Anomalies of Sausal Creek: The Delta

11 November 2019

This is the last of four posts about Sausal Creek from the hills to the Bay focusing on its odd features, stuff that’s been bugging me like a seed stuck in a tooth. Here I’ll talk about the creek segment below Foothill Boulevard, where the floodplain ends and the delta begins.

A delta is a wedge of sediment, built near a river’s mouth where it deposits most of its muddy load. Streams tend to spread out in their deltas, sending sediment here and there like a Vegas card dealer. While the mouth of a big river like the Mississippi or the Nile fans out in a set of multiple distributaries, little streams like Sausal Creek move their courses every once in a while so that over thousands of years, every part of the delta gets its share of dirt.

Today, Sausal Creek officially meets the Bay in this culvert next to the Fruitvale Bridge . . .

. . . with this monumentation.

But it’s all totally artificial. This body of water is a large canal that was built in the late 1800s as part of the Oakland harbor improvements. Before that, Alameda was not the “Island City” but a town on a wooded peninsula called the Encinal, and the only way to get there on dry land was through here, across the delta of Sausal Creek.

The creek’s delta is unlike the deltas of Oakland’s other major streams. Here’s what I mean. Look at this map of central and east Oakland that shows only young material, whatever is not bedrock.


From USGS map OF 2006-1037

I’ve labeled Sausal Creek, shooting south-southwest straight to the Bay where the Alameda peninsula sits in its way. The dark purple stuff labeled “afem” is all landfill (“artificial fill over estuarine mud”), and the light yellow part marked “Qhf” is young river sediment (“[Quaternary Holocene] alluvial fan deposits”). The three pink blobs labeled “Qds” are areas of old Ice Age sand dunes: one under downtown Oakland, one making up the Encinal, and one under Bay Farm Island.

So before the canal was dug, Sausal Creek, unlike all other Oakland creeks, dumped its load here against a buttress of sand. The willow thickets that gave Sausal Creek its name must have thrived here. The early Anglo settlers were quick to put roads and rail lines through this area, and the brushy marshy creek delta would not stand in their way for long.

Which way did the creek run from here, to the right or the left? The only clues are a few old maps, not all of them trustworthy. The first official map of Alameda County, published in 1857, shows Sausal Creek, at top center, draining to the right into San Antonio Creek, the tidal inlet now known as the Oakland Estuary.

The Haynes map, published in 1878, shows it petering out and not even reaching the Bay. Other Oakland creeks, like Courtland and Upper Elmhurst and Seminary Creeks, were the same before they were diverted into pipes and ditches.

But every other map, of this vintage and later, shows Sausal Creek draining to the left into San Leandro Bay. I have no doubt that the road and railroad builders dug a ditch through the sand to control it (and conveniently mark the boundary between Alameda and Brooklyn Townships). The Thompson map of 1878 is a good example, followed by a map of the same area today. Sausal Creek is at top left.

Is it possible that the creek mouth shifted naturally from the right to the left during those years? Certainly; in fact the notorious winter of 1861-62, the wettest in our recorded history, could have done that by sending a big pulse of mud down the creek as it overspilled its banks. And the 1868 earthquake could have liquefied the ground here. The rarest events make the biggest difference. But in this case I would blame us.

The geologic map of Oakland shows the delta in more detail, outlining especially young wedges of sediment (Qhaf1) that were laid down by the creek in very recent geologic time, probably within the last few thousand years.

It’s even plausible that Peralta Creek, just a few blocks east, joined Sausal Creek off and on over the centuries and contributed to this delta. A prominent example is up in Richmond where Wildcat and San Pablo Creeks form a joint delta, at one point flowing just a stone’s throw from each other. But today Sausal Creek’s mouth is a truncated version of its true self, trapped in culverts for the foreseeable future, a dead delta.

Anomalies of Sausal Creek: The Floodplain

28 October 2019

This is the third of four posts covering the Sausal Creek watershed from hill to Bay. It features the part of the creek between Dimond Canyon and Foothill Boulevard. The early Anglo landowners were quick to establish fruit orchards and berry patches there, giving rise to a settlement and district out east in Brooklyn Township they called Fruit Vale.

That was just the right name for this distinctive locality. Its alluvial land, flat, tractable and naturally well watered, was ideal for planting. I say “planting” rather than farming because within a few decades — the commercial lifespan of an orchard — the farms gave way to grand homes with lush landscaping. That’s the “Fruit” part. The “Vale” part refers, as the word has since the 14th century, to a wide, flat lowland between gentle hills with a stream in it. An 1888 pamphlet described it as “one vast garden,” its roadway “lined with elegant residences and villas surrounded by beautiful lawns and gardens. . . . The soil is easily cultivated and everything planted in it thrives.”

