Primary Productivity Response to Climate Change in the North Pacific

The North Pacific is of special interest because of its societal importance in supporting fisheries and because of the consequent effect of the downstream response over North America to possible feedbacks involving the oceanic biology. Changes in the physical environment of the North Pacific driven by global climate change can be expected to have an important effect on the primary productivity and the entire local ecosystems.

North Pacific primary productivity changes in response to global warming have been examined by driving a diagnostic NPZ model by the environmental changes predicted by a global ocean-atmosphere-coupled general circulation model (O-A GCM), as described in detail by Pierce (2003). The physical environment parameters that affect the biological model are MLD, solar insolation, water temperature, and Ekman upwelling at the base of the mixed layer. Although feedbacks to the physical climate are not included in this run, it provides a remarkable view of how the ocean biology may change in the globally warmed world.

The physical parameters that drive the ecosystem model were obtained from a climate change projection run of the parallel climate model (PCM) (Washington et al., 2000). This uses the CCM3 global atmospheric general circulation model run at T42 resolution (approximately 2.8° in latitude and longitude), coupled to the parallel ocean program (POP) ocean model run at 2/3° resolution. The coupled model includes land surface, runoff, and sea ice components, and is forced by a so-called "business as usual'' (IS92a) scenario of future CO2 and sulfur aerosol emissions. The PCM is relatively insensitive to CO2 forcing (e.g., Allen et al., 2001); most other major coupled climate models show larger global temperature increases by the year 2100 than does PCM.

A result of this simulation is illustrated in Fig. 8, which shows the ratio of primary productivity in the decade of the 2090s over the decade of the 2000s. Growing season values (March through June) are used here (but see below). The main effect of the changes in physical environment is to decrease the primary productivity in the shaded regions of Fig. 8. Detailed analysis (Pierce, 2003) shows that these changes are forced primarily by increased stratification (a consequence of the warmer surface temperatures) leading to a decline in MLDs during the winter. The shallower MLDs keep the NPZ system closer to equilibrium, with a consequent reduction in the amplitude of the spring bloom. In other words, in the shaded regions of Fig. 8, the decade of the 2000s has deep wintertime-mixed layers that are associated with a sharp wintertime drop in phytoplankton (and consequently zooplankton) concentrations. This is partly due to mixing a given quantity of phytoplankton over a deeper layer and partly due to the lower

March-April-M ay-J une

March-April-M ay-J une

Figure 8: Ratio of primary productivity in the decade of the 2090s over the decade of the 2000s for the March through June growing season (Pierce, 2003).

Figure 8: Ratio of primary productivity in the decade of the 2090s over the decade of the 2000s for the March through June growing season (Pierce, 2003).

average illumination levels in a deeper mixed layer. The rapid restratification of the water column in spring holds the phytoplankton near the surface in a well-illuminated region, and the depleted zooplankton cannot graze sufficiently quickly to prevent a spring phytoplankton bloom.

In the 2090s, the shallower mixed layers keep the phytoplankton and zooplankton populations more evenly populated year round, with, as a result, less of a spring bloom. It follows that productivity during the rest of the year is somewhat higher in the shaded regions during the 2090s as compared to the 2000s, but this is generally not enough to overcome the loss of the spring bloom, resulting in a net yearly reduction of productivity over the majority of the shaded region in Fig. 8. In the nonshaded regions of Fig. 8, the general increase in temperature by the 2090s tends to increase primary productivity. The overall result, then, is a combination of a modest, near-uniform increase in productivity due to the warmer water combined with a sharp loss of springtime productivity in the regions where warmer surface waters cap the wintertime mixed layer in the future.

It should be kept in mind that this line of analysis does not include other effects that will likely be important in the North Pacific, such as changes in the relative number of species, some of which might be better adapted than others to the changing environmental conditions. Also, future changes in the biogeochemical environment (such as increased iron deposition from industrial activity in east Asia) could have a strong effect as well.

This type of coarse resolution climate model is unable to resolve the dynamics of ocean boundary current systems. Yet primary production in these boundary currents, e.g., in the eastern boundary upwelling system of the North Pacific, is an important contributor to the earth's carbon budget.

Bakun (1990) suggested, based on observational evidence, that the warming of ocean temperatures associated with greenhouse conditions would lead to an inhibition of nighttime cooling and enhancement of daytime heating near the coast. This leads to an intensification of the continental thermal lows adjacent to the upwelling systems. The increase in onshore-offshore atmospheric pressure gradient would then be translated into an intensification of the coastal upwelling winds. More recently this idea has gained observational and modeling support. Schwing and Mendelssohn (1997) report a strengthening of upwelling favorable winds along the North Pacific eastern boundary current. Snyder et al. (2003) find a significant increase in up-welling favorable winds in a high-resolution regional climate model simulation forced by greenhouse gases.

