USGS

PREVIOUS

Dilute streamflow: Flows in which the sediment load does not affect flow behavior, or imparts no yield strength to the flow, are considered as normal streamflow (Pierson and Costa, 1987). Sediment concentrations up to 50% by volume for mixtures of coarse particles of uniform size and up to 35% by volume for more poorly sorted mixtures impart no yield strength to flowing water (Pierson and Scott, 1985). Turbulence is the primary mechanism for sediment transport in such flows (Smith, 1986). The conditions of sediment transport and deposition are controlled by a complex set of variables, including flow velocity and depth, and channel configuration. Streamflow results in deposits generally associated with flooding and water transport, may range in size from boulders to silt, and generally exhibit a high degree of sorting. High-energy streamflow can result in deposits of open framework boulder and cobble bars that contain very little matrix. These bars occur as high-relief trains of well sorted and often imbricated clasts.

Hyperconcentrated flow: Sediment-water flows in which the concentration, size distribution, and/or composition of the entrained sediment lead to a measurable, but low, yield strengths have been described as hyperconcentrated flow (Pierson and Costa, 1987). Intermediate ranges of sediment concentration and low to moderate silt and clay contents result in generally low yield strengths. Hyperconcentrated flow is usually used to refer to non-Newtonian flows with almost no strength that produce deposits that are intermediate in nature between those of streamflow and debris flow (Smith and Lowe, 1991). Turbulence, particle dispersive forces, and buoyancy all potentially contribute to particle support. Deposition occurs by particles dropping out of the flow as individual grains, and the remaining fluid continues to move.

Hyperconcentrated flow results in deposits with particles in contact with each other (i.e., clast supported), and that show some sorting and gradation, depending on the velocity and depth of the flow at the time of deposition. Deposits can exhibit some weak horizontal stratification and poor sorting, but are generally better sorted and have lower silt and clay contents than debris-flow deposits (Wells and Harvey, 1987). Coarser facies can exhibit horizontal clast orientations and partial filling of interstices by sand or silt matrix material.

Debris flow: Debris flow is characterized by a substantial yield strength and plastic behavior, yet the fluid retains at least partially liquid properties (i.e., it will spontaneously assume the shape of its container). The onset of debris flow in sediment-water mixtures is defined by Pierson and Costa (1987) to occur at the point where the yield strength increases rapidly with increasing sediment concentration due to internal friction that arises from interlocking of grains. In contrast with hyperconcentrated flow (where the particles are deposited as individual grains from suspension and remaining fluid continues to move), in debris flow the sediment-water mixture moves as a single phase, and only the largest particles may fall out of suspension.

Debris-flow deposits are characterized by significant relief and sharp, well-defined flow boundaries. Levees lining the flow path, or a veneer of mud coating the channel sidewalls, as well as steep, lobate deposits of matrix-supported material at the path terminus are characteristic of this flow process. Debris-flow deposits show little, if any, internal stratification and may contain gravel-sized and larger particles supported in a fine-grained matrix. Although for the entire deposit clay contents may only be a few percent of the overall size-distribution (Costa, 1984), clay-sized material often comprises a significant part of the matrix material (e.g., 10-15% of the <2mm fraction, Wells and Harvey, 1987). Matrix material with a significant silt component (30-35% of <2mm) can also result in debris flow (Pierson and Costa, 1987). Vesicles formed by trapped air may be abundant in debris-flow matrix, but are rare in fine sediments deposited by other flow processes (Sharp and Nobles, 1953).

The matrix support of larger clasts and significant clay or silt component of the matrix are particularly useful diagnostic characteristics for recent debris-flow deposits. They are, however, less useful in older deposits; matrix material may be lost with time (Costa, 1984; Blair and McPherson, 1992), or interstices can be filled in with less clay-rich material. Further, because debris flows form relatively high relief deposits that fill channels and depressions, and because debris-flow events include later, more dilute flows, debris-flow deposits are prone to reworking by streamflow processes, thus obscuring their debris-flow origin (Costa, 1984; Wells and Harvey, 1987).

