Mechanism of skillful seasonal surface chlorophyll prediction over the southern Pacific using a global earth system model

2020 
This study investigates the physical mechanism involved in an Earth system model (ESM)-based global marine biogeochemical prediction system providing successful forecasts of surface chlorophyll concentrations over the southern Pacific. The significant correlation skill of the surface chlorophyll concentration over the south-central Pacific (SP region, 160°–110° W, 10°–5° S) appears up to a 15-month lead. In contrast to the previously known role of the vertical nutrient supply on the predictive chlorophyll concentration forecasts, the NO3 budget analysis indicates that this prediction skill over the SP region is mostly controlled by the meridional advection of nutrients. Further analysis indicates that the controlling mechanisms involved in chlorophyll variability over the SP region can be explained by atmospheric and oceanic dynamics during the ENSO events. During La Nina, equatorial NO3 anomalies are increased due to enhanced equatorial upwelling, and the climatological southward current then advects nutrient-rich waters from the equator to the SP region $$\left( {{\text{i.e.,\,positive}}\, - \overline{v}\frac{{\partial {\text{NO}}_{{3}}^{\prime } }}{\partial y}} \right)$$ . In addition, anomalous easterly surface winds blow over the SP region as a circulation response to atmospheric diabatic heating anomalies during La Nina, which leads to southward current anomalies over the surface-layer ocean. This advects high climatological NO3 over the tropics to the subtropical south Pacific, which increases the NO3 anomalies $$\left( {{\text{i.e.,\, positive}} - v^{\prime}\frac{{\partial \overline{{{\text{NO}}_{3} }} }}{\partial y}} \right)$$ . This positive NO3 advection over the SP region is realistically simulated only at lead times shorter than 9 month, and the multi-season persistency of the nutrients contributes to the surface chlorophyll bloom at lead times longer than 1 year.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    72
    References
    2
    Citations
    NaN
    KQI
    []