Rivers are apparently not warming due to increased atmospheric warming

... at least not in all Pacific Northwest rivers, according to available data.

In a recent paper ("The paradox of cooling streams in a warming world: Regional climate trends do not parallel variable local trends in stream temperature in the Pacific continental United States"), Arismendi et al provide a description of what's happening to stream temperature trends vis-à-vis air temperature trends. (That's pretty obvious, because it's in the title of the paper; yay to the descriptive quality of scientific papers.)

In the paper, Arismendi et al describe the importance of water temperature in stream ecology, and provide a brief review of the theory of stream warming and a critique of the quality of existing corroborative, data-driven research:

Observed warming in air temperature (between 0.8 to 2.1°C for the first half-decade of the 21st century relative to the period 1950-1980) [Hansen et al., 2006] and changes in streamflow timing and magnitude [Mote et al., 2005; Regonda et al., 2005; Luce and Holden, 2009] have been hypothesized to lead to increases in the magnitude and variability of stream temperature. Several studies have noted increasing temperature of streams. However, these have been based on data from streams that include those altered by human influences, including impoundments and water withdrawals [Kaushal et al., 2010; Mantua et al., 2010], or through inferences and correlations derived from air-water relationships [Mantua et al., 2010; Isaak et al., 2011]

The authors point to an interesting trend in the stream temperatures of 63 sites in northwestern Pacific coast of the US: some are cooling, even as the air temperatures are rising even while others are warming, but most showed no variability in water temperatures. In the end, few stream sites conformed to the initial hypotheses of both (A) increased temperatures and (B) increased variability in temperatures.

In their conclusions, the authors point to a variety of potential causes:

• Variability in stream shading over the record period
• Even though correlations between baseflow and riparian vegetation seemed to show counter-intuitive results
• Variability in results derived from long vs. short record periods
• Longer term records are more robust, but there are very few long-term records for minimally impacted streams.
• Local driving factors that overpower any theoretical relationship between warming air temperatures and increasing stream temperatures

The authors conclude that there need to be far more stream gaging stations, and - as someone who likes greater amounts of data - I have to agree. Still, I wonder if the study could have been improved by - instead of asking for greater amounts of stations in the Pacific Northwest - looking at a greater number of stations across the nation. (I mean, the USGS provides a lot of data for many hundreds of stations around the country.)