Several Case Studies of Fracture Analysis - Remote Sensing Application - Completely Remote Sensing, GIS, and GIP Tutorial -
Several Case Studies of Fracture Analysis

The East African Rift

To test the merits of space imagery for finding fractures on Earth's continents, we examined a region, in which these are clearly visible and reliably mapped, as a ground-truth site using Landsat and SPOT scenes. The East African Rift is a zone of crustal extension, in which part of the eastern African continent is pulling away from its parent along one arm, separating the divergent blocks that stem from the triple junction, centered in Ethiopia (the second arm includes the Red Sea and the third, the Gulf of Aden off the Arabian Peninsula). Much of the rift consists of downdropped crustal sections, bounded by deep-rooted normal faults (forming grabens) that cut into the basaltic lavas, extruded in the resulting depressions. (In the fault type called "normal", imagine a rectangular block broken by a fault that inclines downward at some angle; the upper part of the block then slides down the fault plane; if there are two normal faults facing each other but the second inclined in the opposite direction (so the two converge), then the part of block between the two faults drops as a unit [the graben] becoming lower and leaving the blocks on either side of the fault pair higher; see illustration near top of page 3-2)

These faulted lava fills, because they are recent and not severely eroded, have strongly-expressed surface scarps. They don't show tonal differences on the two sides of a fault, because the faulting affects the same rock type as seen at the surface. The next ground photo shows a part of this East African Rift in Ethiopia, with prominent individual faults having scarps all facing in the same direction.

Aerial oblique photo showing step faulting in volcanic rocks within the East African Rift Zone, in Ethiopia.

The scene below shows part of the rift in Kenya (Lake Hannington is at the top) as imaged by SPOT (described in the next Section).

A SPOT color composite image displaying the East African Rift Zone in Kenya.

This ASTER natural color image shows the topographic expression of these rift faults:

ASTER image of a small segment of the African Rift Valley.

A map of rock units and fractures (most are faults) of the upper half of this image was made by Kenyan geologists by field mapping and aerial photointerpretating. We reproduce it here (at a quality that precludes reading details) to illustrate the reference database.

A geologic map of the same area of the E.A. Rift Zone portrayed in the previous image.

The next figure shows rose diagrams prepared by the author from his interpretation of Landsat Multi-spectral Scanner and Thematic Mapper images and a stereo pair from SPOT (see page 11-9).

A set of eight rose diagrams (plots of compass bearings) that depict the orientations of faults and fractures in the Kenya scene being examined. The plots are divided into a left set representing areas west of Lake Hannington and a right set for areas to the east. Different sources of data are stated.

Areas west and east of Lake Hannington were treated separately. Hence, there are two sets of diagrams. Judging from the similarity of the orientations in the ground-map diagrams, Landsat and SPOT do an excellent job defining larger fractures. These topographic expressions are so pronounced that shadow bias is minimized, thus allowing accurate measurements of their trends.

The Canadian Shield

Lineament analysis using space images has been particularly valuable in determining regional and subcontinental fracture patterns that reveal some of the stress history imposed on large crustal units. Surface exposures of deep crustal rocks are known as shields which are mostly igneous and metamorphic, and form the basement core of the continental craton. Shields typically contain major sets of fractures developed mainly at the time of emplacement of the individual terrains that collided and assembled into these primary crustal units. In northern Canada, where soil and glacial cover are minimal, the Shield there contains a very high density of fractures, shown in the image below.

Landsat MSS Band 5 image of part of the crystalline Shield in the northwestern part of the Northern Territories of Canada, showing the dual effects of low sun angle and snow cover that bring out numerous lineaments.

A segment of this image has been enlarged, contrast-stretched, and enhanced by filtering to emphasize the linear expression of the fractures.

Enlarged part of the above image, enhanced to bring out the lineament patterns.

