I’ve been digging into the relationship between climate and the early history of water development of the Colorado River, and I’ve run into a bit of a conflict between the conventional history and the climate record (or as much of the climate record as I’ve been able to easily sort out).

(Click through if you’ve any interest in obscure western history or you think you’d be amused by my confusion. The rest of you move along, nothing much to see here.)

The story, in brief, involves one of the seminal events in the Colorado River’s develoment. In 1905, the river busted through an irrigation work that had been built for the Imperial Valley. For nearly two years, the river flooded into the lowland then called the “Salton Sink,” now known as the “Salton Sea.” In fact, that’s how it became the Salton Sea. Here’s how Norris Hundley describes it in Water and the West, the canonical history of the development of the Colorado River:

Because high water had seldom been a threat in previous winters and because his company was in financial straits, in 1904 Rockwood had failed to provide his new Mexican intake with an adequate headgate. Unfortunately for him, 1905 proved to be an unusual year. Flood waters began rising in February, gouging away at the banks surrounding the cut faster than he and his men could fill the breach with pilings and sand bags. Five floods eventually hit during the winter and spring until by August 1905 the entire river was pouring into the intake, now a half-mile wide at its juncture with the Colorado. In a matter of weeks, much of hte Salton Sink becase the Salton Sea.

Here’s the problem. According to flow data from the records of the Upper Colorado River Commission, total flow in 1905 (the water year ends at the end of September) was 16 million acre feet. That’s above the long term average of 14.8, but only a tad above the 1904 flow of 15.6 maf. 1903: 14.8. And in the Colorado River Drainage Basin climate division data, precipitation that year (the winter of 1904-05 or the full water year) was below average.

Now, I don’t have monthly or daily flow data, so it’s possible that the floods came in big pulses that are not reflected in the total flow volume. Or is it possible that the conventional story is wrong, and that these guys just fucked up badly and the “unusual year” story was actually a shifting of blame?