Rockslide at Yosemite National Park
Rockslide at Yosemite National Park, California kills one, injures 4
By Gerald F. Wieczorek, U.S. Geological Survey and James B. Snyder, National Park Service
WARNING: Provisional Report, Subject to Revision
Photograph taken from Royal Arches Route climb from across the canyon by rock climber Lloyd DeForrest while dangling on the rope 2000 feet above the valley floor. Click image for a larger image (315kb)
Latest Report - Updated July 19, 1999
Summary
A series of rock falls from releases at about 5500 ft elevation below Glacier Point and the subsequent development of cracks in an adjacent rock mass has worried geologists concerned with the likely possibility of another rock fall that could impact part of Camp Curry. Fresh hairline cracks were observed from helicopter and telescope that have increased in length, size, and number since June 14. Cracks which were at first hairline-size features, subsequently opened to several inches across as they also extended in length. The propagation of cracks has been accompanied by occasional audible rock popping and very small rock falls; some rocks have fallen from locations near fresh cracks. The process of crack formation began to show signs of slowing on June 24. There have been only very small and occasional rock falls and very limited additional cracking reported since June 26, although a slight increase in rock-fall activity occurred between July 8 and July 14. Two small rock falls on July 14 followed an afternoon thunderstorm on July 13; this was the first rainfall in the valley since the June 13 rock-fall event raising the possibility that high pore water pressure developed in closed joints may have played a triggering role. This slowing of crack propagation does not necessarily indicate that the site has stabilized; with its present cracks, the rock formation has been significantly weakened and can continue to be affected simply by gravity. The weakness of the formation has made it much more susceptible to rock-fall failure from a trigger such as rainstorm raising pore water pressures in joints, earthquake, freeze-thaw, ambient temperature changes or snowmelt that are likely to be experienced within the next year. The following discussion will focus on geologic factors affecting the instability of this site.
Geologic Structure
Figure 1 Site of rock-fall releases below Glacier Point (blue-November 16, 1998; brown-May 25, 1999; red-June 13, 1999; and yellow-June 15, 1999). Dashed line shows area of figure 2c. Photograph taken looking south from helicopter on June 16, 1999.
The rock-fall release area (fig. 1) at about 5500 ft elevation is slightly above the glacial trim line of Matthes (1930) for the Tioga glaciation where the rocks above this line have been exposed to weathering for more than the last 1 million years. The arch-like rock mass involved in this rock-fall event was composed of granodioritic and tonalitic rocks from a unit included in the Sentinel Granodiorite by Calkins (in Matthes, 1930), later mapped as the granodiorite of Glacier Point by D.L. Peck (written commun., 1997). Below the source area this granodiorite overlies the Half Dome Granodiorite along a sharp contact that dips steeply west.
Building upon the foundation of rock mechanics, the study of rock fracture mechanics has in the past several decades made significant progress in understanding and analyzing different modes of fracture and the propagation of cracks. However, the role of discontinuities and other details of geologic structure and their influence on crack propagation and slope stability remains a complex and poorly understood subject, not presently amenable to sophisticated analyses developed for more homogeneous rock masses. The following statement by Whittaker et al. (1992) summarizes the present state of understanding of rock fracture mechanics in discontinuous jointed rock masses:
Rock excavating activities can alter the rock slope equilibrium sufficient to cause stress redistribution, stress and energy magnification, discontinuity dislocation and the generation of new fractures. Stable or unstable fracture propagation may occur, leading to the failure of excavated rock structures and slopes. However, the fracture mechanism of a discontinuous jointed rock mass is generally not clearly understood.
Figure 2 Photomosaic of releases and area of cracking below Glacier Point. Photographs taken June 16, 1999 by John Weller, Ansel Adams Gallery and used with permission of the National Park Service.
Figure 2a Delineation of joint sets (J2-J5); estimates of orientation for joint sets are included in Table 1. Joint set J1 is the plane of the cliff face visible in this photo.