Geological, Ecological and Climatological Frameworks

Geology: As much as 400 m of horizontal to slightly dipping Miocene, Pliocene, and lower Quaternary- age rocks make up the sidewalls of Capulin Canyon and the flanks of Boundary Peak. Principal units, from oldest to youngest, include sedimentary rocks of the Eocene Galisteo Formation, sandstones and siltstones of the Miocene Sante Fe Group, the Canovas Canyon Rhyolite and Tuffs, volcaniclastic rocks of the Cochiti Formation, volcanic rocks of the Paliza Canyon Formation, and the Lower and Upper Members of the Bandelier Tuff (Goff, et al., 1990). The Lower Member of the Bandelier Tuff, which was erupted from the Jemez volcanic field about 1.6 Ma (Izett and Obradovich, 1994), is exposed in the base of the canyon, while the Upper Member makes up the steep cliffs that bound the canyon. The Upper Member of the Bandelier Tuff was erupted from the Jemez Mountains about 1.2 Ma (Izett and Obradovich, 1994). Mapping by Smith et al., (1970) shows an exposure of the Sante Fe Group near the toe of a large Quaternary-aged landslide deposit on the northeast flank of Boundary Peak (Figure 3), and exposures of the Paliza Canyon olivine basalt were observed upstream from the Quaternary-landslide deposits. Further, patches of pumice of the El Cajete Member from the youngest eruptions from the Valles caldera about 50 to 60 ka (Reneau, et al., 1996) mantle portions of the hillslopes in the canyon. The nearly north-south trending Pajarito fault cuts across the canyon approximately 0.5 km south of Base Camp (Figure 3). What appears to be brick red Galisteo sandstones and siltstones are exposed on the down thrown block of this fault on the west canyon wall.

Ecology: For this report we consider three major plant communities within Capulin Canyon: pinon-juniper, which is dominated by a pinon pine and one-seed juniper overstory with a grass/herb, shrub understory; Ponderosa pine, which is dominated by a Ponderosa overstory with a variety of understories depending on stand density and recent fire history; and a riparian vegetation type that includes elements from adjacent slopes along with species requiring enhanced moisture regimes (U.S. Department of Interior, 1996).

The fire history of Capulin Canyon has been reconstructed by dating fire-scar material on trees by Touchan et al., 1996. This record indicates that for the period 1664-1893, surface fires swept the area about every 6.8 years. These frequent fires could be widespread, but generally did not exhibit high burn intensities over extensive areas. In the years subsequent to 1893, fires were infrequent and not as widespread, allowing for the accumulation of significant fuel loads. The change in fire frequency has been attributed to intense grazing by large numbers of free-ranging livestock which reduced the grassy fuels through which most fire spread (Allen, 1989, Swetnam and Baisan, in press; Touchan et al., in press). Active fire suppression practices during this century has allowed the buildup of unnaturally high densities of trees and amounts of ground fuels that were formerly thinned by frequent surface fires (Allen, et al., 1996). In contrast with the historical record, the Dome fire exhibited high burn intensities over an extensive area.

Climate: The climate of the Jemez Mountains and Bandelier National Monument is described by Allen (1989). Mean annual precipitation at the Monument weather station is 40.7 cm. Usually, the period from late April through the end of June is dry, followed by the onset of the summer monsoon season. Sixty percent of the annual precipitation falls between June and September, with thunderstorms reported for 58% of the days in July and August. These convectional thunderstorms bring 40% of the total annual precipitation in July and August. During the rest of the year precipitation is generally associated with the passage frontal storms and tends to be less intense (Bowen, 1990). In winter these storms bring snow to all elevations.

Local climate is also temporally variable, with wide fluctuations in annual precipitation common (Allen, 1989). Cyclic El Nino Climate events bring increased spring and summer precipitation to this area about every four years (Andrade and Sellers, 1988).

The June 26, 1996 Storm: The June 26 storm was a convectional thunderstorm and, due to the lack of an extensive rain gage network, its areal extent and rainfall output is unknown. The hydrologic response of Capulin Creek to the rainstorm, however, was observed by a team of archeologists staying at the Base Camp cabin (Figure 3). The team observed dark thunder clouds concentrated at the head of the canyon where it appeared to be raining, although there were only isolated raindrops falling at the cabin. About dusk on the evening of the 26th, a surge of water approximately 2 m high traveled through the stream channel located approximately 3 m from the cabin; the flow barely overtopped the channel bank and ran up to the cabin door. The team members had only a few moments of warning once they heard the flood wave approaching.

