Open thread

8 November 2021

I’m so busy finalizing the manuscript of my book that I can’t find the bandwidth for my accustomed fortnightly post today. There won’t be one in two weeks, either. Next post will be on 6 December. In the meantime, the comments are open so we can talk to ourselves for the rest of the month. (“Did you even know you have selves you can talk to?” Bob Weir reputedly told the audience at a Grateful Dead show as they took a technical break.)

The photo is a fine specimen of mariposite rock from the Carson Hill quarry in Angels Camp. This popular landscaping stone is shot with a green chromian variant of the metamorphic mica mineral phengite. Portions of the deposit are worked for gold as part of California’s continuing gold crawl. The specimen is a recent addition to the rock garden at Lake Merritt.

What’s up with you?

Stop saying “overdue”

25 October 2021

The last week has had its share of local earthquake news, even though there weren’t any earthquakes nearby. It all centered around the release of volume 3, the last part, of the massive HayWired Scenario report, conveniently timed for 21 October, anniversary of the 1868 Hayward earthquake (not to be confused with the 17 October earthquake of 1989).

HayWired is a virtual magnitude-7 earthquake, complete with aftershocks, that represents a typical Big One on the Hayward fault. Seismologists created it as accurately as their science permits, then asked emergency responders, social scientists, planning agencies, structural engineers and other specialists what they think would happen to the Bay area and how they would handle it. Volume 3, “Societal Consequences,” presents all their answers, as accurate as their expertise permits.

In brief, the consequences would be dreadful. Ace reporter Ron Lin of the Los Angeles Times wrote an able summary that I’ll just point you to rather than write my own. Besides, I covered some of the same ground a few months ago.

The East Bay Times, to its credit, also ran Lin’s story, and two days later it issued a wake-up editorial, “Prepare today for next major Bay Area earthquake,” aimed at goosing its readers into action against the threat. It’s a bit overdone, starting with the opening paragraph: “Gulp.” I don’t really mind that, but the editors went on to say something sloppy that I will focus on today:

“We know that the last major earthquake on the Hayward Fault occurred in 1868 — 153 years ago. We also know that, on average, dating back to the year 1134, the fault produces a major earthquake roughly every 150 years. So, yes, we’re overdue.”

No, we are not overdue. Scientists don’t use that word because it’s a deep error in thinking. Something that’s overdue is late, behind schedule, and earthquakes don’t follow a schedule. I don’t like scaring people with inaccurate statements.

Ron Lin, to his credit, stopped short of using the O-word:

“The Hayward fault is one of California’s fastest moving, and on average, it produces a major earthquake about once every 150 to 160 years, give or take seven or eight decades. It has been 153 years since the last major quake — a magnitude 6.8 — on the Hayward fault.”

Instead, he included the uncertainty around that irresistibly tempting “average.” That was helpful, but he didn’t come up with the best word.

Even the U.S. Geological Survey creeps close to the wrong word in its excellent Fact Sheet 2018-3052 titled “The Hayward fault — Is it due for a repeat of the powerful 1868 earthquake?” It sidles up to this D-word, and by implication the O-word, by saying that “the interval between successive quakes has varied from 95 to 183 years, averaging 150 years, and it is now more than 150 years since the 1868 earthquake….” and trails off with that coy string of dots. The sentence leads with the uncertainty, which is good, but the conclusion it implies is not correct. The graphic it goes with is useful in showing the raw numbers behind the average:

There’s a rhythm to this timeline, but not a good beat. Here’s a longer timeline, currently the best we have, that presents the uncertainty of the radiocarbon-based dates in it:


Source: USGS

Those snappy stars are actually smeared into blurs. For instance, the date of that event “in the year 1134” that the newspaper cited is uncertain by over a hundred years.

Maybe I’ve made it clearer what frosts me (and most seismologists) about the O-word. Now the correct, best word for the situation on the Hayward fault is this: the fault is ready for a major earthquake. It’s primed, loaded, set to go. This is scientifically correct because we’ve measured the actual motions of the crust around the fault and know that since 1868 it has accumulated enough strain energy (the kind in a stretched rubber band) to be released in a HayWired-sized earthquake. “Ready” is not as scary as “overdue,” but sit with it and the word is pretty motivating just the same. Are YOU ready?

The prospect of reading the whole HayWired Scenario report is intimidating. I recommend Chapter R as a good summary that will guide you to specific chapters where you can dive deeper.

Load casts in the Shephard Creek Formation

11 October 2021

I see that I haven’t given my photos of this interesting feature their own page. They show some dramatic load casts on the south side of Shepherd Canyon, near the east end of Escher Drive. A load cast is made when heavy material, like a mudflow, crosses soft sediment and sinks its feet into it.

My most memorable lesson about features of this kind happened in 2008 during a visit to Point Reyes. The path leading to the lighthouse passes a chaotic scene that most visitors ignore.

