Recognition of Faults by Remote Sensing - Remote Sensing Application - Completely Remote Sensing, GIS, and GIP Tutorial -
Recognition of Faults by Remote Sensing

When earth materials within bodies (such as blocks or layers within the crust) are subjected to forces (or stresses, which are forces applied to finite areas), they tend to fail - either by fracture or by plastic flow. Fracture is a general term but it has a second meaning - breaking along irregular, often curved, surfaces. Breaks that tend to create planar failure surfaces are called faults or joints. Most earthquakes result when the blocks on each side of the fault suddenly slip against each other and move (are displaced).

Any fracture that lies beneath the surface and also intersects that surface will produce a line or linear trace (in geometry, the intersection of two planes is always a line).

Because of their role as the prime cause of earthquakes, faults - and their detection (as from space) - are of special interest and concern. Faults are fractures along which there is relative sliding movement (offset) of the blocks in opposite directions on either side. We recognize them by various criteria:

1) Layers of different types and ages of rock units sit side-by-side (offset)

2) Abrupt topographic discontinuities of landforms

3) Depressions along the fault trace (broken rock is more easily eroded)

4) Scarps or cliffs

5) Sudden shifts of drainage courses.

6) Abrupt changes in vegetation patterns

The relative movement direction along the plane of fracture (fault plane) and the angle of that plane determine the type of fault that is involved. Nomenclature of fault types is shown here:

Types of faults
From Contemporary Physical Geology, H.L. Levin, Saunders Publ.

For normal faults (caused by tensional forces), the hanging wall block moves down the fault plane, so that marker beds within the block drop downward relative to their once contiguous counterparts in the footwall. Reverse and thrust faults (compressional forces in the crust) involve upward hanging wall movements; thrust faults also carry their blocks outward, overriding the rocks on the footwall side. Strike-slip faults (shear forces) result in largely horizontal movements along a steep to vertical fault plane.

A ground example of each of the four most common types follows. First, a normal fault:

Two proximate normal faults; note angle of fault plane trace and deduce which side has dropped down.

This is a small reverse fault in a roadcut in Japan:

A reverse fault, with only about a meter of displacement.

And then, a thrust fault:

A thrust fault in the Andes of Chile; note angle of fault plane.

And last, a strike-slip fault.

A strike-slip fault, evident here because the plane itself is eroded into, and the topography mismatch indicates very different units are moved laterally.

If the displacement has only been a few meters or so, the sense of movement for normal and reverse faults is possible, where seen in outcrops, by matching the patterns of bedding. If the displacement is large (tens to hundreds of meters), chances are the outcrop is too limited in size to show the now separated beds. Often, the sense of movement can be gained another way: as the blocks slide past each other, the layers on either side are bent as "drag folds". This diagram depicts the expected bending geometry:

Drag folds indicate direction of footwall-hangingwall movement.

As a general rule, strike-slip and thrust faults - which usually involve considerable displacements - are often easy to pick out in space imagery (assuming little vegetation or soil cover). Normal, and especially reverse, faults are much less frequently visible.

Strike-slip (wrench) faults are usually the best displayed when viewed from above (space or aerial) because they can cause lateral shifts of initially continuous rock units to move cumulatively over millions of years for many kilometers. This brings into juxtaposition units that generally are quite different as now positioned. Thus, abrupt change in lithology across the fault plane at any location along the fault are the hallmark of this kind of displacement.

The discontinuity owing to offset of once continuous layers or other rock units along the strike-slip plane is often remarkably displayed in a space image. The scene below of the Kuruk Tagh fault in part of the Tian Shan mountains of westernmost China shows a sharp fault trace running east-west through the Kuruk Tagh mountains. It's composed of folded sedimentary strata, metamorphic rocks, and igneous intrusions. The block of crust on the north side (top of image) has shifted sub-horizontally to the west at least 60 km relative to the block containing the corresponding segment of mountains to the south. This type is a left-lateral wrench fault (also called a strike-slip fault). (Left lateral means that the block on the far side of the fault plane moves left relative to an observer standing on the near side; right lateral describes a movement to the right.)

The Kuruk Tagh fault is similar to the San Andreas fault, which is a right-lateral fault, with the Pacific plate moving northward against the North American plate (see below). The Kuruk Tagh fault is easy to identify because of topographic offset (as well as equivalent parts of the strata and metamorphosed rock units), fault scarps, and displaced drainage. This and similar major wrench faults in south-central Asia represent crustal adjustments to the stresses induced by the collision of India against Asia (see mosaic in Section 7).

A lateral or strike-slip fault separating two segments of a mountain block in western China, as seen in a Landsat image.

