Are river crests getting higher?

The Susquehanna in Wilkes-Barre crested (that is, reached its highest level) during last week's flood at a height of 42 feet, eight inches - a new record. Are river crests getting higher, and is this a natural consequence of development and deforestation?

Prompted by the discovery by amateur historian Philip Lord of a photo tucked into an old book, I wrote this post that looked at the history of flooding in Wilkes-Barre from 1936 through 2010:

Another Monkey: Sterling Hotel and Market Street Bridge, 1936 and 2010

In that entry I included this tidbit:

From the Luzerne County Riverfront Project site:

Although numerous floods occurred in the Wyoming Valley, and some levees had been constructed to try to prevent wide-spread flooding, the valley was unprepared for the flood that struck on St. Patrick’s Day, 1936. That day the Susquehanna crested in Wilkes-Barre and Kingston at 33 feet, and flood waters flowed for miles across the Wyoming Valley.

The devastation wrought by the 1936 flood brought construction funding from Washington in the form of a valley-long, flood protection levee – at a flood stage of 36 feet. Subsequently, the Susquehanna River rose to flood stage in 1946, 1955, and 1964, with the levees providing substantial protection.

And all was well for thirty-six years until 1972, when Agnes rose above the level of the levees and broke thorough the sandbag dikes that had been built around it, reaching a crest just shy of 41 feet. Twenty-four years later there was again widespread flooding, but this flood served as the final kick to get the levees raised to a nominal height of 41 feet, plus three bonus feet. This project was completed ten years later, and was tested by Hurricane Ivan a few months later.

No other storm came close to overtopping the levees until on Labor Day 2011 and much of the week that followed the remnants of Tropical Storm Lee dumped many inches of water onto lands already saturated by Irene a week and a half earlier. Lee then parked itself over upstate New York, like Agnes, and fed the headwaters of the Susquehanna until the levees were put to the test again with a crest of approximately 42 feet 8 inches. (I love the reports that quote this as "42.66 feet" - two decimal place accuracy from a system that had been off by about 7% just a few hours earlier, when the river gauge was overwhelmed by floodwater and reported a crest of just over 39 feet.)

I don't know if it's reasonable to expect to see a "worst flood" and then have every subsequent flood back off from there. But it seems like we're seeing once-in-a-century storms every 35-40 years or so. Had there ever been a 42 foot crest before 1936? Or are higher crests a result of some evolution of the ecosystem that is causing more water to be dumped into creeks and streams that feed the river?

I can think of two potential causes of increased runoff: deforestation, especially in mountainous areas (which will also serve to increase silting-up of creeks and streams), and removal of water-absorbing land surface area through construction and paving. Could these development activities be the cause of increased river crests during catastrophic floods? If so, we should not expect the levees - at 41 or 44 feet, as the case may be - to protect Wilkes-Barre, Kingston, and Forty-Fort next time, or the time after that, or any of the times after that.


Note: I posed this question to Don Williams, the Susquehanna River Sentinel, and he concurred, and dug out a letter to the editor that he submitted to a local paper eight years ago, a letter which he located only with tremendous difficulty. He has reproduced this letter (formatting marks included) in this entry.