I find in my work that I must consistently counter the popular notion that a warming Western North America is leading to earlier peak streamflow runoff.  It should be relatively straightforward, if only as a stochastic hydrologist, for me to explore this notion and to reduce it to simple reproducible experiments.

The featured image accordingly is intended to capture the exploration of a peak flow for a representative stream in the western United States.  In this instance the focus is on the monthly average record for the Animas River at Farmington, New Mexico, whose headwaters reach deeply into the highest peaks of the San Juan range of the Southern Rocky Mountains of the western US.   I am currently working on an Animas paper for a related conference proceeding sponsored by the New Mexico Water Resources Research Institute (WRRI) based at New Mexico State University at Las Cruces, NM.

In that work I’ve already presented, I’ve run an array of hydrostatistical and performance measures to the time series records, both measured and modeled, of that and related streams.  Here in the featured image I have chosen in part to unroll the specific Animas monthly streamflow data and use Microsoft Excel’s cell color ranking formatting features to quickly detail the deep-green band.  Only that band traces through each month of peak flow.  I’ve added a supporting chart within that image which identifies the peak flow month in a more conventional mode.

The apparent conclusion must be that across this basin there is little or no evidence of a descent of the peak flow month for the full complete available history of the stream.  This result supports an alternate argument that there is no evidence of an earlier spring than normal.

Moreover, there are now excellent arguments that the hydrologic runoff is strongly and significantly correlated across most if not all seasons and years to well mapped ocean indexes.  The next image replicates related items that I and others have explored for approximately the past decade or so.  In this chart the actual (versus “peak rank” of the previous image) Animas flows are simply compared to the Pacific Decadal Oscillation (PDO) over an available continuous Animas record span dating back to the first half of the 20th century.

The Animas and its neighbor basins cannot be highly correlated to both the PDO and to an early spring.  As a result, I have explored the early spring notion in somewhat more detail thanks to a paper [1] I was pointed to by a colleague.  Accordingly in the next attachment I have overlayed the Animas and PDO time series across one of that author’s plots of a time series of snow water equivalents (SWE) values from multiple SNOTEL sites across a similar study area to mine:


It would not be a surprise to many hydrologists that the relative peaks and dips are fairly synchronous.  This is excellent independent confirmation I feel of the SNOTEL values but is only an informal part of an informal exercise.

I don’t mind repeating that the attachment appears to demonstrate that the Animas example and others are consistent with the SWE outcome.   However at least some papers covering the SWE outcome, and which also assert an early spring, do not incorporate the PDO.  It accordingly may be that the early spring notion largely predates more widespread recognition of the influence of the PDO.  In any case, I expect that early spring notions will soon be eclipsed as understanding grows regarding the potential for more accurate ocean and solar based forecasts.

Thanks to other colleagues who have engaged in private conversations with me.

[1] Clow, D. W. 2010. Changes in the timing of snowmelt and streamflow in Colorado: A response to recent warming. Journal of Climate 23(9): 2293–2306.