497 / 2024-09-18 08:28:57
Exploring Global Upper Ocean Marine Heatwaves in Coupled GFDL and NCAR models
climate modelling,climate change,global subsurface marine heatwaves
Session 4 - Extreme Weather and Climate Events: Observations and Modeling
Abstract Accepted
Periods of anomalously high ocean temperatures, known as ‘Marine Heatwave’ (MHW), have severely affected marine ecosystems’ health, function, and services they provide, causing substantial biological and socioeconomic disruptions over the past few decades. While there have been many efforts to understand sea surface MHWs, our understanding of their vertical structures is relatively limited. However, subsurface MHW can have dramatic ecological impacts, and may also influence other oceanic properties (e.g., dissolved O2, pH), causing compound ocean extremes. In this study, we evaluated MHW climatology through the upper 700m during the preindustrial and recent historical times using daily temperature outputs from global simulations performed with the coupled GFDL CM4 and NCAR CESM3 climate models. By comparing the simulated MHW between the two models, we analyzed the role of different representations of ocean physics and vertical coordinate types in model performance. We also assessed the global patterns of the sensitivity of MHW simulations to the temporal resolution of temperatures by comparing them against the detected MHWs using monthly data.
Model preindustrial control simulations demonstrate errors in ocean potential temperature due to temperature anomalies relative to the first year of simulation (i.e., model drift). Given the non-negligible drift shown in previous studies, we calculated the trend in daily temperature of the piControl model runs, and corrected temperatures by removing the trend over preindustrial and historical periods. We then detected MHW characteristics based on the de-drifted temperatures and systematically analyzed the influence of model drift on simulated MHWs to identify locations where the MHW metrics are most sensitive to drift-induced trends. For example, preindustrial model drifts of GFDL CM4 lead to overestimated annual peak heat intensity up to 0.1-0.4 °C over the tropical Pacific and the mid to high latitudes of the upper 700m. This suggests the trend associated with drift can induce a shifting baseline that should also be accounted for when analyzing MHWs from climate model output. Our study contributes to understanding the evolution of MHW through the water column and reveals critical factors for better MHW simulation that could benefit future projections.
Model preindustrial control simulations demonstrate errors in ocean potential temperature due to temperature anomalies relative to the first year of simulation (i.e., model drift). Given the non-negligible drift shown in previous studies, we calculated the trend in daily temperature of the piControl model runs, and corrected temperatures by removing the trend over preindustrial and historical periods. We then detected MHW characteristics based on the de-drifted temperatures and systematically analyzed the influence of model drift on simulated MHWs to identify locations where the MHW metrics are most sensitive to drift-induced trends. For example, preindustrial model drifts of GFDL CM4 lead to overestimated annual peak heat intensity up to 0.1-0.4 °C over the tropical Pacific and the mid to high latitudes of the upper 700m. This suggests the trend associated with drift can induce a shifting baseline that should also be accounted for when analyzing MHWs from climate model output. Our study contributes to understanding the evolution of MHW through the water column and reveals critical factors for better MHW simulation that could benefit future projections.