SF-0018Strange America
The USGS calls 1905 the Salton Sea's most recent formation, not its first
The Colorado broke through a half-finished irrigation headgate in 1905 and took until 1907 to put back. But the river had filled that basin on its own before, and each time the basin dried out again.
The east shore of the Salton Sea in 2010. The swimming signage stands well back from water that has since withdrawn across the pale flat between them.
Receding shoreline on the east side of the Salton Sea, California.
- Image
- Not stated. Published by the US Geological Survey; the item page names no photographer.
- Taken
- 2010-07-02
- Location
- East shore of the Salton Sea, California
- Archive ref.
- USGS media library — "Receding Shoreline", 2 July 2010
- Licence
- Public domain per the USGS media library's item-level Sources/Usage statement (licence record)
- The image itself
- Authentic, unaltered
"Receding shoreline on the east side of the Salton Sea, California", 2 July 2010. US Geological Survey media library. Public domain. Cropped.
Verdict
"California's biggest lake is a mistake — it only exists because someone broke a canal in 1905."
The core of the claim holds up, but a specific detail as it circulates is wrong or overstated.
The event is real; the framing that it created a lake from nothing is not.
TRUE: the USGS describes the 1905 failure directly. "In 1905, the company lost control of the river during a flood, and the Colorado broke through the half-finished headgate of an irrigation ditch", and "It took engineers and work crews until 1907 to return the river to its proper course, by which time a considerable lake had formed."
NOT TRUE AS FRAMED: that this was unprecedented. The USGS's topic page says the sea "was most recently formed in 1905" — most recently — and its EROS description gives the older pattern: "once in a while, the Colorado would break out and pour down into the Salton Basin, partly filling it. Then the levee break would fill with silt, the river would revert to its normal channel, and the basin would dry up again."
So the 1905 accident did something the river had done on its own. What is genuinely unusual is not the filling but the persistence, and this file is careful about that: the USGS calls the Salton Sea "a terminal lake (it has no outlets)" and describes it now as "a shrinking hypersaline lake" whose bed "dries to a dusty playa". The basin is doing what the record says it always does.
Mostly True requires 1 primary or authoritative source. This file cites 3.
In short
The Salton Sea is usually introduced as a purely man-made accident — a lake that exists because engineers lost control of the Colorado River in 1905. The engineering failure is real and the USGS describes it plainly. What the shorthand loses is the agency's own qualifier: it was "most recently formed" in 1905. The Colorado had broken into the Salton Basin repeatedly before, filling it and then abandoning it, and the basin dried out every time. It is a terminal lake with no outlet, and it is shrinking now.
What happened in 1905
The Colorado Development Company had been cutting irrigation channels to bring Colorado River water into the Imperial Valley. The USGS describes what went wrong in one sentence:
In 1905, the company lost control of the river during a flood, and the Colorado broke through the half-finished headgate of an irrigation ditch.
The river went where the ground fell away — north, into the Salton Basin, which is below sea level and has no outlet. Putting it back was not quick. “It took engineers and work crews until 1907 to return the river to its proper course, by which time a considerable lake had formed.”
That is the story as it is usually told, and every part of it is true.
The adverb
Here is how the USGS’s topic page introduces the same event:
It was most recently formed in 1905 when the Colorado River broke through an irrigation canal and flowed uncontrolled into the Salton Sea watershed.
Most recently. Not was formed — was most recently formed.
The EROS description says what the earlier formations looked like:
once in a while, the Colorado would break out and pour down into the Salton Basin, partly filling it. Then the levee break would fill with silt, the river would revert to its normal channel, and the basin would dry up again.
Break out. Fill. Silt up. Dry. Repeat.
In 1905 a company put a half-finished headgate in the way of a river that already did this, and the river did it again.
Why it stayed
The basin is a trap. The USGS calls the Salton Sea “a terminal lake (it has no outlets)”, which means water leaves it in exactly one way: upward. Every previous filling ended in evaporation, and so will this one.
The agency’s description of the present is not a lake in trouble so much as a lake behaving normally, slowly: “a shrinking hypersaline lake. As the lake bed dries to a dusty playa, air, water, and wildlife managers will depend on careful monitoring and research.” Freshwater inflow, it notes, has been reduced by changes in regional water management.