Figure 2b Photomosaic of releases and area of cracking below Glacier Point. Cracks are shown by red dashed lines. Some selected joints and other structural features are identified by solid red lines for reference. Window notation refers to area of inset exfoliation sheet. Note water seeps over surface of exfoliation sheet J1 issuing from joint set J2 (left).
Figure 2c Sketch showing progressive propagation of cracks with time in area below rock-fall releases.
In addition to the plane of the steeply dipping exfoliation joints (J1), the site of the rock-fall releases and developing cracks is traversed by five other joint sets (J2-J6) (fig. 2a). Characterization of the orientation of joint sets (Table 1) is approximate because of the inaccessibility of the site; estimates of the orientations of joint planes were made from stereo photographs made from helicopter, camera with a telescopic lens, and from a Questar telescope.
The sheet or exfoliation joint set J1 closely parallels the cliff face. Sheet structures typically form in environments of high differential stress, particularly upon vertical unloading of a rock mass that formed at depth under high triaxial compression and is now exposed at the surface due to uplift and erosion. Individual exfoliation sheets thicken downward (or into) the rock mass and sheets tend to be thinner in fine-grained rocks than in coarse-grained rocks (Holzhausen, 1989). Gilbert (1904) noted that sheet structure in the domes of the Sierra Nevada tends to parallel all topographic surfaces and that the separation of sheets penetrated to depths between 50 and 100 feet perpendicular to the surface. At the site of this rock-fall release the thickness of the most exfoliation sheets visible on the cliff surface is thin, probably ranging from 3 to 6 ft thick.
The most prominent joint set J2 is a pervasive ledge forming discontinuity which dips steeply to the east and is oriented similar to the set of discontinuities forming the Staircase Falls and the abandoned "Ledge Trail" beginning behind Curry Village and proceeding up a prominent ledge towards Glacier Point. In the release area, this discontinuity is smooth, regularly closely spaced, and planar. The "roof" of the November 16, 1998 release forms a crude arch with the intersection of J2, J5, and J6 (fig. 2a). This pattern of "roof" approximates a rough joint-defined arch which is repeated in many place on the cliff of Glacier Point, most closely about 150 feet to the west of the recent releases, although the dark-stained rock face indicates that this collapse did not occur recently.
Inspection of the cliffs below Glacier Point in the vicinity of the release and stereoscopic analysis of the joint set orientations (Table 1) using the ROCKPACK II software package (Watts, 1994) indicates that none of these joint planes or joint planes intersections form plane or wedge conditions favorable to sliding. The lack of plane or wedge conditions favorable to sliding on the north face of Glacier Point is due to the orientation of the joint sets (J2-J5) which dip either due east or west, perpendicular to the cliff face which dips due north. These conditions do not apply to the east face of Glacier Point, where abundant potential plane and wedge failures exist due to joint sets and east dipping slope of Glacier Point (Gilliam, 1998).
The principal role of the joints and their intersections at the release points on the north face of Glacier Point is to define the top and lateral boundaries of the rock-fall releases of exfoliation sheets that split along these joints and their intersections. Thus joint sets (J2-J6) on the north face of Glacier Point do not form the surface(s) along which sliding occurs, but determine the size of exfoliation sheets that fail.
Table 1 - Joint sets and their orientations from near the rock-fall release area. Estimates of dip direction and dip amount from stereo photos are approximate due to inaccessibility of site.
With exfoliation joint set J1, closely paralleling the very steep slope face, a slight variation in the degree of dip of either the exfoliation sheet or the slope face could affect the distribution of stresses and the local stability of the cliff. If the exfoliation sheets dip less steeply than the cliff face, then the exfoliation sheet tends to "daylight" in the cliff face, a condition favoring sliding along exfoliation joints. If exfoliation joints dip more steeply than the cliff face, then toppling would be the favored mode of failure. If the exfoliation sheets parallel the face, then neither sliding nor toppling are favored, but a vertical load would tend to buckle, open and separate the exfoliation sheets. Extension of joints may also occur due to concentration of loads at the tip of joints. The relatively thin (3-6 ft-thick) nature of the sheets, makes flexural bending and opening between sheets more likely as the height of exfoliation sheet increases. The maximum height of the exfoliation sheet forming the November 16, 1998 collapse was about 50 ft, approximately a 10:1 height to thickness ratio. Generally it is not possible to tell the extent to which the sheets are attached and maintain connections along their backsides. The extent to which the sheets are open not only affects the strength of the sheets and local stability, but also influences the patterns of groundwater flow through the jointed and fractured system.
