River, and Yolla Bolly respectively; the Light Blue marks Elk River, the Purplish tone specifies Pickett Peak, the Brownish-red unit against the ocean is Sixes River and the Medium Tan represents the Tertiary units. Yellow denotes Quaternary deposits - both river and lowlands fill; Red near the right corresponds to intrusive granites (Quartz Monzonites).

 The principal terranes in the above map have been assigned colors and positioned on a computer-generated map to better indicate to the eye their distribution.

By comparing the above terranes map with the geologic-units map for all of western Oregon (below), you may be able to see that there is some correspondence between some units and their enclosing terranes, a broader correspondence for others, a number of individual units that are scattered about within the terranes, and some units that seem to cross over or overlap terrain boundaries.

Geologic (stratigraphic) units map of the Klamath study area.

Because there are so many units [generally, at the formation level], I don't show the map's legend. Also, in the legend sequence are units that aren't present in the Klamaths. Two examples of exceptions are the deep purple unit (Jui), which is actually outliers of ultramafic (ophiolitic) lavas, once part of oceanic crust, which appear similar in the field but originated locally within their terranes, and the pinkish-red unit, which consists of intrusive rocks that invaded several terranes, already in place. One expects this general agreement between terranes and their constituent stratigraphic units (i.e., internal consistency), along with the exclusivity of some units to single terranes, inasmuch as terranes typically develop from rocks formed in different source areas at different times.

The Oregon Klamaths are among the most rugged landforms units in the state. We can gain a sense of their appearance by inspecting this 3-D perspective image of nearly all of Oregon, generated from DEM data:

3-dimensional shaded relief map for nearly all of Oregon generated from DEM data.

Features of interest in this shaded map are: A = Klamath Mountains.; B = Coast Ranges; C = Willamette Valley (Portland at the north end); D = High Cascades; E = Mount Hood; F = Crater Lake; G = Paulina Mountains (Newberry Crater); H = Blue Mountains.; I = Abert Rim; J = Summer Lake area (the circular area is somewhat an artifact of the illumination)

You can get a feel for the appearance from the ground of several of the Klamath terranes through these photographs taken during the writer's time in the field:

Ground photos of typical landscapes within the Oregon Klamaths.

The upper left scene shows terrain around Canyonville, with Yolla Bolly terrane in the foreground and Sixes River terrane in the distance. Next to it is a view from the coast of the Elk terrane. In left center is a segment of a canyon dissected by the Rogue River, in which sedimentary units of the Rogue Valley terrane (Jrv) are exposed. To its right is landscape typical of the higher elevations in the Klamath Mountains. The lower left scene is an exposure of serpentinite (the metamorphosed ophiolitic basalts associated with oceanic crust). The lower right shows alternating layers of siliceous (cherty) shales that constitute island arc sediments. Here they are part of the Yolla Bolly terrane.

The wildest part of the Klamaths is the Rogue River which cuts deep canyons into the rocks. This is one of America's premium whitewater rafting streams. This photo captures it ruggedness:

Photo of a narrow gorge formed by the Rogue River in the Klamath Mountains.

We proffer a preview of the topographic expression of the Klamath Mountains, which will be presented in Landsat format shortly, by this enlargement of the area made from the same DEM data. Notice Crater Lake (nearly circular on top of a mountain) in the upper right part. Another simple terrane map is placed below the DEM image. This map is a photo negative of a transparency (terrane boundaries in black) made to superpose onto a Landsat mosaic (discussed later).

Enlargement of the Oregon DEM image to cover mostly  the Klamath Mountains; Crater Lake near right center.
The terrane boundary lines in white, covering the Klamath region that occupies the western 2/3rds of the above DEM image (horizontal dimensions deviate a bit.

In this visual version, we have trouble seeing any significant variations in terrain that call attention to noticeable terrane differences. Of course, we can vary the illumination directions to emphasize contrasted ridge and valley orientations. We did vary the illumination directions on this data set, giving several distinct expressions but, again, obvious patterns that correlate with the various terranes as mapped did not stand forth. However, when one learns where to look, after familiarization with terrane locations, strong hints of certain differences for several of the terranes are discernible. These differences can also be disclosed by appropriate analysis of topographic data in map-sheet or DEM formats and, in fact, this approach proved superior to visual differentiation.

The reason we would postulate for expected geomorphic differences is that each terrane consists of a group of rock types that normally differ from other nearby terranes - this is known from field work. Fault discontinuities also bound each one, at least partly. A terrane will respond to regional erosive action according to its mix of lithologies. Thus, any one terrane may develop landform characteristics that differ from its neighbors and can appear visually as separable. We should be cautious about expecting the differences to stand out sharply across terrane boundaries, because, as the landscape develops within any one terrane, the equilibrium forms tend to exert some influence on terrains outside the boundary. Nevertheless, the hope in testing the hypothesis of distinctive terrane-controlled topographic expression is that real differences do occur when each terrane is considered in its totality as compared to its neighbors.

In planning the geomorphic analysis, I set forth this strategy:

Geomorphic analysis strategy chart.

This effort is predicated on my being able to make a series of quantitative measurements that can be tested numerically and statistically to ascertain valid differences in the topographic character of each terrane. As it turned out, most of the measurements could not be made from Landsat as a test bed. In general, it is quite difficult to extract any of these measures with confidence from Landsat imagery. However, one overlapping SPOT image pair was available, so that a stereo image is achieved, allowing limited recovery of measurable variables that depend on topographic variations. So, in the exposition that follows, Landsat's prime role is to help to confirm the terrane-terrain association by visual recognition rather than extraction of terrain parameters.