The effects of the June storm on hillslopes in the Capulin Creek drainage were observed to concentrate near Boundary Peak and westward (Figure 2). The easternmost extent of evidence of tributary surface flow generated by this storm was observed on July 3 in a small drainage to Capulin Creek shown as location A on Figure 2. The flow was at most 2 m wide and its lateral extent was marked by 15-cm-high levees consisting of rafted ash, charcoal, pumice fragments, and pine needles. The effects of surface overland flow on hillslopes were observed, in general, to increase from this point to the west. The impact on the steep, south flank of Boundary Peak was particularly severe, where cobbles of up to 20 cm in diameter were observed to have been transported short distances down the steep slopes, and up to 3 cm of the mineral soil was removed in places.

Because the storm appeared to be concentrated near Boundary Peak, and its effect on the hillslopes appeared to dissipate toward the east, the proximity to high intensity rainfall was quantified by considering the distance of each of the 15 sites from a north-south line drawn through the Boundary Peak trail head (Figure 3, Appendix A).

The August 19-25, 1996 Storm Sequence: The effects of this storm sequence were again observed by a team of archeologists, this time camping near the junction of Capulin Canyon and the Rio Grande (Earl Ruby, Hydrologist, oral commun, Sept. 1996). Flood crests of 45 cm on August 20 and 94 cm on August 22, 1996 were measured at Base Camp. The flood of August 25 was estimated to be 2 m deep in lower Capulin Canyon and 3 m deep at Base Camp. The archeologists measured an accumulation of 1.27 cm of rain in 29 minutes during the August 25 event, and described dark rain clouds at the head of the canyon, in addition to the cloud cover at their camp. The leading edge of the flood was reported to have reached the camp 45 minutes after the rainfall stopped. This event was described as flood waters that were partially bulked with ash, charcoal and sediment. A rain gage installed at Instrument Site 1 (IS-1, Figure 3) on August 23, 1996 (considerably upslope from the archeologist’s camp), measured only 0.30 cm of rain in an 8 hour period on August 25, 1996 (Figure 4). Note also that an accumulation of 0.30 cm of rainfall was measured on August 27 in a time span of one hour and nine minutes. The effects of this high-intensity rainfall event were not observed by anyone in the canyon.

Remainder of the Monsoon Season:. Summer thunderstorms occurred throughout the summer in the canyon, but a recording rain gage was not installed at Instrument Site 1 (Figure 3) until August 23, 1996. Thus, no record of rainfall accumulations exists until this date. For the balance of the monsoon season, the rain gage at Instrument Site 1 recorded accumulations of 3.58 cm from September 12-16, 0.89 cm on September 18 and 4.22 cm in a very short time period on October 4 (Figure 4). Although no one was in the canyon to personally observe the effects of each of these storms, a visit during November 11-15, 1996 revealed up to 2 m of incision in places in Capulin Canyon. Some reaches that had experienced deposition during the June storm were scoured clean, and other reaches showed considerable deposits of primarily sand, ash, and charcoal. Some large boulder and cobble-sized material had also been moved, and large trees uprooted, by the high-energy summer flooding.

A pending comparison of the summer of 1996 rainfall data with the historic rainfall records from Frijoles Canyon by Jack Veenhuis of the U.S. Geological Survey, Water Resources Division, will indicate the magnitude of severity of these summer storms in the watershed.

Influence of Geologic and Geomorphic Factors

The geomorphic and geologic factors listed in the Introduction have been identified as contributing to the generation of debris flows from hillslopes following fires. The presence of these factors, given an appropriate rainstorm, can indicate a high susceptibility to debris flow. Note that at this time the relative importance of each of the factors is unknown, and methods to quantitatively evaluate their possible effects have not been explored. In the following section, the influence of fire intensity and the geologic and geomorphic factors on the generation of sediment for debris flows in Capulin Canyon, in part evidenced by the hillslope response to the June thunderstorm, are described.

