Northwest Fisheries Science Center

Coastal Upwelling

An important process affecting primary productivity during the spring and summer off the Pacific Northwest is coastal upwelling. Upwelling is caused by northerly winds that blow along the Oregon coast from April to September.  These winds transport offshore surface water southward (orange arrow in Figure CU-01), with a component transported away from the coastline (to the right of the wind, light green arrow).  This offshore, southward transport of surface waters is balanced by onshore, northward transport of cool, high–salinity, nutrient–rich water (dark blue arrow).

Diagram of factors affecting coastal upwelling Figure CU-01. Forces affecting coastal upwelling.

The strength of an upwelling process can be calculated based on estimates of wind speed.  Using such data, Dr. Andy Bakun (1973) developed the coastal The next link/button will exit from NWFSC web site Upwelling Index

The Upwelling Index is, as its name implies, a measure of the volume of water that upwells along the coast; it identifies the amount of offshore transport of surface waters due to geostrophic wind fields.  Geostrophic wind fields are calculated from surface atmospheric pressure fields measured and reported by the U.S. Navy The next link/button will exit from NWFSC web site Fleet Numerical Meteorological and Oceanographic Center (FNMOC) in Monterey, California. 

The Upwelling Index is calculated in 3–degree intervals from 21°N to 60°N latitude, and data are available from 1947 to present.  For the northern California Current, relevant values are from 42, 45, and 48°N.  Year–to–year variations in upwelling off Newport (45°N) are shown as anomalies of the upwelling index Figure CU-02.  The years of strongest upwelling were 1965–1967.



Anomalies of the coastal Upwelling Index during May to September each year, 1946 - present Figure CU-02. Anomalies of the coastal Upwelling Index during May to September each year, 1946 - present

Upwelling was anomalously weak in all but 8 of the 21 years from summer 1976 to summer 1997, and this is expected during warm PDO phases.  When the PDO was in a cool phase (late 1998–2003), upwelling strengthened.  With the change in PDO sign to positive in 2004–2005, upwelling again weakened.

Many studies have shown correlations between the amount of coastal upwelling and production of various fisheries.  The first to show a predictable relationship between coho survival and upwelling were Gunsolus (1978) and Nickelson (1986).

The relationship of spring and fall adult Chinook salmon returns to the Bonneville dam and coho salmon survival (%) with upwelling is shown in Figure CU-03. The relationship is weak for spring Chinook and coho salmon and there is no relationship with adult returns and upwelling for fall Chinook salmon. Although the relationships are weak, the strongest correlations with survival were found with upwelling in April and upwelling in April and May combined. A significant, but weaker correlation was also found between upwelling and survival during the months of April, May, and June combined.

Figure CU-03. Relationship of spring and fall adult Chinook salmon returns to the Bonneville dam and coho salmon survival (%) with the April upwelling anomaly at 45°N from 1969-present. The relationship is weak for spring Chinook and coho salmon and there is no relationship between adult returns of fall Chinook salmon and upwelling.

Scheuerell and Williams (2005) showed that the upwelling index in April, September, and October is also related to returns of Snake River spring Chinook salmon. Moreover, they developed a 1-year forecast of spring Chinook salmon returns based on this composite upwelling index.

Knowledge of upwelling alone does not always provide good predictions of salmon returns.  For example, during the 1998 El Niño event, upwelling was relatively strong, as measured by the upwelling indices; however, plankton production was weak.  This occurred because the deep source waters for upwelling were warm and nutrient–poor.  Low levels of plankton production may have impacted all trophic levels up the food chain. 

Upwelling was also strong during summer 2006, yet SST anomalies only averaged −0.3°C.  On the other hand, upwelling was relatively weak during the summers of 2007 and 2008, yet these summers had some of the coldest temperatures in the time series, −1.0°C.  These observations demonstrate that some care is required when interpreting a given upwelling index.  We hypothesize that although upwelling is necessary to stimulate plankton production, its impact is greatest during negative phases of the PDO. 

Upwelling in 2013Figure CU-04 illustrates the pattern of upwelling through the use of a cumulative upwelling plot. This method simply adds the amount of upwelling on one day to that of the next day, and so on. The plot begins with day 1, on 1 January. Due to "downwelling" during winter months, upwelling values are increasingly negative for several weeks after day 1. But with the onset of the spring transition and upwelling, the downward trend reverses, and the cumulative line trends upwards.

Figure CU-04. Cumulative upwelling index for 2013. Vertical arrows indicate the date of physical spring transition (Day 97) and fall transition (Day 261) and the black bars highlight relaxation events. This year had an average start date for upwelling, but the overall "season" ended 20 days earlier than usual.

One can see in Figure CU-04 that there was very little downwelling through the first three months (90 days) of the year. Upwelling started on day 97 (07 April). The winds started out strong, but relaxed throughout much of May and June, with a majority of the upwelling occurring during July. The total amount of upwelling for 2013 was 6,265 m3/s per 100 m of coastline, which is nearly the same as the 40-year average of 6,163 m3/s per 100 m.