Examining the link between low summer stream flows and climate


A 2023 paper in the Journal of the American Water Resources Association says future low flows in Puget Sound streams could be lower than are projected by climate models that do not reliably account for multidecadal climate variation. The paper is co-authored by Nicholas Georgiadis and Joel Baker of the Puget Sound Institute.

When rains finally returned to Puget Sound last fall, they provided welcome relief for spawning salmon. Drought conditions had left streams at record lows, causing many creeks to become dangerously warm for fish, and in some cases physically blocking salmon from higher elevations.

This scenario is increasingly familiar. Declining summer stream flows are one of many threats to Puget Sound’s salmon, and there is widespread expectation among scientists that a warming climate will only make matters worse. All signs point to a clear and present danger: Snowpack and glaciers are dwindling; rainfall patterns are changing; salmon are dying.

Summer low flows – the lowest flows of the year – provide an index of weather conditions throughout the year. Much of the blame for declining low flows has been attributed to anthropogenic (human-caused) climate warming. However, a 2023 paper in the Journal of the American Water Resources Association emphasizes that anthropogenic warming is not the only cause for concern – other factors should also be accounted. The paper, co-authored by Nicholas Georgiadis and Joel Baker of the Puget Sound Institute, analyzed low flow trends in Puget Sound streams over 80 years leading up to 2015. The clearest pattern was a single oscillation, formed by an increasing trend until about 1960, followed by a declining trend until the 1990s, when flows trended upwards again.

By focusing on streams in undeveloped basins, the authors could rule out local factors that might influence low flows, such as human development and the drawdown of groundwater from increasing numbers of wells. That the observed pattern was climate driven was confirmed by detection of a multidecadal oscillation, similar to that in the flow data, in both precipitation and temperature data from long-term weather stations located near each for the featured streamflow gages. The oscillation in flow data itself tilted slightly downwards over time, a pattern consistent with anthropogenic warming. “But if anthropogenic warming was the dominant factor driving streamflow trends, there would have been no upturn in the 1990s,” the authors observed.

Multidecadal oscillations like those described by Georgiadis and Baker are well known to climate scientists, but their causes are not settled. Some studies invoke ‘internal’ climate variation caused by ongoing physical interactions between the oceans and the atmosphere, some of which oscillate. But a single oscillation does not necessitate an oscillating driver – it could equally result from a sequence of ‘external’ climate-perturbing events, such as intermittent, explosive vulcanism.

The findings are important, the authors say, because they remind us that anthropogenic warming is not the only climate driver we need to better understand. Projections from climate models that do not reliably reproduce these multidecadal patterns (and their interactions with anthropogenic processes) are likely to underestimate the range, and rate of change, of future climatic variation. Remedial measures intended to maintain low flows may fail to protect salmon if they are designed to offset only the effects of anthropogenic warming.

Georgiadis, N. J., & Baker, J. E. A multidecadal oscillation in precipitation and temperature series is pronounced in low flow series from Puget Sound streams. JAWRA Journal of the American Water Resources Association.