Fire intensity: The intensity of a fire can be characterized by the size of fuels consumed and the completeness of consumption, which is indicated by the color of the remaining ash. For this assessment, the presence of black ash, scorched needles that remain on the trees, and an intact ground cover were taken as indicators of a low fire intensity. Moderate fire intensity was defined by the consumption of materials greater than 0.5 cm and less than 2 cm in diameter, including ground cover material, and the presence of gray or mixed-color ash. The consumption of materials greater than 2 cm in diameter, including nearly all tree needles and ground cover, nearly total tree mortality, and the presence of white or red ash indicates a high fire intensity.

In general, areas affected by high-intensity fires are considered to be a higher erosive risk; higher intensity fires are more conducive to the formation of hydrophobic layers and generation of dry-ravel, and result in more thorough removal of ground-covering and riparian vegetation and fibrous root mat material.

Fire intensity was evaluated at each of the 15 field sites in Capulin Canyon (Figure 3, Appendix A). In general, areas of low fire intensity exhibited very little sediment movement following the June rainstorm. One effect of low intensity fire was to simply scorch the needles on the Ponderosa pine; following the fire, these needles dropped, and could provide a mulch protection to the underlying soil. At most sites the layer of pine needles mantling the surface was only slightly disturbed by surface runoff from the June storm, although in areas of high rainfall intensity, the needles were rafted from the hillslope by surface runoff. Interestingly, a noncontinuous hydrophobic surface was observed in one low intensity site.

Following the June rainstorm, of the four moderate fire intensity sites, two exhibited abundant rills that measured up to 5-cm wide and 4-cm deep, and the removal of approximately 60% of the blackened surface pumice. The other two sites showed only slight rill development in areas not mantled by surface stones. The occurrence of surface erosion at the moderately burned sites can be explained by the close proximity to the high intensity storm cell.

 

Figure 4. Cumulative rainfall record for the period Aug. 23-Nov. 18, 1996,l recorded at Instrument Site 1 (fig.3).

Friable bedrock units and cohesionless soils: The Capulin Creek drainage is underlain by a number of volcanic and sedimentary rock units (Goff, et al., 1990) which show a variety of responses to weathering. Welded ash-flow tuffs of the Upper Bandelier tuff that forms the cliffs and lower sidewalls of the canyon are not decomposing into significant deposits of loose, friable material. The porphyritic dacites and andesites of the Paliza Canyon Formation are also quite competent, and are not contributing substantial colluvial material that could be mobilized into debris flows. According to the 1978 Soil Survey of the Bandelier National Monument (Earth Environmental Consultants), the steeper slopes underlain by these units show almost no soil development, and very cobbly or stony surfaces. In addition, even though the lithic-rich ash-fall and ash-flow tuffs of the Canovas Canyon tuffs could potentially erode into loose, friable material, the aerial extent of this unit in the canyon is limited, and is thus not considered to be a high risk. The El Cajete pumice, however, which blankets some hillslopes in the region, is an extremely erodible unit.

Although the El Cajete Pumice does supply abundant loose, friable material that is potentially available for mobilization into debris flow, and the soil developed on this unit is, in general, noncohesive, observations following the summer of 1996 thunderstorm season indicate the potential for the generation of only minor debris-flow activity. The primary response of hillslopes mantled by the El Cajete pumice to summer thunderstorms was considerable surface overland flow; low-density pumice, ash, charcoal, pine needles, and some mineral soil were rafted down the hillslopes and draws by this process. In some places rills that were at most 4-cm deep and 15-cm wide stripped the burned soil and ash from the hillslope, as shown in Figure 5. When surface water flows passed through an area mantled by abundant pine needles, significant volumes of pine needles were often incorporated. These flows then took on the character of debris flows; the yield strength imparted by the pine needles allowed for the formation of discrete levees lining the flow margins (Figure 6). The levee and lobe deposits lining the draws and rills were, in general, clast supported, with very little silty matrix, and appear to have formed by rafting of the low-density materials along the surface of the flows, rather than being incorporated into a slurry. Note that these debris flows were at most 50-cm wide, and of low strength; Figure 6 shows a debris flow path easily deflected by a fallen log. Further, the passage of the flows served only to clear a path of loose, low-density materials, and no incision into the base of the paths were observed. Such incision would be necessary to generate debris flows of significant size.

NEXT PREVIOUS TOP

 

USA.gov logo