The Rose Canyon fault zone in San Diego, California, has many well-exp
ressed geomorphic characteristics of an active strike-slip fault, incl
uding scarps, offset and deflected drainages and channel walls, pressu
re ridges, a closed depression, and vegetation lineaments. Geomorphic
expression of the fault zone from Mount Soledad south to Mission Bay i
ndicates that the Mount Soledad strand is the most active. A network o
f trenches excavated across the Mount Soledad strand in Rose Creek dem
onstrate a minimum of 8.7 m of dextral slip in a distinctive early to
middle Holocene gravel-filled channel that crosses the fault zone. The
gravel-filled channel was preserved within and east of the fault but
was removed west of the fault zone by erosion or possibly grading duri
ng development. Consequently, the actual displacement of the channel c
ould be greater than 8.7 m. Radiocarbon dates on detrital charcoal rec
overed from the sediments beneath the channel yield a maximum calibrat
ed age of about 8.1 +/- 0.2 kyr. The minimum amount of slip along with
the maximum age yield a minimum slip rate of 1.07 +/- 0.03 mm/yr on t
his strand of the Rose Canyon fault zone for much of Holocene time. Ot
her strands of the Rose Canyon fault zone, which are east and west of
our site, may also have Holocene activity. Based on an analysis of the
geomorphology of fault traces within the Rose Canyon fault zone, alon
g with the results of our trenching study, we estimate the maximum lik
ely slip rate at about 2 mm/yr and a best estimate of about 1.5 mm/yr.
Stratigraphic evidence of at least three events is present during the
past 8.1 kyr. The most recent surface rupture displaces the modern A
horizon (topsoil), suggesting that this event probably occurred within
the past 500 years. Stratigraphic and structural relationships also i
ndicate the occurrence of a scarp-forming event at about 8.1. kyr, pri
or to deposition of the gravel-filled channel that was used as a pierc
ing line. A third event is indicated by the presence of several fault
strands that displace the channel but did not move during the most rec
ent event. Other events may also have occurred, but these data suggest
that the return time for surface-rupturing earthquakes is no more tha
n about 4 kyr.