The creek itself is hard to see in this stretch today, where homes and overgrowth block the streambanks. It also runs a good 20 feet below the surrounding land, a typical East Bay arroyo with banks like cliffs. Except for a lovely little fenced-off area on Barry Place, distant glimpses are the closest you can get to the water.

By textbook standards, Sausal Creek has the most well-formed floodplain in Oakland. A floodplain forms wherever a stream runs out of downhill energy and has to drop the sediment it’s carrying. The elevation profile I made for Sausal Creek shows the floodplain portion.

This part of the stream is at grade, like the canyon above it and the delta below. It’s very stable. Neither fully erosive nor fully depositional, a stream in a floodplain has little it can do except rearrange the sediment.

Floodplain streams tend to turn snaky, meandering and migrating like the trickles you’ll see on a rainy windshield. They nip off bits of the bounding hillsides and gradually widen their corridor.


From Phil Stoffer’s geologycafe site

A floodplain is a place where sediment languishes during its slow journey to the delta. A lot of that sediment enters the floodplain from the hills on either side. Here are three different views across the floodplain showing its flat floor and abrupt edges. Fruitvale Avenue at E. 27th Street:

Farther upstream at MacArthur Boulevard:

And still higher at Coloma Street.

For real detail, I always look for a digital elevation model or DEM. The DEM of the upper end of the floodplain shows how flat its floor is and how fragile the west side of the valley is. It also shows how much the stream itself is meandering within the straight floodplain valley.

Unlike the rocky flanks of Dimond Canyon, the hills lining Sausal Creek’s floodplain consist of sediment, ancient gravel beds that I’ve named the Fan. Time to look at the geologic map of the floodplain section.

The material of the Fan, labeled Qpaf on the map (“Quaternary Pleistocene alluvial fan”), is unconsolidated sand and gravel that’s prone to landslides on steep slopes. That scoop out of the hillside just left of center in the DEM image is the notorious McKillop slide, in progress for more than a century. Much of the crumpled land there was smoothed over and preserved as William Wood Park, but the adjoining slope to the south, which failed in 2006, is still a mess.

Two things about this stretch of Sausal Creek strike me as odd. The first is how straight the stream valley is here. Streams tend to curve, so if you find a naturally straight one you look for what’s forcing it to be that way.

My tentative theory, which I alluded to at the end of the previous post, is that the Fan was tilted up, not built up, along with the Piedmont bedrock block thanks to tectonic interactions along the Hayward fault. (In other words, the Fan is not really an alluvial fan but an uplifted relic of an older coastal plain.) I suppose that the stream was incised in a straight line by large flood events early in its history, then widened into the floodplain we see today. It’s easy to imagine big pulses of floodwater coming out of Dimond Canyon like a firehose, most likely after a temporary lake above the canyon, dammed up by an earthquake landslide, gives way. Such floods would rush in a straight line to the sea and set the stage for the floodplain that evolved.

The other odd thing is how closely the stream hugs the west side of the valley. So does Peralta Creek, just east of Sausal Creek. I think this may not be an accident. Glen Echo Creek, on the other side of the Fan, hugs the east side of its valley. Perhaps there is some subtle warping of the crust across central Oakland.

Sausal Creek’s fruited plain may be just a vale, but it’s a vale with some intriguing features.

Anomalies of Sausal Creek: Dimond Canyon

14 October 2019

This is the second of four posts about the peculiarities of Sausal Creek, going from its headwaters to the Bay. Here I’ll address Dimond Canyon, the 2-kilometer segment between the Warren Freeway and the flats of Dimond Park. The steep walls of the canyon, which is several hundred feet deep, are entirely hard sandstone of the Franciscan Complex, part of the Piedmont block.

This is the same stone quarried for decades in Rockridge (the Bilger quarry) and the land that would later become Piedmont (the Blair quarries and the Davie Stadium quarry). In fact the Diamond Cañon Quarry was one of two here in the canyon. It’s now occupied by the Zion Lutheran Church, as seen here from across the canyon.

The quarry scar appears on this terrain map as a big round nick in the canyon wall next to Park Boulevard.

A while ago in this space I described Dimond Canyon as a classic water gap — a stream-cut gorge crossing a bedrock ridge that otherwise seems impenetrable.

