Dissolved oxygen concentrations from three depths (5 dbar, 1020 dbar, and 4750 dbar) at Station ALOHA were investigated to spot trends and suspicious data. The oxygen data used in this section span HOT-1 through HOT-148. Only data flagged as good were used.

     The near surface dissolved oxygen concentrations at Station ALOHA vary seasonally by approximately 10 µmol/kg, or ~5%, therefore a direct comparison of concentrations at this depth over time would not be very useful. Conversion of the concentrations into percent saturation allows for a better comparison since near surface waters should be near equilibrium with the atmosphere and therefore should be near 100 percent saturated with oxygen. Significant trends or offsets in the historical HOT oxygen data could potentially manifest as large differences from the 100 percent mark.

     Percent saturation for each bottle sample was calculated with the equations developed by Weis (1970) via Owens and Millard (1985), using the measured salinity of the same bottle and the CTD temperature and pressure from the time each respective bottle was triggered. The resultant data were plotted in Figure 3 and had a mean of 98.7%.


Figure 3. Percent saturation of dissolved oxygen in near-surface waters (5 dbar) at Station ALOHA, from HOT-1 through HOT-148. The horizontal dashed line is the overall mean. The vertical dashed lines indicate cruises 10 and 30.

     The majority of percent saturated values that are plotted in Figure 3 fall between 97% and 101%, suggesting that the majority of oxygen determinations are being performed correctly. Deviations from these values can be split into two groups: a few points between HOT-30 and HOT-50, and the section of data from HOT-142 through HOT-147.

     Two bottle samples from HOT-32 and one from HOT-49 have percent saturation values of 105% or more. There were four near surface bottle samples collected during HOT-32, and all of them were more than 100% saturated. The HOT-32 saturation values are significantly greater than the cruises before and after. One bottle sample from HOT-49 was 111% saturated, but the other two near surface oxygen samples from that cruise fall within the general cluster of data. Three more bottles collected between HOT-32 and HOT-50 group themselves together and stand out from the surrounding data. One of the two bottles from HOT-40 and two of the seventeen reported from HOT-41 were between 101% and 102% saturated. The saturation values before and after those cruises tend to range between 97%-99%. Supersaturation is possible when surface waters are warming and have not yet equilibrated to the new temperature, as well as when strong winds generate higher levels of wave breaking that can enhance gas transfer between the surface waters and the atmosphere. Undersaturation of the surface waters can occur when deeper water that is lower in dissolved oxygen upwells or becomes entrained into the surface layer.

     The section of data from HOT-142 through HOT-147 (right side of Fig. 3) displays a downward trend in percent saturation, falling to as low as 93% during HOT-145 and HOT-146. The near surface dissolved oxygen concentrations for HOT-122 through HOT-148 are plotted in Figure 4 along with the respective theoretical saturated values. A cyclic change in concentration due to seasonal ocean temperature fluctuations is apparent, with the observed and theoretical values following each other closely up through HOT-141. Beginning with HOT-142, the observed values decreased while the theoretical values increased, and this continues through HOT-147. The observed concentrations from HOT-148 correspond well with the theoretical data.


Figure 4. Theoretical and observed dissolved oxygen concentrations near the surface (5 dbar) at Station ALOHA from HOT-122 through HOT-148.

     The trend of lower than normal oxygen values was recognized soon after HOT-142 and several experiments were made to isolate and eliminate the cause. Although some small sources of error were found, there was nothing of sufficient magnitude to explain the observed differences. For HOT-148, several things had been changed at once, including the redox probe and the Dosimat, so it is difficult to determine what exactly remedied the situation (Fujieki, L., personal communication, 2003). The possible causes of the problems during these cruises are analyzed below (Sect. 4.2.4).

 

Back to Table of Contents

Previous Next