Sensitivity of ocean heating rate estimate to time series processing methods

Ocean heat content (OHC) is a key indicator of global warming, and OHC rate can be used to quantify Earth's energy imbalance (EEI) as ~90% of EEI is stored in the ocean. However, when estimating global ocean heating rate that is the first derivative of OHC (dOHC/dt), various time series processing methods were used, including smoothing, differentiation, and annual mean definition, which introduced differences in the resultant dOHC/dt estimate. This study utilized nine OHC datasets to analyze the impact of the above-mentioned approaches on dOHC/dt estimation and compared with the EEI observation at the top of atmosphere. The results show that the choice between the January-December and July-June annual means introduces 14% relative trend uncertainty into the estimation of the long-term trend of dOHC/dt. Different time derivative methods affect the interannual signal phase and amplitude, where the center difference for dOHC/dt and Clouds and the Earth’s Radiant Energy System (CERES) strongest EEI consistency. The smoothing process further enhanced the consistency between dOHC/dt and CERES. These results indicate that the choice of time series processing techniques introduces a non-negligible uncertainty in the OHC-based EEI estimate. Moreover, this study shows that different OHC datasets show larger differences compared with the methodology choices, suggesting the data uncertainty is still the primary source of OHC-based EEI estimation errors. This study provides a methodological basis for improving EEI estimation and also for the evaluation of climate models with observations.