The sign in the photograph was put up for swimmers. It is standing on gravel.
What this file does not claim
That the flood was natural, or excusable, or anything other than an engineering failure — the timing, the works, and the two-year repair are all human. That there were a specific number of earlier fillings, or when they were, or how big: the sources say “once in a while”, so this file says “once in a while” and names no ancient lake. What kept the sea full for a century after the breach was closed — agricultural drainage is the obvious answer and no source read here states the inflow budget, so it is left as an open question rather than asserted.
And nothing about what the drying playa does to the people who live around it. That is a health subject, it is live, and it deserves its own file rather than a sentence borrowed from a topic page.
The evidence
The engineering failure is described directly by the USGS: "In 1905, the company lost control of the river during a flood, and the Colorado broke through the half-finished headgate of an irrigation ditch."
Cited sources:S1
It took roughly two years to undo. "It took engineers and work crews until 1907 to return the river to its proper course, by which time a considerable lake had formed."
Cited sources:S1
The USGS does not describe 1905 as the basin's first filling. Its topic page states that the sea "was most recently formed in 1905 when the Colorado River broke through an irrigation canal and flowed uncontrolled into the Salton Sea watershed."
Cited sources:S2
The older pattern is described explicitly: "once in a while, the Colorado would break out and pour down into the Salton Basin, partly filling it. Then the levee break would fill with silt, the river would revert to its normal channel, and the basin would dry up again."
Cited sources:S1
The basin has no way to shed water except upward. The USGS describes the Salton Sea as "a terminal lake (it has no outlets)", which is the physical reason the historical cycle ends in evaporation rather than drainage.
Cited sources:S2
The present condition is contraction, not stability. USGS characterises it as "a shrinking hypersaline lake. As the lake bed dries to a dusty playa, air, water, and wildlife managers will depend on careful monitoring and research."
Cited sources:S2
The inflow that sustained it is being reduced. "Changes in regional water management practices have reduced freshwater inflow to the Salton Sea."
Cited sources:S2
Evidence that cuts the other way
Findings that complicate or argue against the verdict above. They are listed because leaving them out would make the file look stronger than it is.
The accident was still an accident, and this file does not soften that. A private irrigation works failed, the river took the breach, and it required two years of engineering to put back. Reading "the river had done this before" as "so it was natural" would be a worse error than the claim being corrected — the timing, the scale of the diversion works, and the fact that a company was in the river's way are all human.
Cited sources:S1
Nothing consulted here establishes how many previous fillings there were, when they happened, or how large they got. The USGS phrasing is "once in a while" and "most recently", neither of which is a count or a date. A reader should not take this file as establishing a specific prehistoric lake, and it names none.
What we still don’t know
- How many times the Colorado filled the basin before 1905, and when. The USGS says "once in a while" and "most recently", which are not counts or dates. No geological or archaeological source was consulted, and this file names no prehistoric lake and asserts no chronology.
- What sustained the sea between 1907 and now. The obvious answer is agricultural drainage into a terminal basin, and the USGS notes that "changes in regional water management practices have reduced freshwater inflow". But no source read here states the inflow budget, so the file describes the sea as shrinking without asserting the mechanism that kept it full for a century.
- Present dimensions, elevation, salinity and rate of retreat. None of the pages read gives figures, and none is asserted. The photograph is from 2010 and the shoreline has moved since; the caption says so rather than implying the image is current.
- The health and air-quality consequences of the exposed playa. USGS names dust and air management as concerns. This file does not go further: that is a health subject requiring its own sourcing, and gesturing at it from a page that only mentions it would be worse than leaving it out.
Sources
Every source below was read in full before it was cited. Archive snapshots are recorded at research time, because a file whose sources have evaporated is indistinguishable from one that never had any.
S1Tier 1Read in full
The Lake's Formation — Earthshots: Satellite Images of Environmental Change
US Geological Survey, EROS Center
The USGS EROS account of how the Salton Sea formed. Establishes the 1905 loss of control and the breached half-finished headgate, the two-year effort ending in 1907 with "a considerable lake", and — the finding this file turns on — the pre-existing cycle in which the Colorado would break into the Salton Basin, partly fill it, then silt up its own breach and leave the basin to dry.