Pattern of Cracks
The pattern and timing of cracks that developed below the rock-fall releases may indicate something about the present stability of the mass, and the mechanism of how it will eventually fail. The cracks are restricted to a roughly rectangular area of about 60 by 75 feet (fig. 2b). The distribution and progressive pattern of the cracks with time are shown in fig. 2c. The cracks appear to be of three apparent types: 1) shear, and 2) tension cracks exposed in the plane of the exfoliation sheet that are in the process of splitting the top exfoliation sheet, and 3) parting or opening cracks that indicate separation between successive exfoliation sheets in the plane perpendicular to the slope face. These cracks are concentrated in an eastern and western section with a center section devoid of cracks. This center section is an exposure of an exfoliation plate extending behind the two sections to the east and west as an apparent "window" into the deeper exfoliation sheet (Fig. 2b).
Using the length of new shear and tension cracks detected since June 14, we compared the percentage length of new cracks formed during each 48-hr period. This rate of crack propagation varied widely although some of the variation could be attributed to the selected time intervals and to observational difficulties. Initially, the rate of crack propagation rate was high during the June 13-14 period, but gradually decreased during the following week. The rate again increased dramatically during the period of June 20-22. Between June 22 and 26 the rate of new crack formation notably decreased. Subsequently, only a few very short additional cracks have been detected, although a spike of small rock-fall activity and a few short cracks occurred between July 8 and 14.
The questions of how the cracks formed and how to interpret the pattern and timing of cracking are difficult to answer due in part due to the inaccessibility of the site and the complexity of the jointed rock mass. One possibility to explain the initiation of cracking, suggested by small fragments of rock missing from the face, is that impacts from the June 13 and June 15 rock falls hit the area below the release and initiated cracking. Another possibility is that changes in the loading conditions at this site caused the initiation and propagation of cracking. The large previous rock falls, particularly the events of November 16, 1998 and June 13, 1999, changed the geometry and structure of the site, probably causing a redistribution and concentration of stresses. Along a steep nearly planar rock face near the releases, the primary load can be assumed to be principally vertical and the stresses in the rock mass chiefly compressional. The distribution of vertical stress on the face would be concentrated at the points above and below the previous releases because of the changes in geometry of the cliff face with the removal of material caused by the rock falls. Compressive forces acting on the relatively thin (3-6 ft) exfoliating slabs could initiate shear cracking and possibly failure. The orientation of many of the cracks at a steep angle of about 60-70 degrees to the horizontal, not aligned with any joint set, is indicative of a pattern of shear under high confining stress. High shearing stresses could induce small movements that could have generated small rock falls, by forcing small fragments from the face. Popping and other sounds similar to rock bursts have been noted from the site during the development and extension of cracks.
The rate of cracking and connection of individual cracks leading to failure will depend upon the fracture toughness of the rock mass. The dilation brought about by movement along cracks in some materials causes a slight increase in friction angle, strengthening the material. Under the forces of gravity and other external triggers, movement and interconnection of the cracks may continue at the present very slow rate (creep) or resume at a rapid rate. A more thorough analysis of fracture toughness and of the conditions necessary to cause failure is beyond present capability. Measuring the properties of this remote site and modeling site complexity to determine the relative stability of the rock face require direct contact with the near vertical face.
