Full disclosure: I just happen to be an employee of the U.S. Geological Survey so I probably know more about this than the average Joe. This entry represents my opinion only, and not the opinion of the USGS.
My sister sent me this link today. It's a blog entry by Jeff Masters, one of the co-founders of the popular Weather Underground site. Jeff reports on recent record flooding in Oklahoma City, where 8-11 inches of rain fell yesterday. This flood left 136 people injured.
Four days earlier, at least twenty people lost their lives Caddo Gap, Arkansas when a sudden flash flood wiped out a campground in the early morning.
Both of these extreme weather events probably had global warming contributing to the disaster. After all, the more heat in the atmosphere, the more water can be absorbed and at some point fall. Meanwhile, hundreds of stream gauges nationwide maintained by the USGS, and which are used by NOAA to predict floods, are being cut nationwide, including in both Arkansas and Oklahoma. This pennywise and pound-foolish approach may result in further loss of life.
The first thing to understand is that the USGS does more than detect earthquakes. Water is also a huge part of its mission and it monitors the thousands of rivers and lakes across the country. The USGS has this site where, if so inclined, you can look up water levels of streams and lakes near you or across the country.
Unfortunately, the way the law is written, most of these stream gauges require some sort of matching funding, usually from states, counties and other local cooperators with a vested interest. Of course, states are currently under enormous budgetary pressure. So they are cutting their funding for stream gauges. As Jeff says in his blog:
With over 50 people dead from two flooding disasters already this year, now hardly seems to be the time to be skimping on monitoring river flow levels by cutting funding for hundreds of streamgages. These gages are critical for proper issuance of flood warnings to people in harm's way. Furthermore, most of the northern 2/3 of the U.S. can expect a much higher incidence of record flooding in coming decades. This will be driven by two factors: increased urban development causing faster run-off, and an increase in very heavy precipitation events due to global warming.
To get an appreciation for the size of the problem, look at this map of discontinued or threatened stream gauges on the USGS web site:
As Jeff says:
According to the USGS web site, river stage data from 292 streamgages has been discontinued recently, or is scheduled for elimination in the near future due to budget cuts. In Tennessee, 16 streamflow gages with records going back up to 85 years will stop collecting data on July 1 because of budget cuts. Five gages in Arkansas are slated for elimination this year. Hardest hit will be Pennsylvania, which will lose 30 of its 258 streamgages.
You don't need a rocket scientist, or even a hydrologist to see the problem here. To make predictions of the future, we need to know what happened in the past. By removing these stream gauges, the country is losing a valuable historical information. In the short term of course you need to know what very recent water levels are and how fast they are rising or dropping to make flood predictions. Moreover, since NOAA uses USGS stream gauge data for flood predictions, by retiring these gauges, it's like we are choosing to make ourselves blind, or at least worsen our vision.
Stream gauges are doubtless not cheap but most can be maintained for a year for less than the cost of a new compact car. They have to be maintained and monitored, which means even after they are put in the USGS has to send out hydrologist periodically to service the station. Some 9,000 of these gauges periodically broadcast recent streamflow information. If all these gauges are discontinued clearly the impact will be felt.
You can see a graph of the event in Arkansas on the Caddo River here:
At least 20 people died on the Caddo River. Fortunately, the gauge on this river is not one of those scheduled to be cut. I do not know if NOAA issued a flood warning for the river. Perhaps the rain fell too quickly to give a proper notification, or no one was available on site monitoring the river to give warning. It does appear that procedures could be improved that might have saved some or all of the lives lost on the Caddo River.
One thing is for sure: you cannot save lives without data, and NOAA cannot create flood forecasts without data from various stream gauges that are out there. I know money is tight, but with states cutting their funding for these stream gauges, in a way it's like they are telling their citizens they are getting out of the public safety business.
Shame on 'em.