Groundwater pumping and subsidence in the Central Valley
It's been called the largest alteration of the earth's surface. In the San Joaquin Valley, since the 1920s, farmers have relied on groundwater to varying degrees, and over time, overpumping of groundwater basin has caused the land to subside – over 30 feet in some locations. In this presentation, USGS hydrologist Michelle Sneed discusses subsidence in the San Joaquin Valley, specifically along the Delta Mendota Canal, where there’s been some problems associated with subsidence, as well
She began with a photo that’s very well known for depicting land subsidence. “This is Joe Poland and he’s using a telephone pole or a power pole to illustrate where the land surface was in 1925, where the sign is, 1955, and where he is standing in 1977. Almost 30 feet of land subsidence has occurred at this location during that time period.” She noted that in the picture, he is standing is southwest of Mendota, but today she would be talking about a different area, a new area where they were surprised to find subsidence occurring.
“What we found were 1200 square miles subsided in the northern San Joaquin Valley area in an area bounded by Mendota on the south, Merced on the north, Madera on the east, and Los Banos on the west,” said Ms. Sneed. “The subsidence is occurring at rates ranging from about a half inch a year to almost a foot a year over a 2 year period, from 2008 to 2010.” She noted that surveys done since then by Reclamation and DWR indicate that these rates of subsidence have continued through 2013.
“The problem with subsidence in this area is that there’s a lot of water conveyance infrastructure that’s getting impacted by the subsidence,” she said. “This includes the Delta Mendota Canal, which is the original study area that we were working on, also the California Aqueduct, the East Side Bypass, which is the most important flood control channel east of the San Joaquin River, the San Joaquin River itself in the restoration area, and many local canals.”
“The subsidence is permanent, meaning even if we stopped pumping so much groundwater and groundwater levels actually start to recover, the land surface will not come back up,” she said. “This subsidence occurred when groundwater levels declined to historically low levels as a result of pumping.”
She said the recent subsidence is occurring about 25 miles northeast of where it has occurred historically. “What we need is long term monitoring of groundwater levels and subsidence to detect and track groundwater conditions to help with decision support, because when the subsidence stopped in the 80s, we moved resources to other issues throughout California because subsidence was essentially arrested,” she said. “We turned our heads for 40 years and when we turned back around, we found that these very fast rates of subsidence. Long term monitoring would help us avoid those kinds of surprises.”
The Central Valley, and the southern half, which is the San Joaquin Valley, is very important to California, she said. “It’s large; it takes up the center part of our state – about 20,000 square miles or so. About 250 crops are grown here, and they’re worth about $17 billion a year, so it’s very important for California’s economy,” she said.
“It’s also important to feed the nation – about 25% of food that feeds the nation comes from the Central Valley, so it’s a very big contributor to our nations’ food supply,” she continued. “But only about 17% of the irrigated land is in the Central Valley, so while 17% of the irrigated land is in the Central Valley, we’re producing about 25% of the table food. Another interesting statistic is that 20% of all the groundwater pumpage in the nation occurs in our Central Valley aquifer system.”
There are many kinds of subsidence, but principally only two kinds in California, said Ms. Sneed. The type of subsidence that occurs in the Delta is a shallow process that is unrelated to the subsidence occurring in the Central Valley, she noted.
The type of subsidence that is occurring in the Central Valley is due to aquifer system compaction, which is concentrated in the find grained deposits called aquitards, she said. “The reason that these fine-grained deposits, especially clay, are different is because they are platy minerals, and when they were laid down originally, they were laid down in random orientations,” she explained. “But when we start to lower groundwater levels, and we lower the pore pressure, which is the pressure the water is exerting on the grains that is keeping them apart, what happens is that the grains start to rearrange themselves into more like a stack of pancakes. You can see there’s a lot less room between a stack of pancakes for water to be stored than in randomly oriented ones. So clays are the big player in aquifer system compaction.”
The level at which the grains rearrange themselves is called ‘pre-consolidation stress’, and it tends to be the previous lowest groundwater level, said Ms. Sneed. “The result is you have reduced storage capacity. The subsidence is largely permanent, so these grains will not go back into random orientations, even if water levels come back up.”