Lateral displacement (one mode of offset) is also evident in this Landsat image of the Altyn Tagh fault, also in western China near Tibet and the Kunlun Shan (mountains):

The Altyn Tagh fault.

Another major fault zone in western China near Tibet is the Kunlun strike-slip (left lateral) system. In the ASTER image below, the fault has split into two parallel segments. The lower one has produced a topographic barrier against which water (black) has been impounded to form a long, narrow lake. Above a series of alluvial fans is the upper segment from within which water has emerged to flow downslope and to enable vegetation (red) to grow.

Aster image of the Kunlun Mountains showing two prominent east-west wrench faults.

Discontinuities are revealed at space observation scales when a sequence of layers abuts abruptly against another sequence trending in a different direction. This is evident in this Landsat subscene in western China:

Two sequences of strata, one trending SE, the other NE, in China; this is evidence of a major fault at their plane of juncture.

Comparable to the Chinese faults is the Dead Sea Fault that runs from just below the mountains of east Lebanon southward through the Sea of Galilee and the Dead Sea (both actually lakes), thence into the Gulf of Aqaba. This right lateral fault marks one of the three arms of the Afar Triple Junction (see page 17-3). Here it is shown in a mosaic made from two Landsat images.

The Dead Sea wrench fault, running through Israel, the West Bank, and the east Sinai.

The fault is recognized in part by topographic discontinuities (mountains in Jordan not fitting with those on the western side) and because the Jordan River and the two lakes follow zones of weakness that are erodes so that they are lower than their surroundings (land adjacent to the Dead Sea has the lowest elevation on land anywhere on Earth; in places below - 300 feet).

Now, take a look at this Landsat image and 1) try to identify its location (w/o peeking at the caption), and 2) see what evidence you find for possible faulting.

The landscape of southwestern California.

This region in the U.S. famed for its propensity to experience frequent earthquakes is much of California from the Mexican border to about 100 km (62 miles) north of San Francisco. North of the Los Angeles Basin is a series of mountain ranges trending towards an east-west orientation. These collectively are known as the Transverse Ranges, and include the San Gabriel and Santa Monica Mountains north of Los Angeles. They are bounded by faults that have a strong effect on their topography. These faults appear in this physiographic map of southern California:

Map of part of southern California showing major faults and topography (reds denote higher altitudes; blues lower.

Best known is the famed San Andreas Fault, the earthquake-maker running from the Gulf of California through the California Coastal Ranges north of San Francisco (see the Los Angeles mosaic further down this page and the L.A. image in Section 4 and the San Francisco images in Section 6).

Although most of these western California faults are wrench or strike-slip types (dominantly horizontal slip motion), they are capable of influencing the fronts of ranges in a manner similar to the normal-type of fault. Here is a perspective view of the Transverse Ranges made by combining Landsat imagery with topographic data acquired by JPL's SRTM mission (radar altimetry). Fault lines (in white) have been drawn on the resulting image, which also depicts the Los Angeles Basin (bottom center) and the Mojave Desert (right).

Landsat7-SRTM perspective view of California's Transverse Ranges; the San Andreas fault marks the northern edge of these mountains (to the right).

No doubt the most famous fault in North America is the San Andreas, a major strike-slip type, that runs from the Gulf of California northward more or less parallel to the California coastline until it finally passes out to sea as a transverse fault in Bodega Bay north of San Francisco. In the ranges north of Los Angeles the San Andreas marks a prominent straight boundary with the southern Mojave Desert. This is evident in this image which is actually a beautifully produced aerial photomosaic that includes the Los Angeles Basin, with its many cities, to the south. In the mosaic, note the arcuate fault trace which is the topographic expression (makes a narrow valley) of the San Gabriel fault.

Aerial photomosaic of southern California from Los Angeles to the Mojave, showing the strike-slip San Andreas and Garlock faults that bound the angular salient along both sides of the Mojave Desert.

An aerial oblique photo shows the fault in an area of the Coast Ranges in the Carizzo Plains northeast of Morro Bay.

Aerial oblique photo of the edge of the Coastal Range northeast of Morro Bay, in which the San Andreas fault is expressed as a straight low ridge.

Segments of the San Andreas have been imaged by Landsat, SPOT, and radar systems many times. This is an airborne radar mosaic of the fault in the Carizzo Plains area:

Radar image of the San Andreas fault in the Carizzo plains area.