Groundwater Cleft Pressures
The presence of water issuing from the release areas following the June 13, 1999 rock fall suggests that groundwater flowing along joint surfaces could have increased pore water pressures (cleft pressures) and further weakened the mass triggering the rock failure. Some of these seeps dried up in the days following rock-fall events, further suggesting that water within the joints had been backed up and then released by the failure; once released the flow quantity of water quickly decreased. Whereas most of the seeps issued over the exfoliation sheets (J1), the water appeared to be emerging from the exfoliation of joint set J1 with J2, the thorough going main ledge-forming joint set that proceeds towards Glacier Point. The infiltration of groundwater from spring snowmelt along the complex pattern of joints to the release area would be slow perhaps taking weeks or months. According to the 1:24,000 Half Dome topographic map and the orientation of joint surfaces, no well defined catchment area near Glacier Point feeds this rock-fall release area. The peak of runoff of the Merced River in late May and early June suggests that the two rock falls of May 25, and June 13, 1999 were possibly related to the buildup of cleft pressures from infiltration and fracture flow from spring snowmelt, but it is not possible to substantiate the timing nor the magnitude of the cleft pressures. The timing of the initial November 16, 1998 rock fall does not fit the pattern of influence from spring snowmelt; however, it does fit the pattern of freeze-thaw in early winter storms.
An afternoon thunderstorm in Yosemite Valley on July 13 caused noticeable changes in the locations and amounts of water flowing from joints at the site in the subsequent days (July 14-15). Quite possibly two small rock falls on July 14 (9:57 am and 8:20 pm) and an additional short crack noted on July 14 are attributable to the buildup of groundwater cleft pressures in joints caused by the infiltration and flow of water through the jointed rock mass from the thunderstorm. The triggering of a rock fall in Yosemite by an increase in cleft water pressure in a joint during an intense rainstorm has been previously documented (Wieczorek and Jäger, 1996, p. 23).
Acknowledgements
The U.S. Geological Survey acknowledges the cooperation and financial assistance of the National Park Service in the investigations of rock falls inYosemite National Park. The coauthors are also indebted to those who voluntarily observed and noted rock-fall events at Curry Village. Likewise, we are grateful for the graphical preparation of figures by Chris Catherman of the U.S. Geological Survey.
References
Gilbert, G.K., 1904, Domes and dome structures of the High Sierra, Geological Society of America, v. 15, pp. 29-36.
Gilliam, D.R., 1998, A structural and mechanical analysis of the Happy Isles rockfall, July 10, 1996, Yosemite National Park, Mariposa County, California [Master's thesis]: Radford, Radford University, 172 p.
Holzhausen, G.R., 1989, Origin of sheet structure, 1. Morphology and boundary conditions: Engineering Geology, v. 27, pp. 225-278.
Matthes, F.E., 1930, Geologic history of the Yosemite Valley: U.S. Geological Survey Professional Paper 160, 137 p.
Watts, C.F., 1994, ROCKPACK II, ROCK slope stability computerized analysis PACK, User's Manual, 79 p.
Whittaker, B.N., Singh, R.N., and Sun, G., 1992, Rock Fracture Mechanics--Principles, Design and Application, Developments in Geotechnical Engineering, 71, Elsevier, 570 p.
Wieczorek, G.F., and Jäger, Stefan, 1996, Triggering mechanisms and depositional rates of postglacial slope-movement processes in the Yosemite Valley, California, Geomorphology, v. 15, p. 17-31.
Contact Information
| Gerald Wieczorek U.S. Geological Survey National Center MS 955 Reston, VA 20192 gwieczor@usgs.gov |
Report from June 21 & 23 and July 6 1999
A series of rock-fall events continue to threaten the Curry Village area of the Yosemite Valley. The release site of the rock falls is on the north-facing wall below Glacier Point from an elevation of about 5500 ft.
Figure 1 Photograph looking south at rock-fall release areas below Glacier Point (dashed colored lines indicate individual release areas: blue-11/16/98, brown-5/25/99, red- 6/13/99, yellow 6/15/99). Dark shaded areas indicate water seeps from joints and cracks. Photograph slightly oblique to rock face with approximate scale as indicated.
No previous historic rock falls have been recorded from the area of this release. The release is slightly above the glacial trim line of Matthes (1930) for the Tioga glaciation where the rocks have been exposed to weathering for more than the last 1 million years. Curry Village was established in 1899 and eventually extended upslope onto the two talus cones that are the result of prehistoric rock falls. Part of the charm of Curry Village is the presence of huge boulders, the result of prehistoric rock fall(s). The currently active release is above the eastern of these two talus cones that in its lower portion contains tent cabins used for seasonal housing for employees.
