Trends and persistence in ocean acidification as measured by station ALOHA

Abstract

Ocean acidification, largely driven by the uptake of anthropogenic carbon dioxide, is reflected in a sustained decline in seawater pH. This paper examines the dynamics of surface-ocean pH at Station ALOHA over the period 1985–2024 using fractional integration methods, which allow for a flexible characterisation of persistence and trend behaviour. The differencing parameter is estimated under alternative assumptions concerning the error term, namely white-noise and autocorrelated Bloomfield disturbances, and recursive estimation is used to assess the evolution of the relevant parameters over time. The results show a negative and statistically significant time trend in both the original and logged pH series. The estimates of the differencing parameter are positive and significantly different from zero in all cases, providing evidence of long-memory behaviour. Under white-noise errors, the estimate of the differencing parameter is 0.89 and the unit-root hypothesis cannot be rejected, whereas allowing for autocorrelation yields an estimate of approximately 0.55, implying mean reversion with long-lasting but transitory effects of shocks. Recursive estimates further indicate that both persistence and the magnitude of the negative pH trend have increased over time. These findings suggest that pH dynamics at Station ALOHA are characterised by a persistent decline and slow adjustment following disturbances, highlighting the importance of long-term ocean monitoring and modelling approaches that account for fractional dependence. Nevertheless, the results should be interpreted with caution because the analysis is based on a single long-term record combining observational and reconstructed data.