PFEL is using state-space modeling techniques to look retrospectively at historical oceanographic and atmospheric data sets from the Steller environment.| An advantage of the state-space approach is its ability to separate seasonal and long-term contributions to the variability observed in environmental time series. Both contributions are ecologically important, as many marine organisms respond strongly to environmental changes on both seasonal and decadal time scales. The following three data sources were chosen for analysis due to their long record, easy access, and relevance to the Steller habitat: (1). Sea level pressure (SLP) fields for the dominant centers of action in the northeast Pacific (the nominal positions of the Aleutian Low and North Pacific High), from the National Center for Environmental Prediction (NCEP) reanalysis, for the period 1948-2000; (2) Sea surface temperature (SST) and surface wind data for 1° boxes in the eastern, central and western Gulf of Alaska, from the COADS data set, for the period 1950-1997; and (3) Mixed layer depth and upper ocean heat content throughout the coastal northeast Pacific from several dynamical models, including those that are part of this project. In order to isolate the influence of ENSO on the atmospheric changes in the northeast Pacific, CDC is executing a large ensemble of experiments with the GFDL Atmospheric General Circulation Model (AGCM) coupled to a gridded 1-D mixed layer ocean model, with observed SSTs specified in the tropical Pacific. CDC is also working with NCAR~Rs ocean GCM (OGCM) driven by observed surface fluxes over the period 1958-1997. SIO is using the Regional Ocean Modeling System (ROMS) on an ETOPO-V based, 10-km grid covering the GOA north of 50°N and east of 170°W.| The model is forced by NCEP-derived climatologies from the 6 years before and after the 1976-1977 regime shift.| The biological model that is coupled to ROMS is a nitrogen currency, eight-component ecosystem model consisting of parallel paths representing old and new production. Each path consists of a single phytoplankton compartment, associated zooplankton grazer, and feed into a common detrital pool.
RESULTSThe following is extracted from a progress report submitted to CIFAR, 15 July 2002, and available here [WORD format] PFELs decomposition of the SLP series documented significant changes in the phase and amplitude of the seasonal cycle in the Aleutian Low and North Pacific High. The Aleutian Low seasonal amplitude nearly doubled over the study period, mostly due to wintertime deepening. Changes in the amplitude and phase of both the annual and semi-annual components contributed to a strongly time-varying structure of the North Pacific High. In particular, spring arrived earlier in the California Current System through the 1990s compared to the 1950s. These changes in seasonality can impact lower trophic levels by affecting local Ekman processes and disrupting the timing of seasonal events. Changes at lower trophic levels, in turn, may impact the diet of higher trophic level organisms, including marine mammals in the northern GOA. This study demonstrated the importance of considering a non-stationary and non-deterministic seasonal cycle when looking for the effects of environmental variability on marine ecosystems. Results from this analysis have been published in Geophysical Research Letters (Bograd et al., 2002), and a reprint is included in this report. Analyses of the other two primary data sources, focused on looking at long-term trends and climate-driven change points, are continuing. Ocean conditions in the GOA are sensitive to changes in the tropical Pacific on interannual and decadal time scales.| For example, the 1997-1999 El Niño-La Niña cycle had a substantial impact on the northeastern Pacific. The CDC group, using simulated SSTs, achieved 1°C warmer SSTs and a 40m shallowing of the January-February mixed layer in 1998 versus 1999.| On longer time scales, wintertime SSTs were also warmer in the central GOA in the decade after the 1976 regime shift compared to the decade before. This result was also mirrored in the OGCM model runs (using surface fluxes over the period 1958-1997), which indicated a large scale warming of 0.5-0.75°C in the GOA beginning in the mid-1970s.| The warming was mainly a surface phenomenon, however, especially in coastal regions where temperatures at 100m depth were actually colder after 1976. Based on analysis of the NCAR OGCM, CDC has found reduced Ekman pumping the northern part of the GOA in 1977-97 as compared with 1958-75. This resulted in decreased upwelling and a deepening of isopycnal surfaces between 40m and 200m along the northern rim of the gyre. The decrease in upwelling occurred between 125°W and 155°W and within 500 km of the coast. The CDC group is now investigating the processes responsible for these ocean changes. The SIO group is currently validating the physical accuracy of the GOA ROMS output.| The model adequately reproduces the general circulation patterns of the GOA, in particular the Alaskan Stream and Alaska Coastal Current.| Runs have been made with both pre- and post-regime shift wind sets. Once validation of the physical model is complete, SIO will repeat the current runs and make longer, decadal-scale runs with active biology. Analyses will concentrate on differences in biological response between the eastern and western GOA. Our initial results suggest at least two ways by which environmental changes can affect the GOA ecosystem: (1) by changing the intensity of Ekman processes, as demonstrated by the CDC results, and (2) by changing the seasonal timing of Ekman processes, as shown by the atmospheric changes revealed in the PFEL retrospective.| These processes are likely responses at different time scales of the same atmospheric phenomenon.| In the former case, changes in upwelling could lead to substantially different nutrient concentrations in the surface waters. In the latter case, shifting seasonality could alter not only the health of primary producers dependent on nutrient availability but the interaction between adjacent trophic levels. The hypotheses generated by these analyses are being fully tested in the SIO coupled physical-biological model runs.
PUBLICATIONS
TEAM MEMBERS
[ BARX |
Steller Sea Lion]
|