Now we show an unusual portrayal of its appearance in a C-band image specially processed to give landform information, with elevation variation shown as a series of color bands. The imaging instrument is the Shuttle Radar Topography Mission (SRTM) that is discussed more fully in Section 11. Compare this image with the aerial photomosaic shown above: the straight boundary along the northern Tranverse Ranges in both images stands out. The image orientation is shifted somewhat from the mosaic, with the Mojave salient apex now pointing down at the left.

A interference image made by processing Shuttle Radar Topography Mission (SRTM) data, showing the Mojave salient and the sharp boundaries made by faulting.

As is described in Section 11 and elsewhere, imagery coupled with elevation data (in STRM's case, from its own stereo-like capability) can be recast in the perspective mode as though it is being viewed much like an aerial oblique photo. Here is an STRM construction of the topography west of Palmdale, Calif. (an earthquake-active area) in which a somewhat straight valley (holding the lake) roughly coincides with the San Andreas fault.

 A perspective view made from SRTM radar data; the area shown is in the northern Transverse Ranges just south of the Mojave Desert.

Moving on: A second fault type often well-displayed in space imagery is the thrust fault, in which the fault plane is at low angles relative to Earth's surface and the usual direction of movement carries the upper (near-surface) block over the lower block, causing rocks of different ages to be juxtaposed. Thus, a shallow, tabular slice of crust slides (thrusts) over the fault plane and on top of the surface ahead of it. Thrusting can occur on a large scale when a terrane attempts to "dock" on a continental margin and slides up and over. Another way is for a fold to overturn onto one limb, then break off, and slide over the terrain (landscape) ahead, as illustrated in this diagram"

Development of a thrust fault.
From Contemporary Physical Geology, H.L. Levin, Saunders Publ.

If each block consists of rock types that are different in composition and erosive response, these will appear at the surface as intervals of rock with contrasting topography. When seen in the field, the disparity between rock types can be dramatic. Here is a scene witnessed many times by the writer (NMS) as he flew into Las Vegas from San Francisco enroute to the Nevada Test Site. On top of Mesozoic sandstone (buff-yellow) are much older Paleozoic limestones (left; dark gray); this superposition of old on top of younger defies the Law of Superposition (young over old) unless the overlying older unit was emplaced by thrusting:

Exposed thrust fault sheet of older limestone above younger sandstone, near Las Vegas, NV.

We return to western China to see an excellent example of topography related to multiple thrust sheets; fault planes dip to the south, the thrusting is towards the north:

Ridges related to thrust fronts in the Tian Shan of China.

This radar image shows the Kalpin Tagh thrust in China; its boundaries are a pair of wrench (tear) faults on either side.

Radar image of the Kalpin Tagh thrust belt; possibly more than one thrust sheet is involved.

Several thrust slices (sheets) may stack one on top of the other in a staggered pattern horizontally, leading to a sequence of side-by-side thrust block bands that outcrop in mountainous terrains. This topography is superbly displayed in the Landsat scene below of the Pindus Mountains of western Greece and Albania.

Part of Greece and adjacent Albania - the Pindus Mountains - in which there are notably different topographic units, distinguishable visually, each being a segment of the crust that was thrust westward; a series of thrust faults determine this terrain as seen from Landsat.The accompanying map indicates the boundaries of the thrust blocks in the Pindus scene.

One can differentiate five tectonic zones, named in this generalized map, by sight because of distinct topographic variations and tonal differences related to contrasting rock types. The direction of tectonic transport is from east to west (right to left) causing the sheets to partially overlap below the surface, but each front edge occupies a different geographic position relative to the one it overrode and the one overriding it. This zone of thrust belts is part of the Balkan Alpine system, an offshoot of the European Alps that runs sub-parallel to the Apennine Mountains of Italy (see Alps mosaic in Section 7). As part of the tectonic adjustments caused by the African Plate shoving northward against the European Plate, a small tectonic plate underlying the Tyrrhenian Sea (off Sardinia) is squeezed against the Adriatic plate. The Adriatic plate then pushes it eastward against the Aegian plate, underthrusting it and causing the thrust slicing shown here.

As you would expect, thrusting is common in the Appalachians, as a result of compression tearing of sedimentary units during terrane docking and again as the African-North American plates collided. This thrusting is strikingly displayed in this Landsat image of the Pine Mountain thrust in eastern Tennessee; the thrust front is located at the upperleftmost ridge (see caption):

The Pine Mountain thrust, it is bounded by tear faults (southern one shown here); the Appalachian Plateau is in the upper left - the Great Smoky Mountains in the lower right.