In contrast to these observations of increased upwelling favorable winds, coastal observations over the last 50 years in the California Current System reveal a transition towards conditions that are more typical of reduced up-welling, such as a freshening of the surface waters in the coastal upwelling boundary (Bograd and Lynn, 2003; Di Lorenzo et al., 2005) and a decline in macrozooplankton abundance (Roemmich and McGowan, 1995). A possible explanation for these seemingly contradictory lines of evidence involves the observed increase in upper ocean stratification associated with warmer temperatures. In this scenario, the stratification exerts a stronger control than the winds on the ability of upwelling to supply subsurface nutrient-rich water at the coast (McGowan et al., 2003). This hypothesis has been tested with an eddy-resolving model of the coastal ocean driving a prognostic NPZ ecosystem model (Fig. 9). Di Lorenzo et al. (2005) performed model experiments that included as forcing conditions both the observed strengthening of the upwell-ing winds and the warming trend over the last 50 years. The effect of the increased stratification is strong enough in these experiments to inhibit the otherwise upwelling favorable conditions. The model chlorophyll response indicated a reduction in primary production in response to the combined effects of upper-ocean heating and increased upwelling favorable winds (Fig. 9).

Chai et al. (2003) used a coarse-resolution ocean model hindcast to determine that decadal climate variability has the largest impact on oceanic biological variability in the central North Pacific, a region bounded by two oceanographic fronts at approximately 30-32° N (Subtropical Front) and 42-45° N (Subarctic Front) in the central Pacific. This was based upon their analysis of the response of the MLD and the Ekman pumping to Pacific decadal variability. To extend this analysis and depict the possible changes in the carbon budget, Fig. 10 shows the spatially averaged modeled TCO2 concentration for the central North Pacific (defined as follow: 35° N-45° N, 170° E-150° W).

The modeled monthly averaged TCO2 concentration in the central North Pacific shows several scales of temporal variability between 1950 and 1993

Figure 9: Time series of model surface Chl-a averaged over the eddy-resolving California Current model coastal boundary within 50 km from the coast. (a) Model experiment that include the strengthening of the upwelling winds but no warming conditions. (b) Same as (a) but the warming conditions are also included as forcing. See Di Lorenzo et al. (2005) for details of the physical model experiments.

Figure 9: Time series of model surface Chl-a averaged over the eddy-resolving California Current model coastal boundary within 50 km from the coast. (a) Model experiment that include the strengthening of the upwelling winds but no warming conditions. (b) Same as (a) but the warming conditions are also included as forcing. See Di Lorenzo et al. (2005) for details of the physical model experiments.

Figure 10: Time series of TCO2 (unit: mmolkg 1) in central north Pacific (35°-45° N, 170°-150° W). Top panel shows the time series of surface TCO2 concentration. The lower panel shows the vertical profile of modeled TCO2 concentration from the surface to 250 m. The contour interval is 25 mmol kg-1 (For colour version, see Colour Plate Section).

Figure 10: Time series of TCO2 (unit: mmolkg 1) in central north Pacific (35°-45° N, 170°-150° W). Top panel shows the time series of surface TCO2 concentration. The lower panel shows the vertical profile of modeled TCO2 concentration from the surface to 250 m. The contour interval is 25 mmol kg-1 (For colour version, see Colour Plate Section).

(Fig. 10). First, the TCO2 has a strong seasonal cycle in the central North Pacific, which is due to the seasonal cycle of upper-ocean physical conditions and biological uptake. The second most pronounced temporal variability between 1950 and 1993 is the influence of Pacific decadal variability. For example, the modeled TCO2 concentration is higher during the 1980s and early 1960s, and the values are lower during 1970s and late 1960s. One of the reasons for the increase in modeled TCO2 concentration during the 1980s is the change in ocean circulation and MLD in response to the wind pattern changes in the central North Pacific. Chai et al. (2003) found that the modeled winter MLD shows the largest increase between 30° N and 40° N in the central North Pacific (150° to 180° E), with a value 40-60% higher (deeper mixed layer) during 1979-1990 relative to 1964-1975 values. They also found that the winter and annual mean Ekman pumping velocity difference between 1979-1990 and 1964-1975 shows the largest increase located between 30° N and 45° N in the central and eastern North Pacific (180° to 150° W).

Beside the impact of decadal climate variability on the modeled TCO2 concentration, there is another temporal trend (increasing from 1950 to 1993) in the modeled TCO2 concentration in the central North Pacific. This increase is due to anthropogenic effects because the model is forced with the observed atmospheric pCO2 from 1950 to 1993, which increased significantly during this period. In order to separate the anthropogenic uptake and storage of CO2 from the natural cycle of climate impact, Chai et al. (private communication, 2006) conducted a second experiment in which the atmospheric pCO2 is fixed at 1950 values and other surface forcing are unchanged. By comparing the results from this twin experiments, Chai et al. (2005) estimated the anthropogenic uptake and storage of CO2 in the Pacific Ocean solely due to the changing of atmospheric pCO2, eliminating the effects of changing upper ocean circulation and mixing. The modeled anthropogenic CO2 has a linear trend since 1950 with a surface-increasing rate of 0.57 mmolkg-1y-1, which agrees with several independent estimates based on the observations (Sabine et al., 2004).

Was this article helpful?

0 0

Post a comment