Mapping Technique and Limitations
The approximate locations of four of the faults shown on map 2 were known from the published works of McClelland Engineers (1966) and Van Siclen (1967). Their published maps, however, are at scales that are difficult to adapt to local studies; for purposes of land-use planning all four faults were remapped in detail (at a scale of 1:12,000). Two of the faults, the Long Point and Eureka Heights faults, are among the longest and, by virtue of their location, the most destructive faults in Houston. All seven faults shown on map 2 were mapped primarily by tracing an associated zone of damage through developed areas.1
Fault damage may take a variety of forms, but all damage is due to the same cause—“sinking” of the downthrown side* of a fault relative to its upthrown side*. Rigid structures, such as paved surfaces and slab foundations, break along the traces of faults and become progressively more offset as movement continues. Vertical offset is commonly the most visible aspect of fault movement, but inasmuch as typical fault surfaces are inclined at angles of 60°–75° from a horizontal plane (Van Siclen, 1967), any movement on the fault must be accompanied by horizontal extension as well (fig. 4).
Damage due to faulting may locally mimic damage due to other causes (such as swelling clays, general deterioration of structures with age, or substandard construction practices), but usually is distinguishable by the following characteristics:
(1) Damage to cultural features is significantly more severe within a narrow (1 to l0-m width), rectilinear to gently curvilinear zone than in adjacent areas.
(2) The zone of damage is parallel to and coincident with a topographic slope—the fault scarp*—that separates two blocks of land of different elevations. Together the fault scarp and associated zones of damage define a linear feature whose length is typically 20 to 2,000 times greater than its width.
(3) Damage to affected structures is consistent with vertical offset of the land surface as deduced from the fault scarp. For example, if the land surface is lower on the south side of a fault, observed damage to buildings, roads, and so forth must be consistent with south-side-down movement on the fault.
Observations of fault damage provide information on the exact location of a fault for only short segments of its length—that is, only at those locations where manmade structures have been built and show clear evidence of offset along the fault. The fault scarp, however, connects discrete zones of fault damage and thereby establishes a reliable and continuously mappable fault trace. Although the fault scarp may be modified by erosion or locally obliterated by man’s activities, the elevation difference between the upthrown and downthrown blocks remains to suggest the approximate trace of the fault (fig. 5).
Faults in urban areas were investigated at every location where they were crossed by public roads. At some locations where additional detail was required, permission was obtained to trace faults across selected residential lots and through commercial and industrial complexes. Normally, however, visual inspection from a public road sufficed to establish the fault trace through private property. The faults shown on map 2 could thus be traced on a nearly continuous basis.
Techniques in Undeveloped Areas
Typical fault scarps in the Texas Gulf Coast range in height from about 0.2 to about 1.5 m—very low compared to fault scarps in many other areas of the United States. Scarps in the Texas Gulf Coast nevertheless commonly stand out as prominent, relatively abrupt topographic features within the otherwise much flatter and more gently rolling landscape characteristic of much of the region. At many locations a fault scarp is the dominant feature of the local landscape.
Faults in undeveloped areas, such as those depicted on map 1, are commonly visible on aerial photographs as sharp, gently curvilinear to rectilinear boundaries between two areas that differ in tone on the photo. Large-scale (1:10,000 to 1:40,000) black-and-white or color-infrared photographs are particularly well suited for this purpose. Tonal contrasts on photographs reflect, among other things, fault-associated vegetation and soil-moisture contrasts in the field. The topographically low, downthrown side of a fault, which catches and holds much of the surface runoff from adjacent areas during periods of rainfall, commonly supports a different plant community than is found on the upthrown side. In extreme cases, shrubs, small trees, and prairie grasses dominate the upthrown side, whereas the downthrown side is a marsh and supports only those plants that can grow in a permanently wet environment. In winter, when bare ground is visible through remnants of the previous season’s growth of annual plants, tonal contrast between the dark, relatively wet soils of the downthrown side and the lighter and better drained soils of the upthrown side may be particularly evident—often more so than the fault-related vegetation contrasts of the growing season. During extended periods of little rainfall, however, soils on both sides of the fault will dry out and photographic evidence of the fault may all but disappear.