Both normal and reverse faults are far less likely to appear in a recognizable format on space imagery - even high resolution aerial photos can fail to detect them. There are several reasons for this: 1) reverse faults are uncommon (in tectonic settings requiring vertical uplift(; 2) many of these faults are relatively small (hundreds of meters or less; 3) the vertical displacement along a normal fault may be only a few to a few hundred meters, so that rock units differing in lithology are not always juxtaposed; and 4) the downdropped side of a normal fault, if large and resulting in significant displacement over time, produces a tectonic basin (the graben and horst style of faulting); this basin is the destination of sediments carried in from higher surrounding fault blocks, fills up (covering the downdropped block rock units), and laps against the trace of the fault plane.

Here is an aerial photo that shows how disparate lithologies on either side of a normal fault surface expression can appear:

An oblique normal fault which has displaced inclined sedimentary layers so that these are offset in the hanging wall (stream deposits hide part of the discontinuity.

Active normal faults (those still developing; earthquake sites) can drop down the hanging wall in one or more movement events a few to ten or more meters. The result is an exposed fault plane - called a fault scarp. This is an aerial view of the scarp along the active Wasatch Front fault in Utah; this fault has been moving for millions of years (intermittent; a few meters at a time), so that the mountains in this photo are the upthrown side - now much higher - and the foreground is the downdropped hanging wall block:

Fault scarp in alluvial beds in Utah.

The scarp can be visible in higher resolution space and aerial imagery. But in three decades of viewing Landsat images, the writer has never seen evidence for normal faults as juxtaposed rock units. IKONOS-type images probably can show this.

There is one setting where normal faulting can be inferred. This is the tectonic situation in which large sections of the crust are uparched or downdropped where the near surface rocks are being subjected to tension. Under tension, normal faulting is expected, leading to the classic graben structure (downdropped block) and complimentary horst (higher block left behind or actually pushed up) (see page 2-9 and page 3-2 for pertinent diagrams [look for Kenya callout). In East Africa, the zone of the crust that is rupturing from tension as an overriden spreading ridge is pulling off part of the eastern continent is responding by having melted basaltic magma reach the surface as lava outflows. As these cool into basalt units, those are pulled apart to form normal faults with some slices of rock between faults dropping down as grabens while others remain behind as horsts. This SPOT image shows a small segment of the Rift Valley in which dark lines mark the fault traces and the blue depicts basaltic blocks on either side:

SPOT image of normal faulting in the East African Rift in Kenya.

This image, extracted from a photo taken by astronauts using the Large Format Camera (LFC), shows a number of step faults (of the 'normal' type) which cuts into basaltic flows making up rift valleys:/p>

Part of the East African Rift zone, here in Ethiopia; normal faulting has produced steep cliff faces in most of the series of subparallel faults.

A spectacular example of a single downdropped block bounded by normal faults on either side is the Rocky Mountain Trench in Alberta, Canada. Here is a Landsat-1 view of this structural graben that is a flat valley midst high mountains:

The Rocky Mountain Trench, cutting diagonally across the Alberta Rockies.

The faults have been verified in the field. But if all one has to go on is the Landsat image, this faulting would be deduced from its effect on the topography. While there are rivers within the Trench, they do not cause the valley, since rivers seldom are straight over long distances and seldom carve out straight bluffs (here, these are fault scarps).

In the United States, the classic region where normal faulting abound is the Basin and Range of Nevada and neighboring states (see page 6-8. The geologic factors that have caused this faulting are discussed on that page (suffice to state that the entire region has been uplifted owing to thermal expansion so that the arched crust has failed in tension). Here are typical horsts (the ranges) and grabens (basins). The infilling of the basins masks the bounding fault planes. Reproduced here is a Landsat subscene that includes several ranges and the alluvial-filled intervening basins:

Basin and Range topography.

Sometimes it is difficult to identify a structural feature in a Landsat-type image. Examine this scene:

The Pei Shan system

The area shown is in the eastern Sinkiang desert of western China. The Pei Shan is the mountain system at the bottom. Cutting across the left third of the image at a slant is a thin zone that separates different land cover types (desert surfaces; hills). Although the writer (NMS) has no maps of this region, he photointerprets the feature to be a fault. However, close inspection of the original image indicates some low topographic expression along the fault line. This could be explained as a dike (rock intruded along the fault plane) but there is obvious uncertainty here.

Thus, from the above evidence we can conclude two things: 1) faults can often be discerned in space imagery by their expressions as discontinuities, differences in topography and lithology on either side, and signs of displacement; and 2) the type of faulting cannot always be inferred until the rock ages are known; reference to maps or even field work may be necessary. One last item: Use of space imagery has led to the discovery of new, previously unknown faults; this can be important because it can reveal potential locations of future earthquakes.