Geology textbooks will tell you there are two ways for streams to make a water gap. In the first way, the stream was there first (an antecedent stream) and a ridge of resistant rocks rose up around it. In dynamic California, this is a straightforward explanation of our water gaps. In the second, the ridge was there first, inherent in ancient deformed rocks buried under younger strata, and the stream (a superposed stream) cut down to, then into it while stripping off the overlying material. That’s how they explain the Delaware Water Gap and other examples in the gentle Appalachians.

Dimond Canyon is actually a semi-classic water gap. Yes, the ridge it crosses must have risen while the stream was cutting down, but the story is complicated by the fact that the watershed upstream lies across the Hayward fault, and is constantly being moved to the right. This means the canyon has hosted streams from several different watersheds over the past million years or so.

Therefore the streams feeding Sausal Creek today could not have dug the canyon; some predecessor watershed did it. There must have been gaps and surges in the water (and sediments) flowing through this canyon. If we ran things backward a million years, what would it show? The exercise would be blurred by serious uncertainties, but the matter is not beyond all conjecture.

I beg your indulgence as I present some slides from my talk to the Friends of Sausal Creek last month. They’re Google Earth views looking west across the fault. Here’s today, with the fault trace shown in red.

The view may be a bit confusing as I rewind the motion on the fault at about 10 millimeters per year. The far side looks the same because we’re focusing on it while it moves leftward, toward San Leandro. For a long time, Sausal Creek has been carried past small watersheds that, like today’s, could not possibly have carved Dimond Canyon. But about a million years ago, Dimond Canyon would have lined up with the watershed of Arroyo Viejo.

This looks promising because the watershed (the part above the fault) is about twice the size of Sausal Creek’s, giving it roughly twice as much water and cutting power to match.

But to make the canyon, you have to have something pushing up the ridge while the stream across it keeps cutting its way down. There’s nothing obvious that would have been pushing up the bedrock ridge at this time.

Going back a bit further, though, we line up with the great big watershed of San Leandro Creek, a dozen times larger. This stream has plenty of cutting power, evident in the canyon it’s dug where the dam and reservoir sit.

And finally, we have a mechanism here for uplifting the ridge that Dimond Canyon cuts across. The hills of San Leandro consist of a large slab of gabbro so big and strong that it deflects the Hayward fault slightly. Back when the sandstone of Dimond Canyon was grinding past the gabbro of San Leandro, the jostling between these two bodies of rock, caught in a vice by the geometry of the fault (a restraining bend), would have pushed both sides upward because that’s the only way out of the vice. And all the while San Leandro Creek would have been cutting a nice deep water gap as that hard rock rose.

Eventually, inevitably, the fault carried the water gap out of reach, and ever since then Dimond Canyon has housed lesser creeks for episodes of a few hundred thousand years. Sausal Creek trickles down the canyon today not doing much to it, the shrunken tenant of a structure built by a mightier maker.

This story (and that’s all it is really) appeals to me because it would also explain the presence of the Fan — the swath of gold on the geologic map representing Pleistocene sediment.

I’ve always regarded it as a fossil alluvial fan because of its shape on the map, but maybe that’s accidental. Maybe it’s just a chunk of old East Bay land that was lifted along with the Piedmont block, or washed off of it afterward.

I first posted about the problem of Dimond Canyon more than 10 years ago. Takes a while to figure out some things.

Anomalies of Sausal Creek: The Headwaters

30 September 2019

The Sausal Creek watershed is full of anomalies and questions from its headwaters to its San Francisco Bay outlet. Here I’ll look at the top end of Sausal Creek — the part that isn’t Sausal Creek. The creek originates where three different tributaries join: Shephard Creek, Cobbledick Creek and Palo Seco Creek.

The divide between Shephard and Cobbledick runs up Chelton Drive, then Darnby and Carisbrook Drives, up to Skyline. The divide between Cobbledick and Palo Seco runs mostly up Castle Drive to Skyline, so if you know the area these are easy to visualize.

That map, from the Alameda County Flood Control and Water Conservation District, has Sausal Creek proper beginning at the junction of Shephard and Cobbledick, which is also where the Hayward fault crosses the creek (more about that below). But because that’s in a culvert deep beneath the Warren Freeway, for purposes of this post I prefer to put the origin a little farther down, in an easy-to-miss opening east of the parking lot of the Montclair Golf Course driving range at the head of Dimond Canyon. That’s where Palo Seco Creek, the third tributary, comes in from the redwood-filled canyon of Joaquin Miller Park.

So with that settled, let’s look at the headwaters on the geologic map. This area includes a wide variety of rock units — the Sausal Creek watershed touches more different rock types than any other Oakland stream.