These are highlighted by three factors: 1) in this winter scene, the Sun's elevation angle is only nine degrees;, so that the low-angle rays emphasize small differences in topography associated with linear depressions, 2) the snow cover acts as a dusting which emphasizes these features similar to sprinkling a powder on a fingerprint, and 3) many of the fractures have been gouged out by glacial scouring.

The writer (NMS) was part of a team at Goddard Space Flight Center that produced lineaments maps for selected areas within the Canadian Shield, which comprises most of the eastern 2/3rds of Canada. Their goal was to better understand its structural history. Because glacial scouring has removed covering soils and has sculpted lineaments into lakes, fractures over large parts of the shield are clearly evident, as demonstrated in this example.

Part of the Canadian Shield, northeast of Sudbury, Ontario, showing glacially sculpted fractures, many occupied by lakes. Fractures north of the Grenville Front (a diagonal boundary running through the middle of the image) lie in the Superior Province; those south of the Front are in the Grenville Province. Rock types are mainly metamorphic and igneous.

It is the frequent filling of scours by lakes that brings out the presence of the lineaments; the lake waters are very dark in IR images. This aerial photo illustrates the control by more easily eroded lineaments on lake distribution after scouring.

Aerial photo showing glacial-scoured linear fractures in the southern Canadian Shield, most now containing lakes.

Simply by looking at the Landsat image, we can tell that the terrain in the middle of the image (below the upper-left corner) is different from the remaining terrain to the south. The terrain in the upper left corner marks the Superior Province of the Shield, consisting of meta-igneous rocks, whose ages go back beyond 2 billion years. The terrain to the right is the Grenville Province, whose rocks are generally younger and were emplaced northwestward against the Superior Province some 900 million years ago. The line, is the Grenville Front, made of thrust faults that are seen from above as a NE-trending boundary. Many of the fractures in each province stand out as linear lakes. These were glacially-scoured gouges that are now filled with water. When the orientations of the lineaments are plotted in rose diagrams for each province, two characteristics emerge:

Rose diagrams (showing azimuthal orientations) of the two tectonic provinces of the Canadian Shield on either side of the Grenville Front; on the left = Superior Province; on the right = Grenville Province.

1) both provinces have a dominant ENE fracture trend and 2) the second maximum is NNE for the Superior (left diagram) and NNW for the Grenville (right diagram) Province. This example underscores the power of space imagery in providing observations that are well-suited to scientific analysis of regional structures.

Over part of a one year span lineaments orientation and frequency measurements were carried out by a group at NASA Goddard for the entire Canadian Shield (this probably was the largest area so studied by regional fracture analysis attempted up to that time [1986]). One objective was to see if fracture patterns differed from one tectonic province to the others. From this general rose diagram for the crystalline shield, the answer appears to be a firm yes. This suggests that as each province was emplaced as an accreted terrain (see Section 17), the stress field at the time was responsible for the observed lineaments pattern for that province; stress fields for different provinces as they docked and added to the continental craton had different orientations relative to today's geometry.

Composite rose diagrams for lineaments and fractures located in and averaged for the Tectonic/Age Provinces of the Canadian Shield.

Lineament analysis has a practical side, as well. For example, fractures and faults can serve as channelways for circulating solutions that eventually deposit ore minerals. They also are involved in traps for oil and gas and can be instrumental in storing and moving ground water. Original or improved lineament maps, therefore can payoff economically.

Atlanta Groundwater Survey

Our next example emphasizes this economic point. In the mid-1970s, the State Geologist of Georgia was commissioned to show where in a small area northwest of Atlanta conditions would be favorable to locate a new industrial park, which would require large amounts of groundwater for plant use. He included space imagery as a means of locating fracture patterns that could control water accumulation. The results of this effort are summarized in the sequence of image panels shown here.

Six panels consisting of aerial photos, Landsat images, and ground views of an area northwest of Atlanta, GA., which summarize a practical use of space imagery in searching for groundwater; see text.