The first rock fall in this sequence occurred on November 16, 1998 and had a volume of about 300 yd3 (700 tons). The block(s) fell about 150 ft to a ledge breaking up against the cliff face in the process. The rock further broke up on hitting the top of the talus. The size and velocity of the rocks were sufficient to knock over and clear large trees from the upper part of the talus. The rocks continued to bounce, roll, and slide down the talus, their path determined by the distribution of larger boulders from prehistoric rock falls and large trees which either stopped or directed their paths. The rocks that traveled furthest, about 1600 ft from the top of the talus, took a NE direction.
Figure 2a Map of recent rock-fall impact zone near Camp Curry in the Yosemite Valley. Outlines of impact of 11/16/98 (blue), 5/25/99 (brown), and 6/13/99 (red) rock-fall events. Map does not include locations of all structures within Curry Village.
Rocks reached within about 75 ft of the closest cabin, highest on the talus.
Smaller mostly fist-sized pieces of fresh rock were found considerably beyond the limit of where the larger rock blocks stopped.
Figure 2b Map of recent rock-fall fly rock zone near Camp Curry in the Yosemite Valley. Outlines of fly rock from 11/16/98 (blue), 5/25/99 (brown), and 6/13/99 (red) rock-fall events. Map does not include locations of all structures within Curry Village.
The paths of the larger rock blocks could be traced through the forest whereas the small pieces were presumeably airborne projectiles from points of impact high on the talus. These small pieces were presumably airborne projectiles from points of impact high on the talus. A few of these airborne splatter pieces reached the tent cabins and, falling at steep angles, pierced the canvas tops, broke beams, and fell to the floor. One of the larger projectiles, approximately the size of a football, pierced the canvas roof of a cabin. Although large dustclouds were generated, no airblasts were observed from any of these rock falls, presumably because the rock blocks broke into moderate-sized pieces in impacts along the cliff by the time they reached the talus.
At 9:12 am on May 25, 1999, a much smaller (7 yd3, 16 tons) rock fall occurred from the same general release area that reached the talus. The pattern of distribution of rock onto the talus was similar to that of the November 16, 1998 event, but covering a smaller area. Rock did not travel nearly as far, the farthest rocks traveling about 500 ft. Similarly the zone of airblast splatter was confined to a smaller area on the talus. No well-defined trigger for the May 25 event could be assigned, for there had been only a trace of rain several days before and no observed seismicity in the Yosemite Valley.
On June 13, 1999, an intermediate-sized rock fall of about 225 yd3 (525 tons) occurred that killed one climber and injured several others who were climbing along a route beginning at the top of the eastern talus cone at Curry Village, immediately below the release area. The travel of rocks along the talus and airborne splatter was again very similar to the previous observed distributions, extending to nearly the limit of the November 16, 1998 event (fig. ). Following this event, the NPS (Jim Snyder) and USGS (Gerald Wieczorek) began to monitor the release area by helicopter and on the ground using telescope for development of cracks, new failures, patterns of seepage, and other changes possibly related to inherent instability. Comparing observation on the mornings of June 15 and June 16, we noted a series of new extensional fractures and fine hairline cracks in fresh rock that encompassed an area of potential instability of approximately 2500 ft2. These new fractures apparently accompanied a very small rock fall (4 yd3) from the same release area that was heard about 10:20 p.m. (June 15) throughout the Yosemite Valley. From the exposed thickness of exfoliation sheets at this site, the thickness of the instability could range from 3 to 6 ft thick. The pattern of developing cracks indicated that a piece, approximately 1/3 of the total area, would be likely to fail before the remaining mass. If the entire area of potential instability failed as one single piece (at the same time), assuming a 3-ft thick section, then the resulting volume (280 yd3), would nearly equal that of the November 16, 1998 event; if the section were thicker, the volume (and probably the travel distances of rocks and flyrock splatter) could exceed the previously observed rock-fall events at this site. Consequently, on the afternoon of June 16, temporary evacuation of a larger portion of the employee housing area (Terrace) of Curry Village was recommended to the Superintendent of Yosemite National Park. Immediate actions secured the perimeter of the area removing people from tent cabins within the approximate boundary of the flyrock splatter zone of the November 16, 1998 event.
