Seasonal variations in the aragonite saturation state in the upper open-ocean waters of the North Pacific Ocean

Abstract

Seasonal variability of the aragonite saturation state ((AR)) in the upper (50m and 100m depths) North Pacific Ocean (NPO) was investigated using multiple linear regression (MLR). The MLR algorithm derived from a high-quality carbon data set accurately predicted the (AR) of evaluation data sets (three time series stations and P02 section) with acceptable uncertainty (<0.1(AR)). The algorithm was combined with seasonal climatology data, and the estimated (AR) varied in the range of 0.4-0.6 in the midlatitude western NPO, with the largest variation found for the tropical eastern NPO. These marked variations were largely controlled by seasonal changes in vertical mixing and thermocline depth, both of which determine the degree of entrainment of CO2-rich corrosive waters from deeper depths. Our MLR-based subsurface (AR) climatology is complementary to surface climatology based on pCO(2) measurements.