At a few locations in the North Pacific Ocean the estimated carbon offsets at the crossovers were not consistent with the offsets from the basinwide MLR analysis. In an attempt to determine whether the crossovers were biased by the limited number of stations analyzed, we expanded the crossover analysis to include additional stations along each cruise and/or stations from neighboring cruises. The deep (> 2200 m) station data were averaged at specific potential density values (σθ or σ-3), and the mean values were compared at several isopycnal intervals for each cruise. The range of values observed for a particular cruise at each isopycnal level indicated whether the initial stations used in the crossover analysis were offset from the surrounding stations. The additional stations also provided more robust estimates of the difference between cruises, since more data points were included in the analysis.
Analysis of P2 - P16N - P14 TCO2
Results of isopycnal analysis of P2 and P16N TCO2 near 30°N, 152°W indicates consistently higher TCO2 values (8-10 µmol/kg) in deep water on the P2 line as compared with P16N (Figure). The P14 data along 179°W also show slightly higher deep-water values (mean = 4 µmol/kg) than the P16N results. In the previous crossover and MLR analyses P14 showed no obvious offsets, whereas the P2 were generally found to be high and the P16N data were consistently low. The results of the isopycnal analysis can be reconciled with the findings of the other methods if the P16N data are increased by ~4 µmol/kg and the P2 data are decreased by 4-6 µmol/kg.
Analysis of P17N - P17C - CGC91 TCO2
In the region near 37°N, 135°W, the isopycnal analysis indicated that the P17N data are consistently higher (range: 6-16 µmol/kg) than the P17C and CGC91 data (Figure). The CGC91 data agree reasonably well with the P17C results at higher densities, but are somewhat higher at the lowest densities. The crossover and MLR analyses of P17N indicate that the P17N data need to be decreased by 3-10 µmol/kg. An average decrease of ~6 µmol/kg in P17N will account for most of the observed differences with P17C and CGC91, but not all of the observed 6-16 µmol/kg range. The MLR techniques suggest that CGC91 should be increased by 4-5 µmol/kg, but the crossover results are much smaller and none of the techniques indicate a large correction is necessary for P17C. All of these results cannot be reconciled, but a small increase of ~2 µmol/kg in both CGC91 and P17C would minimize all of the observed offsets from the various approaches.
Analysis of P2 Alkalinity
An isopycnal analysis was used to evaluate the P2 data relative to the meridional cruises that it crosses. The P2 data were plotted as a function of longitude for three isopycnal intervals representing data between 3000 and 4000 m (Figure). Values from P10, P13N, P14N, and P15N, with latitudes between 25 and 35ºN, were consistently higher than the P2 data, indicating that a significant positive correction is necessary for this cruise.
Analysis of P31 Alkalinity
TALK for section P31 was compared with a nearly parallel section, P21, that was run along 20ºS and the intersecting meridional cruises P15N, P15S, and P16C (Figure). The P31 data at either end of the section, where the meridional lines intersect, are only slightly higher than the data for the other lines. In the middle of the section, however, the P31 data are much higher than the P21 values, suggesting that a correction is in order.
Analysis of P17C Alkalinity
TALK for P17C and neighboring cruises were examined as a function of latitude for σ-3 values between 41.44 and 41.51 (Figure). The larger density range was necessary to encompass enough data for comparison. The P17C data appear to be slightly higher than those for P16C or P15N. The low P2 data relative to P15N are also seen in the Figure.
Analysis of P17N - P16N Alkalinity
The data for P17N-P16N TALK was reexamined at the intersection of P17N with P16N to include data from a larger region and a wider depth range than initially considered with the crossover analysis (Figure). The calculated TALK values for P16N also include a +4 µmol/kg adjustment in the TCO2 values that is proposed from this study. The resulting extended crossover has an average offset of 12 µmol/kg.