We care about land subsidence for two reasons: Infrastructure damage and flood protection and damages to our natural resources, she said. “Water conveyance systems and other water infrastructure get damaged by subsidence because it’s happening at different rates at different locations. If the whole San Joaquin Valley was subsiding at the same rate and in the same way, then nobody would really care, but it’s this differential subsidence, the different amounts of subsidence in different places that really damages canals, roads, railways, pipelines, bridges – anything that crosses these areas of differential subsidence can get damaged.”
Canals are particularly sensitive because gravity is oftentimes used to move water, and this means that every point downstream needs to be at a lower elevation than every point upstream or pumps are needed, she said. “We rely a lot on gravity which means canals are built at very specific elevations and very specific relative elevations, so if you start to lower one part of a canal, then all of the downstream parts of that canal are impacted by the subsidence upstream. This results in reduced conveyance capacity which is probably the most cited problem with land subsidence in canals – we can’t push as much water as we used to.”
Subsidence reduces freeboard, which is the distance between the water surface and anything that crosses it, such as bridges or roads. “When you have subsidence, now sometimes the water will be running into the bridge which will damage the bridges integrity, causing erosion and other problems,” she said. “For lined canals, we start to get water coming up over the top of the concrete liner if it’s misaligned, because of subsidence. Water will go over the top of the liner and erosion problems will occur subsequent to that.” She noted that with unlined canals, subsidence can cause deposition and erosional problems.
She then presented a photo of a well in the San Joaquin Valley. She explained that when the well was drilled in 2010, the painted the top of it orange so farm equipment wouldn’t hit it. “However, two years later, two additional feet of that well casing were sticking out of a ground, so we’re looking at about a foot a year at this location,” she said.
She then presented a slide with several pictures depicting damage that subsidence is suspected to have caused to infrastructure. “Along the Delta Mendota Canal, there’s a buckling in the concrete liner, and while we can’t say for sure that this is the result of subsidence, this is an area along the Delta Mendota Canal that has had serious issues with subsidence. They’ve had to do a lot of infrastructure retrofit in this area so it’s in a suspicious location.”
She then presented a slide with more pictures of damage to infrastructure. “The top and bottom photos are both of the same area, and on the left, you can see that this is a double-decker structure here. When it was originally built, it was not double-decker structure, so they had to build it up so they could maintain the elevation of the water surface at a certain elevation in this canal,” said Ms. Sneed. “The Delta Mendota Canal is a federal canal, and they have the resources to put into the canal to build up the infrastructure and to do the mitigation that’s necessary to keep the flow capacity near design capacity. This canal is a place where federal money is backing infrastructure improvements to keep up with subsidence.”
The picture on the right is the Outside Canal just a few miles north. “You can see that there are sidewalls built up on the bridge; they weren’t there when the bridge was first designed,” Ms. Sneed said. “They had to build up the sidewalls to keep the water off of the road. This whole bridge is actually going to be torn down and redone because it’s starting to erode the bridge structure, so it needs to be replaced. So this is an area where they have lost freeboard. You should see the water going under the bridge but instead you see the water going into the bridge.”
Natural resources can also be affected by subsidence besides the reduced aquifer storage capacity, she said. “Wetlands and rivers flow downhill, and they are in the low spots in the land, so as we are differentially lowering the land surface, the wetlands and riparian corridors may get moved around, so rivers may shift course and the aquatic ecosystems that depend on those will also have to modify, and then of course, we get restricted land uses based on those kinds of problems.”
Historically, subsidence or elevation change was measured using benchmarks and networks, first by spirit leveling and then later by GPS surveys, she said. Now we are using a satellite-based technique called InSar. “Essentially, the idea is that a satellite goes over the same area and takes multiple images. We take two or more images and process them together, and it makes a change map that is sensitive to vertical change so we’re able to image subsidence from space,” she said. “Sometimes, depending on the conditions, we can see anywhere from 5 millimeters of change, although 10 millimeters change was more the resolution for the San Joaquin Valley because it’s very agriculturally active, and you can imagine trying to image 5 mm of change in an agriculturally active area where they till the land and things grow. We have a lot of extra filtering to do with agricultural lands that makes the resolution not quite as good as in desert areas or urban areas.”