S2Tier 1Read in full
US Geological Survey
The agency's current topic page. Source of the decisive qualifier — the sea "was most recently formed in 1905" — and of the physical framing: a terminal lake with no outlets, now "a shrinking hypersaline lake" whose bed "dries to a dusty playa", with freshwater inflow reduced by changes in regional water management. Read independently of the EROS page and consistent with it.
S3Tier 1Read in full
Receding Shoreline — USGS media library item
US Geological Survey · 2010-07-02
The image's rights record and caption. Establishes the item-level "Sources/Usage: Public Domain." statement, the date of 2 July 2010, and the caption "Receding shoreline on the east side of the Salton Sea, California." Also establishes, by its absence, that no photographer or affiliation is named — which is what narrowed this file's public-domain basis.
Checked for circularity: S1 and S3; S2 and S3 were confirmed not to derive from one another. Two sources that trace back to the same origin are one source, and counting them twice is the most common way a false claim looks well-evidenced.
How we checked this
What we went looking for, and what we found
Three checks, two of which shaped the file.
First, against overcorrecting into "so it's natural after all". It is not. A company's diversion works failed and the river took the opening; putting it back took two years of engineering. The contradicting-evidence section says so directly, because "the river had done this before" is one short step from excusing the thing that happened, and that step is not supported.
Second, against inventing a prehistory. The tempting move is to name the ancient lake that occupied this basin and give it dates and dimensions. The two USGS pages read here say "once in a while" and "most recently" and nothing more specific, so the file names no lake and gives no chronology, and records that as an uncertainty rather than filling it from memory.
Third, on the image. The Salton Sea's most-shared photographs are documentary and art work with live rights, and the temptation is to find the striking ruin shot and clear it afterwards. The search began and ended inside the USGS media library, and the item chosen turned out to name no photographer — which narrowed the rights basis rather than being quietly ignored.
- We searched for evidence against our own conclusion
- We traced where the claim started
- The claim is stated in its strongest form, not a weak version
- The photograph was verified independently of the story
- We checked whether this was settled after the popular version froze
- Image rights were traced to a stated licence
- We checked what we were leaving out
- The verdict sat before publication, and was re-read cold
- Confidence in the verdict
- high
- Verdict assigned
- Most recent authoritative source
- Researched by
- The Strange File
NOT YET CLEARED. Cooling-off, independent review and owner approval are three controls and none substitutes for another. Cooling-off is a wait, and the wait has elapsed. The verdict was assigned 2026-07-30; the fresh-eyes pass was made on 2026-07-31 by a session that neither researched nor wrote this file, reading the dossier, the evidence, the sources, the verdict reasoning and the rights record cold. Nothing required changing. The headline attributes "most recent formation" to the USGS, which is where the phrasing comes from, rather than asserting it in our own voice.
That pass is cooling-off and nothing else. A cooling-off reader is not an independent reviewer, and this file names no second reviewer — it carries no sensitivity flag, so none is required. The owner has not approved publication, and READY is not permission to publish.
No sensitivity flag. The subject is a landscape and an agency's description of how it formed. The file deliberately does not enter the public-health argument about playa dust, which is live, is about a named region's residents, and would need its own sourcing and its own care — it is listed as an uncertainty and left there.
No individual is named anywhere in this file, including in the image credit, because the source names none.
Part of a collection
The Qualifier Is The StoryThe interesting content is in the clause everyone skims.
Places America LeftTowns emptied by industry, and the sequence the retelling flattens.
Also related
Lead began Picher's relocation. Collapse risk widened it. A tornado hastened it
Oklahoma's 2004 relocation covered only families with children under seven. Then the Army Corps found a third or more of homes had collapse potential. Then, in 2008, an EF4 tornado hit the town.
Kīlauea makes hair out of glass, and it blows into people's gutters
Strands of volcanic glass over a metre long and under a millimetre thick, spun from lava the way sticky dough pulls into threads. The USGS advice for finding some is to leave it alone.