The upper left panel is a Skylab photo of Atlanta with the survey area outlined in a box. The image to its right is part of a Landsat MSS image that shows layered rock units in the Piedmont Province of the Appalachians. In the center-left image, data from Landsat were computer-processed by a technique known as edge enhancement that highlights linear features by bounding them with a light-dark contrast. The strata are out clearly visible from this technique, but also exposed are numerous parallel lines, running across the strata from upper left to lower right (almost east-west when properly reoriented). These parallel lines looked suspiciously similar to a single, prominent, joint set. When we see them in a high altitude aerial photo, taken from an RB-57 aircraft, they appear as small linear valleys (right center). When visited on the ground, they appear as joint valleys, such as in the lower left panel. When we walked out several of these joint valleys until they intersected the Interstate 285 Beltway, we found that they flowed into road cuts, where exposures such as that shown in the lower right panel, identified them as fracture zones. The porosities associated with these zones made them favorable as potential aquifers for groundwater storage.

The area was drilled and large supplies of groundwater were proven. This led to an approval for developing the industrial park. The results of this study were brought to the Governor's attention, and he was so impressed that he insisted on going into the field with the State Geologist to see the evidence. A year later that governor became President Jimmy Carter, who was always a great friend to the Landsat program. This study also was cited in Congress as prima facie evidence of the value of space technology to the nation.

Oklahoma Lineaments Analysis

As mentioned on the previous page, the writer (NMS) was for several years at Goddard Space Flight Center the Landsat Discipline Leader in Geology. As such, I was responsible for evaluating 57 initial Principal Investigator studies. One, known as the Eason Oil Study, is summarized on page 5-5. Its main concern was to look for surface indicators of subsurface oil deposits in the Anadarko Basin of southern Oklahoma. One clue that the investigator team reported was the widespread occurrence of lineaments, some of which they argued were possible factors in the development of these deposits (their principal claim, stated on page 5-5, was that over many of the developed oil fields leaks along the fractures led to chemical changes in surface rocks and soils which were expressed as tonal anomalies in ERTS images).

The writer (NMS) has tended to be a skeptic about much that was reported in the early days of geologic studies of satellite imagery. That attitude was paramount in my evaluation of the Eason Oil results (as will be presented on page 5-5). It was doubly re-enforced by my doubts about their lineaments mapping in Oklahoma. So, an experiment was designed. I and three other geology colleagues at Goddard each made our own lineaments map from the prime Landsat image (black and white Band 7 transparency) that included the Anadarko Basin study area in Oklahoma. Prior to this effort, all of us met for an hour to try to establish some common ground for our decision-making in selecting the linears we would then map. The results are summarized in these two illustrations:

Maps of linear features made by the four Goddard geologists.
Plots of orientations and frequencies of linears chose by each geologist operator.

The writer, Operator 3, was the most experienced of the four in lineaments mapping. Thus, I found the most. Operator 4 was confused about some of the mapping criteria and probably found too many spurious linears. Operators 1 and 2 - both experienced structural geologists - were more conservative in what they considered bonafide linear features. One positive note, all the first three identified the principal aximuth trend of linear directions in the region. Comparing our results with those reported by Eason Oil resulted in a 59% agreement in picking out the larger lineaments. Conclusion: once more the subjectivity and variability of individuals in mapping lineaments was confirmed.

To close, we give you a mini-minitest to see if you have learned how to search a space image for specific information. Look over this image.

Test image.

A closing remark: During the first ten years of Landsat, its fullest use, as measured by sales of data to the user community, was in geology. Landsat, and later SPOT and other space systems, including radar, became respected new sources of information of particular interest to those who explore for oil and gas, mineral deposits, and groundwater. We will describe remote sensing applications to geological resources exploration, such as metal deposits and petroleum fields, in Section 5. But first, in the next two Sections we introduce two other major application regimes: in monitoring vegetation and assessing land use and urban planning.