The USGS has also been refurbishing extensometers which Joe Poland installed and used to measure subsidence originally. “We’ve learned a couple of things since the 50s and 60s about how to measure aquifer system compaction using extensometers, so we made a few design upgrades and are now measuring aquifer system compaction at four locations in the San Joaquin Valley by retrofitting these 50 year old extensometers.”
She then presented a slide showing the location in the San Joaquin Valley where subsidence has historically occurred, the area in the original photograph by the telephone pole. “This shows a map of where that subsidence occurred. … The brown areas indicate where there is subsidence between 1926 and 1970, and you can see the darker areas here are near the California Aqueduct.”
The subsidence continued until 1970 when the well levels started to recover and subsidence started to cease. “It slowed down, and then it ceased,” she said. “This is because the California Aqueduct was put online and it started to deliver water in 1970, and so we saw an abatement of subsidence essentially because now they didn’t have to pump as much groundwater and instead they were using water from the California Aqueduct.”
“This recovery occurred until a drought, and because there wasn’t enough water coming down the California Aqueduct, the groundwater use was increased to make up for the deficit,” she said. “Subsidence was reinitiated; there was about 150 feet of water level decline in just a couple of years. After the drought, California Aqueduct deliveries went back to what they were, or similar to, and groundwater levels again recovered. Subsidence stopped until the next drought in the late 80s, and then we saw subsidence reinitiate. So essentially we found that in areas of historical subsidence, that the problem was essentially arrested except during droughts because less water is coming down the canal, and groundwater pumping was increased to make up for that deficit.”
She then presented a slide and noted that the location of the red circle is where they are seeing subsidence recently. She said there was concern in recent droughts about subsidence on the west side where historically it had occurred historically, and even more so now in the current drought. “As it turns out, subsidence is not just a problem during droughts as it has been in the past because we are starting to use land in a different way,” she said. “Where there isn’t any access to surface water, whether it’s the wettest year on record or a drought, they pump pretty much the same all the time, regardless of climatic conditions, so it’s just not a problem during droughts for some areas that don’t have any access to surface water.”
She then presented a graph of groundwater level and subsidence plotted over time for the location in the previous slide, noting that the red portion of the graph is a continuous GPS site near Mendota. “It was pretty flat until 2007 during the drought and we had some subsidence, and then from 2010 to 2012, it essentially flattened out, still subsiding a little bit, but at a greatly reduced rate, and then during the latest drought, we can see that subsidence has reinitiated.”
“At the same time here, we’re looking at groundwater levels,” she said, pointing out the thin blue line on the graph. “This is in a deep well, we can see that groundwater levels were recovering, but then in the last drought, 2007-09 … You can see groundwater levels declined, and in the last drought, 2007-09, they actually got below the previous historic lowest level, which is the pre-consolidation stress, and that was set during the last year of the previous drought in 1992 – it was about 155 feet or so. We saw in 2009 that we went deeper than that, and again in 2013.”
One of the main reasons we are concerned about subsidence is because of its impact to infrastructure, Ms. Sneed said, presenting a map showing the area of new subsidence. “The impacted area is in the red circle, and you can see that the California Aqueduct is on the edge of that, the Delta Mendota Canal is in there, and so is the Eastside Bypass, other local canals, and also the San Joaquin River in the restoration area. So there’s a lot of infrastructure in this new subsidence area.”
She then presenting a slide of the results from the InSar satellite technique. She noted that the focus of the study was on the Delta Mendota Canal which is shown in red on the left. “We had to choose our data to make sure to cover the DMC and be efficient at the same time. We could see that most of the Delta Mendota Canal looks pretty stable – some subsidence a little bit, but relatively stable until you start to get to the lower stretches of this canal, and you can start to see some yellows and greens come in here on the edge … We could see we were on the edge of something, but our focus was the Delta Mendota Canal.
