The Influence of North Pacific Atmospheric Circulation on Streamflow in the West

  1. David H. Peterson
  1. Daniel R. Cayan1 and
  2. David H. Peterson2

Published Online: 23 MAR 2013

DOI: 10.1029/GM055p0375

Aspects of Climate Variability in the Pacific and the Western Americas

Aspects of Climate Variability in the Pacific and the Western Americas

How to Cite

Cayan, D. R. and Peterson, D. H. (1989) The Influence of North Pacific Atmospheric Circulation on Streamflow in the West, in Aspects of Climate Variability in the Pacific and the Western Americas (ed D. H. Peterson), American Geophysical Union, Washington, D. C.. doi: 10.1029/GM055p0375

Author Information

  1. 1

    Scripps Institution of Oceanographya-024, University Of California-San Diego, La Jolla, Ca 92093-0224

  2. 2

    U.S. Geological Surveyms-496, 345 Middlefield Road, Menlo Park, Ca 94025

Publication History

  1. Published Online: 23 MAR 2013
  2. Published Print: 1 JAN 1989

ISBN Information

Print ISBN: 9780875900728

Online ISBN: 9781118664285



  • Climatic changes—Pacific Area.;
  • Paleoclimatology—Pacific Area.;
  • Climatic changes—West (U.S.);
  • Paleoclimatology—West (U.S.);
  • Atmospheric carbon dioxide.


The annual cycle and nonseasonal variability of streamflow over western North America and Hawaii is studied in terms of atmospheric forcing elements. This study uses several decades of monthly average streamflow beginning as early as the late 1800's over a network of 38 stations. In addition to a strong annual cycle in mean streamflow and its variance at most of the stations, there is also a distinct annual cycle in the autocorrelation of anomalies that is related to the interplay between the annual cycles of temperature and precipitation. Of particular importance to these lag effects is the well-known role of water stored as snow pack, which controls the delay between peak precipitation and peak flow and also introduces persistence into the nonseasonal streamflow anomalies, with time scales from 1 month to over 1 year.

The degree to which streamflow is related to winter atmospheric circulation over the North Pacific and western North America is tested using correlations with time averaged, gridded sea level pressure (SLP), which begins in 1899. Streamflow fluctuations show significant large-scale correlations for the winter (December through February) mean SLP anomaly patterns over the North Pacific with maximum correlations ranging from 0.3 to about 0.6. For streams along the west coast corridor the circulation pattern associated with positive streamflow anomalies is low pressure centered off the coast to the west or northwest, indicative of increased winter storms and an anomalous westerly-to-southwesterly wind component. For streams in the interior positive streamflow anomalies are associated with a positive SLP anomaly stationed remotely over the central North Pacific, and with negative but generally weaker SLP anomalies locally.

One important influence on streamflow variability is the strength of the Aleutian Low in winter. This is represented by the familiar Pacific-North America (PNA) index and also by an index defined herein the “CNP” (Central North Pacific). This index, beginning in 1899, is taken to be the average of the SLP anomaly south of the Aleutians and the western Gulf of Alaska. Correlations between PNA or CNP and regional anomalies reflect streamflow the alternations in strength and position of the mean North Pacific storm track entering North America as well as shifts in the trade winds over the subtropical North Pacific. Regions whose streamflow is best tuned to the PNA or CNP include coastal Alaska, the northwestern United States, and Hawaii; the latter two regions have the opposite sign anomaly as the former. The pattern of streamflow variations associated with El Niño is similar, but the El Niño signal also includes a tendency for greater than normal streamflow in the southwestern United States. These indices are significantly correlated with streamflow at one to two seasons in advance of the December–August period, which may allow modestly skillful forecasts. It is important to note that streamflow variability in some areas, such as British Columbia and California, does not respond consistently to these broad scale Pacific atmospheric circulation indices, but is related to regional atmospheric anomaly features over the eastern North Pacific.

Spatially, streamflow anomalies are fairly well correlated over scales of several hundred kilometers. Inspection of the spatial anomalies of stream-flow in this study suggest an asymmetry in the spatial pattern of positive versus negative streamflow anomalies in the western United States: dry patterns have tended to be larger and more spatially coherent than wet patterns.