9 Comments

  1. John, speaking of forecasting and historical trends: is there an obvious reason for the apparent periodicity in Lake Powell annual inflow? It seems apparent on the “Yearly Cumulative Inflows” chart on
    http://graphs.water-data.com/lakepowell/index.php

    (You have to scroll down and select the chart from a pull-down menu.)

    Are they ENSOs, or (since it seems to be an interval between peaks of about 11 years) something to do with the solar cycle? Has this been observed on the Colorado River (and others in the SW?) before this chart’s beginning in 1964?

  2. David A – Great question re the periodicity. Without actual research, I think the answer is that the apparent periodicity disappears if you look at a longer time series:

    http://www.usbr.gov/lc/region/g4000/NaturalFlow/current.html

    The “natural flow” data is a calculation of the flow at various places on the river back to 1906, based on USBR’s estimates of what that flow would have been absent upstream dams and diversions. That makes it a much better measure than the Lake Powell data. It has a similar look to the Powell data you cited in recent years, but not so much farther back in time. (You want the Lee’s Ferry column on the spreadsheet if you want to go take a look.)

    ENSO is a poor predictor for Colorado River flow because the basin spreads across such a large area north to south. Most of the snow-producing region is in Colorado and to the north, the area where the ENSO precip signal is weak.

    AE Douglass, the guy who invented tree ring stuff a century ago, was actually an astronomer looking for a deeper time proxy for the sunspot cycle. All the other benefits of tree ring research (archaeology, climate) were a happy accident. But the sunspot cycle signal in the tree ring data turns out to be incredibly weak.

  3. With regards to the regular cycle of up and down with inflow, I wrote a paper about this in 2005

    http://journals.ametsoc.org/doi/pdf/10.1175/JHM410.1

    “A Recent Increase in Western U.S. Streamflow Variability and Persistence”

    Since about 1980, river flows across a large area (in Colorado but also Wyoming through up to Idaho) have settled more into a “sine wave” type pattern like you suggest. Previously it was more like random noise, one year would be high, the next would be low, there was no pattern. In other words, this wet-wet-wet-dry-dry-dry-wet-wet-wet cycle this is a new thing.

    But it’s not a good thing for water managers. From the paper: “A period of high persistence, variance, and skewness is perhaps the most challenging scenario for water managers. For example, a series of consecutive wet years (e.g., Fig. 5, 1983–86) may overwhelm reservoirs and inflate stakeholder expectations about the amount of water available. An extended stretch of dry years then exhausts storage reservoirs and does not give them a chance to recover (e.g., 1987 94). Smaller reservoirs that do not have multiple-year storage capacity would be especially vulnerable. In comparison, individual dry years interspersed among wet years are much more tolerable.”

    There’s subtle and technical reasons why this scenario can make the accuracy of river forecasts go down, as well (details on request and in the paper).

    This creates double headaches for water managers- not only is the river being more variable, but it’s also faking water managers out before the flows come.

    20 years from now are we still going to be in this series of persistent up and downs? To know that, you’d need to know why it was happening in the first place. I only got as far as documenting that it was happening, not the cause. I looked at the last 100 years, some people have been doing tree ring research to go back even further.

    It’s possible there could be no cause! Even if the flow was random, you’d expect a couple wet years to pile up in a row now and then. Naturally people want to look for patterns but that leads to things like the “Hot Hand Fallacy” http://en.wikipedia.org/wiki/Hot-hand_fallacy

    Happy reading!

    Tom

    http://tompagano.blogspot.com
    The River Seers

  4. Thanks a lot — you’re right, I plotted the Lee’s Ferry annual flow and there’s no apparent periodicity prior to about the 1984 peak. The linear trend is only -32,000 acre-ft/yr per year, and not a very good fit anyway, but still that would be about -16% in a century if that trend continued.

    Do you know what occurred in 1984 to cause the big peak in annual flow?

  5. Pingback: Another Week of GW News, April 22, 2012 [A Few Things Ill Considered]

Comments are closed.