She noted that the two images were from slightly different time periods. The northern image is from 2007-2010, and results showed about 3” near the Eastside Bypass, which was near the edge of their data and far from the Delta Mendota Canal. On the southern image, from 2003-2008, there was about 6” during that time near the Eastside Bypass. “It looked like we were on the edge of something big, but we had limited resources so we focused on the Delta Mendota Canal for the time being,” she said.
“Here are the results that we found,” she said, presenting a slide with two graphs that depicted subsidence along the Delta Mendota Canal. On the x axis is subsidence, and on the y axis are the check stations for the Delta Mendota Canal. The portion circled in yellow on the graph corresponds to the area circled on the map of the Delta Mendota Canal; the upper graph covers check stations 1 through 17, and the lower graph depicts 17 through 21. “It looks like check stations 1 through 10 are fairly stable, but what really gets your attention is between check station 16 and 17,” she said, pointing to the end of the upper graph.
Moving to the lower graph, she pointed out the subsidence between checkpoints 17 and 18, and between 20 and 21. “There’s a lot of differential subsidence here between 17 and 18, a lot of differential subsidence between 16 and 17, and also between 20 and 21,” she said. “Those are where we figured there might be some problems on the canal, based on this data. More or less fortunately, checks 18, 19, and 20 are subsiding about the same, so they are more or less maintaining gradients in that area. That is sort of good news for the Delta Mendota Canal water authority and the Bureau of Reclamation.”
“However, about a month ago, I got a call from the Delta Mendota Water Authority saying that they were having problems pushing water past check 7,” she said, pointing to the edge of the yellow circle on the upper graph. “Canals are very sensitive to subsidence; you need elevation and gradients maintained between the check stations so that the canal can flow and deliver water as designed. But just this little bit of subsidence upstream … they were having a hard time getting it past this check station here, because it had to essentially go uphill as that gradient was disrupted. … They had problems pushing past check 7, and so a very short-lived opportunity to try and fill up San Luis Reservoir somewhat was missed because the capacity was impacted at check 7.”
“We couldn’t help but look to the east a little bit after we got a call from the Department of Water Resources who had a consultant do a GPS survey in this area and they couldn’t believe the results of their surveys,” Ms. Sneed said. “They did them 2 years apart, in 2008 and in 2010, and they were finding remarkable difference in elevations between those two times at the same benchmarks they were surveying.”
She said they did some quick work to see if they could confirm the survey results. “Not only did we confirm the results, but we found this very large subsidence area that was covering 1200 square miles,” she said. “We not only confirmed that they were seeing about a foot a year of subsidence between 2008 and 2010, but essentially the subsidence bowl went all the way from I-5 to 99 … this is starting to impact that far west, and this is impacting the Delta Mendota Canal. We did not expect to find subsidence here.” She noted that the subsidence is occurring in the area of the Eastside Bypass and the San Joaquin River, which is a concern. “The Eastside Bypass is the most important flood control channel east of the San Joaquin River, and that’s severely impacted by subsidence here.”
She then presented a close-up of the area between the Eastside Bypass and the San Joaquin River, noting that the Eastside Bypass transects the area of subsidence on the north and the San Joaquin River transects the area on the south.
She next presented a map of the Eastside Bypass and noted that the water flows from A’ northward to A on its way out to the Delta for flood control. [Note that on this map, water is then flowing northward.]
She then displayed a profile for the Eastside Bypass, noting again that water flows from A’ to A, which would be right to left on this graph. “As water is flowing down, there’s a hole and then this very large, essentially depression … it’s going to have to fill this area up before it continues flowing down the canal, so it’s going to flood this and flood all points at lower elevations than it, so the Eastside Bypass is certainly impacted by subsidence, and it will not be able to move water out like it used to be able to.”
