McGee Creek Leveling Line

Latitude: N 37°33'53.9"
Longitude: W 118°47'04.2"

Length: 627.8m
Number of Bench marks: 29

Number of Surveys to 2000: 18
Date of Initial Survey: 9 July 1982
Date of Latest Survey: 30 June 1998

Trespass Permission:  None needed, although the campground host will not permit parking in any of the campsite slots.

Directions:  From Highway 395, drive 2.0 miles up McGee Creek Road to the entrance of USFS campground. The north end of the level line is in the north end of the campground next to prominent white granite boulders next to the road, 75 m east of the campground entrance.  The latitude and longitude given above were taken from a hand-held GPS and are for benchmark 7317 at those boulders.

Comments:  This line crosses the Hilton Creek fault, 18 km ESE of Mammoth Lakes, and its 15 m high scarp in McGee Creek. It was established to monitor movement along the fault which had 10 cm of vertical displacement after the May 1980 Mammoth Lakes earthquakes.   The line was surveyed by UC Davis students under the supervision of Dr. Gerald Bawden in 1997.

Results:  Rarely does surface rupture occur with earthquakes having magnitudes less than 6.0, and the two M 5.1 earthquakes of June 8 and July 14 (Figure 1) provided no exception to this rule of thumb, judging from a recent UCSB resurvey of a leveling array across the Hilton Creek fault, allegedly the causative fault of the two earthquakes.

According to Dr. David Hill, Chief Scientist of the USGS's Long Valley caldera monitoring program (personal communication, 1998), "The M=5.1 events of June 8 and July 14 were both in the footwall of the Hilton Creek fault. The epicenter of the June 8 event was "on" the Hilton Creek fault trace but at a focal depth of 6 km. Its focal mechanism was dominantly strike-slip (right-lateral on the NW striking plane defined by the aftershock sequence). The epicenter of the July 14 event was about 2 km west of the Hilton Creek fault trace (focal depth also about 6 km). Its focal mechanism had a large normal component with planes striking more or less north-south, consistent with the general distribution of aftershocks to the July 14 mainshock."  Judging from the distribution of the epicenters in Figure 1, it seems more likely that the causative fault coincides with the ESE-WNW caldera margin rather than the NNW-striking Hilton Creek fault.

USGS observers are reported to have searched part of the Hilton Creek fault trace for a surface rupture soon after each of the two earthquakes, but in vain. The only surface manifestation of the June 8 earthquake evident to us is a rounded, SUV-sized granite boulder that tumbled into the McGee Creek campground from its precarious perch on the crest of the lateral moraine 100 m west of the campground, leaving a trail of smashed, dead sage brush, impact craters, and shards of other rocks that it hit along the way.

On July 30, 1998, Arthur Sylvester, researcher Jeff Lee, and three graduate students (Ross Hartleb, Kurtis Burmeister, Marcy Davis) resurveyed the leveling line across the Hilton Creek fault that Sylvester established in 1982, two years after the four M 6.0 earthquakes in the area did produce a 10 cm scarp along the fault where it crosses the road (unpaved in 1980) in McGee Creek. The results of the survey are plotted in Figure 2 as the height changes of 28 bench marks relative to their heights in 1982. The 1982 survey is represented as a straight line through the zero point of the ordinate. Three other surveys are also plotted for comparison, in particularly the previous resurvey in July 1997, done by UC Davis students under the supervision of Dr. Gerald Bawden, and the last previous UCSB survey done in July 1995.

Figure 2A.  Height changes among 29 bench marks, McGee Creek leveling line, for six of 18 surveys done in the period 9 July 1982 to 30 July 1998.  Uncertainty bars for the 1998 survey represent 1 standard deviation, and those for the other surveys are comparable, but are not present here to minimize cluttering the diagram.  Bench mark MC08 was not recovered in the 1989 and 1997 surveys.

Figure 2B - Topographic profile, McGee Creek leveling line.

Each of the surveys has a standard error of about 1 mm, consonant with what UCSB geodesists term "tectonic first order". Uncertainty bars are plotted only for the 1998 survey to avoid cluttering the diagram, but clearly the 1995, 1997, and 1998 surveys are indistinguishable within the levels of uncertainty for each survey.

The general pattern of all levelings is a gradual increase of scarp height to 1994 to 6 mm. The rate of growth, about 0.4 mm/yr, is scarcely noteworthy. It may represent long-term after slip, driven by the abundant, small earthquakes that have occurred in the vicinity of the Hilton Creek fault over the past 18 years, or alternatively, it may represent the culmination of some combination of systematic or nonsystematic errors in the UCSB surveying. The point is somewhat moot, given the small changes relative to the standard errors of the surveys.

Our interpretation of the leveling results is that no significant, vertical surface deformation occurred from 1995 to 1998 within that part of the Hilton Creek fault spanned by the leveling line. This conclusion is hardly surprising, given the relatively small magnitude of the earthquakes and their depths - but we had to check.

Even though the signal is small relative to the noise, it is of parenthetical interest to consider the overall pattern of the leveling results, especially of those bench marks in center of the line. They are set in boulders embedded on the steep face of the 15 m-high scarp. Previously we regarded their subsidence, relative to bench marks off the fault at the extremities of the line, to be a result downslope slumping by means of freeze/thaw. Thus, in the winter moisture in the ground freezes beneath the boulders containing the bench marks, pushing them upward and outward, normal to the scarp face. When the moisture thaws in the spring and summer, the boulders sink downward by the force of gravity. The freeze/thaw mechanism causes the boulders to creep downhill. Other boulders at the foot of the scarp are not embedded in the ground, and they seem to be unstable, especially MC08 (Figure 2).

An alternative interpretation for the sinking bench marks in the middle of the line is that they are located within a zone of extension along the fault, similar to sags or fault-bounded graben at the foot of a normal fault scarp. The boulders did not move downward appreciably from 1995 to 1998, however, in spite of the heavy precipitation in winter 1997-98.  Thus subsidence due to tectonic stretching is a viable hypothesis, especially for those few bench marks that are not on the scarp face. We need a bigger earthquake to prompt a larger signal out of the noise in order to discriminate sense from nonsense.