The seasonal nature of vegetation changes and the dependence of moisture contrasts on recent rainfall necessitate examination of aerial photographs from as many different flights (ideally from different years) as practical if a reliable fault map is to be prepared. Field examination of each linear feature tentatively identified as a fault from aerial photographs is essential—photos provide a powerful but not sufficient tool for positive fault identification. Repeated visits to the field are advisable, and at least one such visit should take place in winter or early spring, after a heavy rain, when ponded water and minimal vegetation may enhance the visibility of subtle fault scarps.
Identification of faults in undeveloped areas is complicated by the fact that no associated zone of damage serves to document recent movement and thus confirm the identity of the mapped feature as a fault. Topographic features locally resembling fault scarps form by a number of natural processes, and care must be taken not to interpret such features as faults. Presence or absence of a fault may, under favorable circumstances, be confirmed by well-log or seismic data, but for most areas this information is not available for the shallow subsurface (upper 1,000 m). It thus seems pertinent to review the evidence for the two features shown on map 1.
The Clodine fault appears at the land surface as a 12-km-long scarp which extends east-northeast from Figure Four Lake to 2 km east of State Highway 6. Parts of the scarp have been modified recently by grading for new homesites and by plowing for crops, but fieldwork in 1975 confirmed the existence of a well-defined scarp and associated vegetation change in these areas. A scarp 0.22 m in height, for example, was measured by a transit-and-rod survey in a field near Farm-to-Market Road 1464 prior to recent plowing. Shrubs and prairie grasses grew in profusion on the upthrown side of the fault, whereas grasses more typical of wetlands dominated the muddy downthrown side. The fault was then clearly visible on every aerial photograph examined of this area. The field relationships have since been modified, and now only a gentler slope exists where a well-defined scarp once stood.
Despite such modifications of the natural landscape by man, good morphological evidence for the fault remains. After a heavy rain, nearly the entire 12-km length of the downthrown block adjacent to the fault is flooded (fig. 6). A well-defined scarp may still be seen along parts of the fault that remain relatively undisturbed, and even where the scarp has been modified by man, the land immediately north of the fault is visibly lower than that south of it by about 0.3 in. More conclusive evidence of a fault origin, however, is seen on State Prison Farm lands where point-bar* deposits near Oyster Creek are offset along the fault (map 1, fig. 6). Despite the inability of confirming subsurface data, little doubt exists that this feature is a fault scarp, and that the fault has moved in the geologically recent past (the last 30,000–40,000 years).
An origin by faulting for the shorter Renn scarp (map 1) is less definite. A large (about 1 m high) scarp may be observed along much of its length, but the scarp is highly irregular and appears to be a composite feature made up of only roughly coincident sections. The scarp at some locations is steep and well-defined; at other locations two scarps of variable sharpness are seen, and at still other locations only a very gentle slope separates two blocks of land at different elevations. These attributes are quite unlike most known fault scarps in the Houston area; they favor instead an origin by erosion. The rectilinear trace and identical surface sediments on both sides of the feature, however, are more consistent with faulting. The feature is, perhaps, an old fault scarp whose slopes have been modified by subsequent local erosion; we tentatively regard it as a possible fault. Additional work (by drilling, or trenching across the scarp) will be needed to further define this feature.
Each feature shown on maps 1 and 2 was examined along nearly the entire mappable length of its scarp, both on the ground an from a helicopter. Repeated visits to the field were made to document seasonal vegetation and moisture changes, and much field work was carried out in winter, when minimal vegetation maximized visibility of topographic features.
All faults mapped in the field were plotted initially on 1:l2,000-scale aerial photographs and later reduced to the publication scale of 1:24,000.