She then presented a slide showing how the rate of subsidence changed at one of the locations on the East Side Bypass. The graph plots the rate of subsidence over time. She noted that between 2003-2004, there was 40-45 mm of subsidence, and then it flattened out. The yellow bar indicates a gap in the data, so they assumed no change. “When we get to 2010, we can see that the slope of the lines are much different. This slope is about twice this slope, so the rate doubled in 2008, and this rate of subsidence has continued through 2013, so we’re looking at about a foot a year of subsidence since 2008.”
Ms. Sneed said they also looked at GPS measurements in the area, and while there weren’t any in the specific area, there are three surrounding the area of subsidence: Madera, Los Banos, and Mendota. “They all show something different,” she said. “Up here by Los Banos, it was subsiding a little bit and then the drought came and the rate increases. During the inter-drought period, it’s still subsiding, but it slowed down a little bit, and then during the latest drought, we have additional subsidence at a greater rate than during the inter-drought period.”
“Down here by Mendota, it was really quite flat until 2007; during the drought we had subsidence, and then it essentially flattened out again,” she said. “Then in the next drought, we see subsidence kick in again.”
“Over here in Madera, it’s a pretty steady rate during droughts and in between droughts with slight increases during drought,” she said. “This shows us something interesting. We notice that these two, P303 and P307, have subsidence even when there isn’t drought. Down here by P304, we really only see subsidence during drought, and that’s a reflection of access to surface water. At P303, they have very little access to surface water, so it doesn’t matter how much is coming down the California Aqueduct or the Delta Mendota Canal; they have to pump the same because they don’t have access to that water.”
“It’s the same with P307 in Madera, where they don’t have access to a lot of surface water so it really doesn’t matter how much is being delivered to anyone else; they still have to pump the same,” she said. “Whereas P304 is the end of the Delta Mendota Canal; they do have access to surface water supplies and that’s why there is subsidence only during periods where the volume of surface water is reduced.”
She then presented a map, and noted that the light green indicates the area they are mapping now, which is northeast of the area where subsidence has historically occurred. She also noted that the red circle is an area that they’ve just started looking at in the last couple of weeks. “The Delta Mendota Canal is right between these two areas, so historically it was being impacted that was occurring to the south, and more recently, it’s being impacted by subsidence occurring to the north,” she said.
She then presented a graph of a cross-section off the San Joaquin Valley aquifer system noting that there’s an unconfined aquifer system and a confined aquifer system. “There’s a Corcoran clay confining layer, a very clay-rich unit that was laid down by an ancient lake, and that confines the aquifer system; it is a pressurized aquifer system, a confined aquifer system,” she said.
“Now I’m going to show you some information about the unconfined and confined aquifer system in terms of groundwater levels and compaction,” she said, presenting a slide with two graphs. “In Mendota, we have the same GPS site that I showed you before that is essentially flat except for during droughts,” she said. “Nearby, there’s an extensometer, and it is anchored in the top of the Corcoran clay, so it’s essentially measuring the aquifer compaction from land surface to the top of the Corcoran clay, which at this location is about 400 feet below land surface,” she said. “We can see that the extensometer is measuring a little bit of compaction, but you can see that the GPS station is measuring a whole lot more. Because this is anchored in the top of the Corcoran clay, this shows us how much subsidence is happening or how much compaction is happening in the upper 400 feet. Theoretically this GPS station goes all the way to the center of the earth, so what this tells us is that most of the compaction is happening below the top of the Corcoran clay. Some is happening in that unconfined system, but it looks like most of it is happening beneath the unconfined system.”
The site near Los Banos is another indication that most compaction is occurring in the deeper system, she said. The bottom graph shows the water level in a shallow well near Los Banos. “You can see that the water level in the shallow well has hardly changed at all, while subsidence has occurred at P303, so what’s telling us is that the stress that is driving this compaction is not occurring in the shallow system. It looks like most of the compaction in the Mendota area and in the Los Banos area is happening beneath the top of the Corcoran clay.”
The high rate of subsidence is due to groundwater level declines below historical lows and clay units. “Groundwater level declines and geology – we need both of those things for subsidence to occur,” she said.