The scene was an offshore basin, probably like the offshore Monterey Canyon today, where every now and then an undersea debris flow, full of gravel and sand, fell rudely upon nice quiet beds of deep-sea clay or soft mud. The results included scour marks . . .

and rip-up clasts, hunks of (easily eroded) clay swept into the flow . . .

and downward-pushing load casts accompanied by upward-pointing flame structures.

It was a great pleasure to come upon high-quality load casts in the Oakland Hills, in the mudstones of the Shephard Creek Formation, and lead group walks past them. Here’s the overall scene, photographed in February 2016. A thick layer of massive (i.e., unbedded) sandstone overlies thin-bedded shale and mudstone.

Near the base of the sandstone, on the lower right side, are these well-exposed load casts.

The previous June, I took a closeup of the underside.

Unfortunately, a Google Maps image from January 2021 appears to show that this feature has crumbled off the roadcut. That’s how geology goes in the Oakland Hills, and that’s one reason I constantly take photos. I also tell myself, in consolation, that new examples could appear on any given day.

The classic 1995 text Sedimentographica has good photos of these and many more features of sedimentary rocks. I treasure my hard copy, but maybe the publisher’s online version will outlast it. That one, YOU can enjoy.

Merritt Canyon

27 September 2021

Like all true Oaklanders, I keep coming back to Lake Merritt. In this visit, I’ll muse about the many times in the recent geologic past when there was no Lake Merritt here.

If we assume, as I do, that the uplifted block of bedrock making up most of Lake Merritt’s watershed is about 1 million years old, then this little arm of the Bay has a fairly deep ice age history. A million years rather neatly fits the period of Pleistocene time after the Mid-Pleistocene Transition, when for imperfectly understood reasons glacial cycles changed from roughly 40,000 years to 100,000 years in length.

Let us then stipulate that this part of town has gone through about ten full glacial cycles. Each time the world’s ice sheets and ice caps grew, the sea level fell by a hundred meters or so — three or four hundred feet! Here’s a recently published set of sea level data for the last nine cycles. Various lines of evidence agree, though never in exact detail, about the timing and magnitude of the changes. That’s what paleoclimate science looks like.


From Rachel Spratt & Lorraine Lisiecki, 2016, A Late Pleistocene sea level stack, Clim Past 12(4)

Each time the sea fell, all of San Francisco Bay slowly drained dry and the coastline withdrew out past the Farallon Islands. Every stream that could do so cut into the newly exposed ground, digging gulches, ravines and canyons into the young sediment where before they were prevented by the high sea level. That’s when Lake Merritt became temporarily Merritt Canyon, with Merritt Creek surging along its bottom.

Each time, Merritt Creek dug out all the gravel and mud that had filled the basin of Lake Merritt and shoved it straight out into the Bay, where the drainage ran south and then around the end of San Bruno Mountain into the Pacific. That’s right — the Golden Gate didn’t exist. Instead, the Bay area streams and the great Central Valley rivers drained through what’s called the Colma Gap.

Here’s an illustration from a publication I refer to often, Sandy Figuers’ “Groundwater study and water supply history of the East Bay Plain,” that shows the typical drainage pattern of those times.


Courtesy State Water Resources Control Board

It’s very interesting: the river ran east of the Potrero Hills in Richmond, east of Yerba Buena Island, and all the way down to around San Mateo. And Merritt Creek, right in the center of that map, pushed a big fan of alluvium — river gravel and sand — into the basin that rivaled the fans of the other major East Bay streams. It went right under Alameda. That’s because Alameda wasn’t there.

This configuration of the Bay lasted from about 630,000 years ago, when the great freshwater Lake Clyde that once filled the Central Valley broke through the hills and cut Carquinez Strait, until the last warm interglacial period about 125,000 years ago. The sea at that time rose even higher than it is today, and motion on the San Andreas fault closed the Colma Gap. That’s when another gap opened up farther north on the fault, which became the Golden Gate.

So when the next glacial age began, the whole drainage pattern of the Bay shifted dramatically. Also the winds: instead of blowing south through the Colma Gap, the ice age westerlies carried huge amounts of glacial sand through the Golden Gate and across San Francisco, across the dry grasslands of the Bay and onto the East Bay slopes. That’s when the big fields of sand dunes accumulated in San Francisco and in three places on this side of the Bay: in downtown Oakland, in Alameda and on Bay Farm Island.

The latest time that Merritt Canyon formed, Alameda sat in its way all of a sudden, and I think Merritt Creek must have drained west, down San Antonio Creek (today’s harbor estuary), not southward as shown on this figure from the same report.

Here’s part of a third figure from that report, showing the depth to bedrock in Oakland as determined in boreholes.

Merritt Canyon really stands out underground: over repeated ice-age cycles, as the Bay floor has gradually subsided, the earliest versions of the canyon now lie 600 feet below Lake Merritt.

There should be a record of successive incisions and fillings of the canyon preserved down there. It would take a concerted campaign of core drilling and seismic reflection profiling to map and characterize it, and if I were a billionaire like some people I won’t mention, I’d spend the money even though it would be a huge hassle to get the permits.