Effective and proper use of these maps is in part contingent upon awareness of their limitations. This report should be viewed only as a preliminary study of faulting; more detailed site-specific studies are warranted in areas where faulting could prove detrimental to anticipated land use.
Four of the faults shown on map 2—the Long Point, Piney Point East, Eureka Heights, and Memorial Park faults—have been mentioned previously in the geologic literature (for example, Weaver and Sheets, 1962; McClelland Engineers, 1966; Van Siclen, 1967). The other faults on maps 1 and 2 were identified during the course of a reconnaissance study of faults in the Texas Gulf Coast. Only a limited attempt was made by us to identify additional faults; time constraints precluded systematic exploration of the entire map area. It is thus probable that other, perhaps active, faults are present. Blank areas on the map are most appropriately regarded as insufficiently mapped rather than unfaulted.
A further complication in preparing this report and similar fault maps is that much of the evidence a geologist normally uses to map surface traces of faults in the Gulf Coast has been destroyed in developed areas. Only the most active and damaging faults, or faults whose scarps are of substantial height, are likely to be noticed during mapping of developed areas. Mapping of faults is most difficult in areas of recent construction, where even rapidly moving faults may not have had sufficient time to cause substantial damage. Faults in urban areas are thus likely to be always underrepresented on maps, regardless of the time available to map them. For those active faults that are found, however, the trace of the fault can commonly be determined quite accurately from the associated zone of damage.
Wherever possible, faults on map are plotted to an accuracy within 1 mm, a distance that corresponds to 24 m on the ground. It is thus difficult to determine from the map alone whether a fault goes through or merely lies near a particular site, except for those parts of the map where individual buildings are shown and where we can, therefore, depict faults in true relation to all cultural features. In so doing, however, minor inaccuracies are unavoidably introduced. The Eureka Heights fault, for example, is shown on the map in proper relation to all access ramps at the I–10 and 610 Loop interchange, but the location of one of these ramps, from the 610 Loop northbound to I–10 eastbound, is printed inaccurately on our base map. This introduced a slight bend into the mapped trace of the fault that is not physically present on the ground. Additional errors of this type may be present.
One other aspect of mapping faults deserves special attention. In developed areas (map 2), two factors combine to facilitate accurate mapping of a fault: the fault can be located quite accurately at many points where it has damaged buildings, roads, and other manmade structures; and each such location is readily identifiable on aerial photographs and maps. Neither fact holds true for the open areas that dominate map 1. In these and similar fields, one sees only a fault scarp. The precise position of the fault itself, relative to the scarp, is somewhat in question, and the degree of positional certainty drops as man modifies the natural landscape by plowing or by grading and smoothing the land in rice fields (fig. 5). Positional accuracy is assured only to the extent that the present landscape resembles its original form. In addition, even where a well-defined scarp may be followed readily in the field, the common lack of nearby control points from which one can determine exact map positions deters accurate portrayal of the scarp on maps. Unless the position of a scarp is apparent from aerial photographs, because of a visible change in vegetation or soil tone from the upthrown to the downthrown side, the geologist must simply estimate the position on the map. These uncertainties affect, to varying degrees, the two features depicted on map 1. Locational uncertainties were reduced by examination of numerous sets of aerial photographs, in conjunction with careful field inspection of both scarps, but even so the mapped traces of the Clodine fault and the Renn scarp (map 1) do not have the accuracy of faults shown on map 2. We therefore stress again that this report, by its very nature, depicts faults in considerable, but limited, detail. For site-specific studies where the location of a fault must be known to within the nearest few meters, subsurface data must commonly be acquired and used in conjunction with surface evidence to pinpoint a fault, and the fault’s location must then be surveyed to required standards of accuracy.
A final and serious limitation to any fault map of the Houston area is the inability of geologists to determine, on the basis of current fault activity, which faults will be active in the future.
1 Not until after the base map had been printed was it apparent that the Pecore fault extended westward into the area of map 2. A map of the eastern part of this fault is included here as figure 3.
Discussion of Faults in Houston Area | Contents