Coincident with the evacuation and cordoning off at Curry Village, a program of monitoring, observation and documentation of the rock-fall release area was initiated from Stoneman Meadow. Observers would note the time, length, and sound of rock falls in an attempt to relate these occurrences to events on the rock face. Periodic observations and photographs would be taken of the release area with high powered cameras and telescopes to document the changes in the slope face. At 5:40 p.m. on June 17 another small rock fall occurred which again was correlated with further extension of fissures and new hairline cracks observed subsequently. There were some crack extensions noted on June 18, but no accompanying rock fall. Observations from helicopter on June 19 confirmed that not only were new cracks appearing, but those that were noted on the previous days were judged to be enlarging. In the evening of June 19, sounds of popping, presumed to be the sound of rock cracking, were heard. Similar sounds were noted a day prior to a rock fall along the Upper Yosemite Falls trail in November of 1980. Although no additional popping sounds or rock falls had occurred before observation on the morning of 6/21, three additional new cracks and crack extensions compared with observations of 6/20.
Update to report — June 23, 1999
Two new fine hairline cracks were observed on the underside of the existing roof of the 11/16/99 failure on the evening of 6/21. At about midnight of 6/21 a popping sound like that of a shotgun was heard in Curry Village indicating that brittle rock fracturing was occurring at the rock-fall release area. At about 10 pm on 6/22 and at 6:13am on 6/23 observers heard small rock falls from the release area.
Report — June 16, 1999
USGS geologist Gerry Wieczorek performed two helicopter flights over the source area of the Yosemite rockfall site. As a result of the first flight, on June 15th, Gerry reported no new cracking in the source area. However, since a new rockfall occurred during the night of June 15th, a second reonnaissance flight today, June 16, revealed large extensional cracks in the source area, with the certainty that there will be another rockfall--probably quite large. Gerry has briefed park officials as to the severity of the situation. A new map of the rockfall hazard is in preparation and will be published soon, for public use. The map will be available through the USGS publications office, at 1-888-ASK-USGS.
The USGS has published a previous report on Yosemite Rockfalls which can currently be obtained from USGS Publications: "Rock-fall Hazards in the Yosemite Valley" by Gerald F. Wieczorek, Meghan M. Morrissey, Giulio Lovine and Jonathan Godt, USGS Open-file report 98-467 (published in 1998).
The Yosemite Rockfall report is published in its entirety on our website under "Landslide Program Publications - Open file reports" on this website and can be directly accessed by following this link: Open-file report 98-467
Initial Report — June 13, 1999
A man was killed by falling rocks while rock climbing at Yosemite National Park, California, Sunday, June 13, 1999. The accident occurred at Glacier Point at the eastern end of the park. Huge boulders reportedly broke loose and killed the victim while he was still holding onto his rope. Gerry Wieczorek, a geogist/engineer from the US Geological Survey, (Reston, Virginia office) will be arriving Tuesday, June 15, to assess the situation. Dr. Wieczorek has studied the Yosemite National Park area extensively, and has published reports on rockfalls in that area.
Contact Information
| Lynn Highland National Landslide Information Center U.S. Geological Survey MS 966, Box 25046 Denver Federal Center Denver, CO 80225 U.S.A. highland@usgs.gov |
Gerald Wieczorek U.S. Geological Survey National Center MS 955 Reston, VA 20192 gwieczor@usgs.gov |
The USGS is prohibited from reproducing copyrighted news articles on its web pages — as a result, we generally report landslides of significant magnitude or those that our research team has investigated. Since we do not receive automatic reports of landslides everywhere in the world, it is suggested that those searching for recent landslides, also engage the use of local newspapers, most of which are readily accessible on the internet.
