204-214 Perez Brunius

We compared the NOAA Southwest Fisheries Science Center’s Environmental Research Division (formerly Pacific Fisheries Environmental Laboratory: PFEL) coastal upwelling indices along the northern Baja California coast with those derived from winds measured by coastal meteorological stations and estimated by the QuikSCAT satellite. With the exception of the PFEL series at 33 ̊N, the three data sets compare reasonably well, having similar typical year patterns, correlations >0.6, and significant coherences for periods three to five days or longer. By contrast, the seasonal variations, the timing and magnitude of maximum upwelling, and the variability of the PFEL indices at 33 ̊N are significantly different compared to all the other time series, including QuikSCAT at that location. The performance of the QuikSCAT winds close to shore was evaluated using the coastal meteorological station data. Although large root-meansquare (RMS) errors in direction were found for the QuikSCAT winds, both datasets have properties similar to the variance ellipses, and show reasonable coherences for frequencies in the weather band and lower, particularly south of 33 ̊N. INTRODUCTION Winds near the Pacific Coast off Baja California blow predominantly from the north-northwest, causing an offshore Ekman transport that results in year-round upwelling of cold, relatively saline and nutrient-rich waters in the coastal region (Lynn 1967; Bakun and Nelson 1977; Huyer 1983; Schwing et al. 1996; Strub and James 2000). Coastal upwelling helps explain the large productivity along the North American coast and upwelling intensity has been linked with variability in fish stocks and other factors affecting coastal ecosystems (e.g., Reid et al. 1958; Ryther 1969; Longhurst 1998). Researchers have also used upwelling variability to explain zooplankton population spatial processes in coastal systems (Peterson et al. 1979), and more recently a link between latitudinal variability in coastal upwelling and intertidal larval supply, population dynamics, and community structure has been hypothesized (Roughgarden et al. 1988; Connolly et al. 2001). Coastal Upwelling Indices (CUI) at 15 standard stations along the North American coast have been generated since 1945 by the NOAA Fisheries Southwest Fisheries Science Center’s Environmental Research Division (formerly the Pacific Fisheries Environmental Laboratory: PFEL), and are publicly available at its website (http://www.pfel.noaa.gov/). The indices are estimates of the offshore Ekman transport obtained from geostrophic winds, which in turn are derived from the surface pressure fields of the operational atmospheric model provided by the U.S. Navy Fleet Numerical Meteorology and Oceanography Center (FNMOC), Monterey, California. For historic reasons, and to be consistent with previous references, we will refer to the Environmental Research Division coastal upwelling indices as the “PFEL indices.” The PFEL indices have been widely accepted, with more than 50 regular users each month, several dozens of additional requests for the data each year, and more than 400 scientific publications referencing them (Schwing et al. 1996). The studies cover topics ranging from descriptions of coastal circulation patterns, climate change, and linkages between environmental and biological variability. They have been particularly popular in linking physical forcing with marine population variability (e.g., Ainley et al. 1993; Parrish and Mallicoate 1995; Rau et al. 2001; Koslow et al. 2002; Ladah and Zertuche 2004). The limitations of the PFEL indices have been discussed elsewhere (e.g., Schwing et al. 1996). The most important may be the fact that upwelling is the combined effect of two processes: the offshore Ekman transport due to the alongshore component of the winds (which is what the PFEL indices represent) and the Ekman pumping that results from the curl of the winds 204 204-214 Perez Brunius 11/17/07 1:32 PM Page 204