Don’t worry, I won’t go into the rocks, although there are a lot of them and they’re interesting . . .

I’ve added the Hayward fault to the map, as a thick red line, just to show how different the rocks upstream and downstream are. That’s because motion on the fault has been dragging the west side to the north for a few million years. That explains two major peculiarities of Sausal Creek, the first being its lumpy longitudinal profile.

I made this stream profile by walking down the creek from the top of Eastwood Court to the Bay, recording elevations with my smartphone altimeter.

A normal stream profile describes a nearly smooth listric curve — steep at the top and level at the bottom. The bottom of the curve represents what’s called the base level for the whole stream, sea level in this case. A stream with a nearly perfect curve is said to be at grade. Two basic things will put kinks in that curve: rocks that are especially hard or soft, and changes in base level. For instance, ice ages lower the sea level by hundreds of feet, and streams have to adjust during that time by digging down their beds toward the new base level. (I alluded to this in my last post with respect to Lake Merritt.)

There’s a big discontinuity in this curve right where the Hayward fault crosses, just above the 3 kilometer mark. The lower half, Sausal Creek, is at grade, even though it crosses hard sandstone in Dimond Canyon and young sediment farther down. But the fault has ripped its head off and put on another head — the Shephard-Cobbledick-Palo Seco system. It’s a Frankenstein creek. I think this head transplant has happened more than once.

Sausal Creek, over the last million years or so, has done fine even with its head ripped off and replaced. The worst that might have happened is that it had less water in it for a while. But in the upper creek system, these changes have drastically affected its base level.

Picture it: the two sides of the Hayward fault are moving past each other at about 10 millimeters a year, or a kilometer every 100,000 years. So for a good long time, Shephard Creek flowed down against the high rocky ridge of the Piedmont hills. Very likely it turned north for a long ways, the way Temescal Creek does today, before flowing around the north end of the ridge. As its route to the Bay got longer and longer, the slope of the stream grew gentler as it remained at grade. The effective base level, that is to say, was up around the elevation of the Thornhill district.

Then along came Dimond Canyon, moseying up on the far side of the fault, and at some point Shephard Creek switched over to that route. All of a sudden, it had a lower base level. It was not at grade. So it started eroding downward into its streambed and eroding uphill, trying to re-establish that nice listric profile.

When that happens to streams, what geologists call a knickpoint appears in the profile. The extreme case of a knickpoint is a waterfall, but more often they’re just rapids. In Shephard Creek, there appear to be two knickpoints.

Pay attention next time you ride down Shepherd Canyon Road. There’s a nearly level stretch in the road between Saroni and Escher Drives, where the railroad trail meets the road, then a steep “rapids” below. The other knickpoint is under the landfill of Shepherd Canyon Park, where the creek is buried in a culvert. A more carefully made profile would show it better, but that may not be possible.

I could conjecture a story that accounts for these features, but there’s a lot I don’t know so it would just be armwaving. For instance, the stream profile is based on the elevations of the roadway rather than the actual streambed except for the part between Shepherd Canyon Park and Mountain Boulevard (which is so thickly wooded I don’t recommend you visit, even though I did) and the one data point at the golf club. For another, the history of vertical movements along the Hayward fault is almost completely unknown, other than that the hills on the east side are rising today at about a millimeter per year. So enough about that.

I mentioned that the creek has two major peculiarities, and here’s the second one. The upper part of the Sausal Creek watershed is not a pretty, textbook stream network shaped like a nice tree — what geologists call a dendritic drainage pattern. It’s more like a bush in a gale, and all of the streams that cross the fault are warped. Here’s how it looks in the set of stream maps on the Oakland Museum website:

And for comparison here’s Temescal Creek:

And Arroyo Viejo, the weirdest of all.

There’s a struggle going on between the stream’s innate tendency, driven by gravity, to dig itself a home with an optimal shape and the motion of the ground beneath, driven by tectonics, that keeps messing it up. The shapes of the streams, and the landscape they live in, are a snapshot of that struggle. They remind me of the shapes of trees high on windy mountains — although one case involves organisms and the other purely physical systems, the similarities are tantalizing.

Geologists are starting to explore this topic with computer models. A recent paper in Geophysical Research Letters, with the dry title “The Role of Near‐Fault Relief Elements in Creating and Maintaining a Strike‐Slip Landscape,” has some state-of-the-art animations that address the exact situation of Oakland’s fault-crossing streams. You don’t need to understand all the modeling details — I certainly don’t — to enjoy the illustrations and the movies.