In the Madera subsidence bowl near El Nido, the geologic setting is a bit different, she said. “A lot of people think that it’s probably the Corcoran clay that is the big player in subsidence, but it’s really not,” she said. “It’s so tight hydrologically that it drains very, very slowly, so even as its being squeezed, even as those particles are rearranging ever so slightly, it is really slow to respond. The Corcoran clay is not a big problem for us now, but we will probably shake our fists at it in 5000 years and wonder why we didn’t do anything about it today. It does have compaction potential in it, but it won’t be realized for quite some time.”
She then presented a slide with three graphs, noting that the top graph are groundwater levels in the shallow system above the Corcoran clay and the two graphs on the bottom are groundwater levels in the sub Corcoran or in the deep system. She noted that the shallow system has stayed above historic lows but the deep system did not. “The deep system reached historical lows, so what that tells us is that the compaction that occurred in the deep system is permanent; the pre-consolidation stress was surpassed,” she said. “In the shallow system, while there may have been some compaction going on up there, it’s probably recoverable because it wasn’t below that pre-consolidation stress.”
We’ve seen groundwater levels continue to decline, she said. “Our study ended in 2010 but we’re certainly not out of the woods,” she said. “We saw historic lows being reached last summer in this well. We have it instrumented with a submersible pressure transducer, and we anticipate it will reach new historic lows this summer.”
She then presented a visualization depicting the sediments on the valley floor created from the Central Valley Hydrologic Model that digitized about 8500 well logs. She noted that clay is blue and yellow and reds are gravels. “So you can see a lot of clay in this area,” she said. “We digitized well logs to get a handle on the geologic setting and the variability throughout the San Joaquin Valley and the Central Valley as a whole, and we put all of that into a database and came up with a sediment texture for each of the layers in the model. Although this is the top layer here, it really does represent what happens throughout the system as you get deeper. The warmer colors are more coarse grained materials and the bluer colors are more fine-grained, and they are kind of shallow. This is finer grained that the surrounding areas and this is exactly where we are finding this new subsidence area.”
“What we found was that we could almost map the subsidence area by mapping the location of the Fresno River fan and the Chowchilla river fan,” she said. “They are different texture-wise than the fans around them to north and the south because they were never connected to Sierra Nevada glaciations, so they have finer grained deposits.”
“They never had those big pulses of glacial till coming down those rivers; they didn’t have those pulses of coarse grained material coming down these fans that are in between the San Joaquin and the Merced River,” she said, presenting a map of Central Valley rivers. “In the red circle where the Chowchilla fans and the Fresno River fans are located, they were not part of the glaciations so they are finer grained. That said, there may be a quite a lot of potential in the unconfined system to compact and subside but we don’t really have the instrumentation to know how this is working yet. We need extensometers in this area that are shallow so we can differentiate what’s happening in the shallow system compared to the deep system.”
Traditionally the shallow system hasn’t been used much for groundwater pumping and irrigating crops, and the reason is because it has poorer groundwater quality, so most of the pumping has been from the deeper system, she said. “As subsidence has been occurring there, farmers and others are looking for solutions to subsidence, and starting to exercise and use that shallow unconfined system more than it’s been used before. So the impacts of that, until we get instrumentation in place, will really be unknown if most of the compaction occurs to happen in the deep system or if we start to see some in the shallow system.”
“It’s a really interesting area for further research, and we’re looking forward to doing that,” Ms. Sneed concluded.
Online resources …
- USGS Hydrologic Information for California
- USGS Hydrologic Studies for the Central Valley
- USGS Drought Information for California
See also …
- San Joaquin Valley: Largest alteration of the earth’s surface. by the USGS
- Land subsidence from groundwater use, report from the California Water Foundation
For more information …
This post is derived from a webinar from the series, Insights: Water & Drought Online Seminar Series, produced by the University of California Agriculture and Natural Resources.
- To watch this webinar, click here.
- To view a list of all available webinars in this series, click here.
Sign up for daily emails and get all the Notebook's aggregated and original water news content delivered to your email box by 9AM. Breaking news alerts, too. Sign me up!