Temescal Canyon isn’t a name anyone uses. It’s kind of a ghost canyon, even though you’ve all driven through it many times — on Route 24 going to and coming out of the Caldecott Tunnel.
Underneath all the spaghetti and labels on that Google map is what remains of a fine little valley with steep sides and permanent streams that was once the principal water source of Lake Temescal, Oakland’s first surface reservoir. Here’s how the 1897 USGS topo map showed it.
Notice the stream northeast of the lake. The solid blue line signifies a perennial, year-round stream, and by the rules of hydrography it claims the name Temescal Creek given to its lower reaches, and hence comes the name Temescal Canyon. The other stream feeding the lake from the southeast is marked with a dot-dash line, indicating an intermittent stream; the map labels it Kohler Creek after the name of a landowner in Thornhill Canyon, but today that’s the creek called Temescal.
This annotated version shows two things: the arrows mark the Hayward fault and the dots outline the rim of the canyon.
Because the east side of the fault is rising, thanks to a bit of compression across the fault, the stream is forced to dig down harder than your average creek as the hills around it rise, and the long-term result is a watershed that’s wide at the top and narrow at the base — a wineglass canyon. As you know from my previous posts about Claremont Canyon and Shepherd Canyon, I have a thing about hiking around the rims of our wineglass canyons. This post is about that.
I call this a ghost canyon because waves of human intervention have modified it pretty seriously since that 1897 map. The original Kennedy Tunnel was punched through the hill in 1903, with the original Tunnel Road leading up to it. The 1915 topo map shows that when they built the road, the mapmakers added new detail to the contour lines on that side of the canyon. (Also, Oakland annexed all the land uphill from Berkeley.)
In addition to Tunnel Road, the Oakland Antioch & Eastern Railway was pushed through, skirting the lake on its way to the Shepherd Canyon tunnel. It ran over the Temescal Creek arm of the lake on a trestle.
Next came more infrastructure: PG&E constructed a big power line along with the Claremont “K” substation on Landvale Road in 1922, a classic industrial Deco structure. And then the Broadway Extension leading to the first Caldecott Tunnel bores carved up the sides of the canyon mouth even more in the 1930s. By this time, the Temescal Creek arm of the lake had been filled in. The work consumed huge quantities of rock, which was quarried from the north side of the canyon ( a few details here). Meanwhile residential development began on the north canyon wall. All of this is visible in the 1947 topo map.
Between then and now there was more residential development in Hiller Highlands, the big Parkwoods apartment complex in the canyon, a major expansion of Route 24 (with more quarrying to support the work), and finally the other little valley in the canyon was filled in to make the North Oakland Regional Sports Center. All of that was before the third and fourth Caldecott bores were added in this century. All that mayhem and erasure is what makes me think of it as a ghost canyon.
This patchwork of development has not created a ready set of roads to follow for a circumambulation, but I think I’ve cobbled together a trek route. And while I’ve walked all of it at one time or another over the years, I haven’t done it in one go. In fact I’m not eager to do so because it’s strenuous and a bit fraught and I need to build up strength first.
You can zoom in and explore this route at gmap-pedometer.com; this image (click to see it big) uses the OpenCycle viewing option. It starts at the Firestorm Memorial Garden at the foot of Hiller Drive, and right away goes seriously off-piste with a steep climb up the far side of the great Tunnel Road cut.
There are three fairly level stairsteps in the cut to choose from; this is the view back from the second. All three have cool views, and you avoid the heavy traffic on Tunnel Road.
The route veers off Hiller Road into a seekrit pathway that overlooks major features of the lower canyon. The power station and freeway lanes squat on the grave of the creek, with the Hayward fault slashing through the lake toward the notch in the horizon.
The spur across the narrow mouth of the canyon, carved flat for the powerline poles, is the endpoint of this trek.
And the ballfields of the sports complex smother this former stream valley. The woods to the left are an impenetrable eucalyptus thicket.
The next three miles-plus call for cautious walking: cars don’t expect hikers and the roads are narrow. But you’re high in the clean air well above the worst of the road roar. A little past the 3-mile mark, where the power line crosses the road, is an excellent place to stop discreetly and look down the ridge that forms the south wall of the canyon. That road in front of the ridge is Broadway Terrace, perhaps Oakland’s most dangerous road for pedestrians.
The long detour between miles 3 and 4-1/2 is unavoidable (trust me on this), but you can skip the last bit of Grizzly Peak Boulevard by turning right onto a footpath that’s part of the Sibley Preserve.
A little ways down Broadway Terrace is where it might feel a bit hinky. At Pine Needle Drive, you climb over the fence and locate a teeny footpath, almost a deer trail, along the power line that plunges about 200 feet to a fire road, which then climbs all the way back up to the ridgeline past a big landslide. Believe me, that is less fearsome than walking on upper Broadway Terrace.
The last leg, from the ridge down to the powerline tower pad, I can’t really vouch for as I last walked it 11 years ago. Assuming it’s not too overgrown, you should be good, and if not, then go back on the fire road down to the sports center. The pad is at the end of Pali Court, and the view back to the starting point looks like this.
Getting to and from these two endpoints is an exercise left for the reader, and I do mean exercise.
Oakland Geology 
2 March 2020 at 10:44 pm
Here’s one of my usual semi-educated questions: These wineglass canyons are created by uplift on the east side of the Hayward fault. So I wonder if geologists consider the Hayward fault to act sometimes (or all the time, but partially) as a thrust fault with more uplift than transverse movement?
5 March 2020 at 9:08 am
Motion along the fault is transpressive, about 10 percent thrust and 90 percent strike-slip.
2 July 2020 at 6:02 pm
The creek headwaters are still evidenced. I was riding the fire road above the sports center a couple of days ago (on unicycle, it so happens) and there is a seep across the road, active enough that I still have clay stuck in my tire. June 30, hasn’t rained for weeks.