Meteorologists frequently use ocean reanalysis data to study the evolution of El Niño-Southern Oscillation (ENSO) over time. These datasets are powerful tools and can paint a reliable picture of equatorial Pacific sea surface temperatures throughout the last 40 years. However, particular ocean reanalysis datasets may reproduce the evolution differently. Because of these divergent data, a team of ENSO experts recently published a study in Advances in Atmospheric Sciences that aims to better understand the reanalysis differences regarding equatorial Pacific upper ocean heat content. This critical parameter is frequently used to diagnose the state of ENSO.
“Interestingly, our study finds that such a difference varies within the life cycle of El Niño events.” said Prof. Wen Zhou, the corresponding author of the study from City University of Hong Kong. Her team notes that the difference among data sets grows as the El Niño develops toward peak phase.
Then, shortly after the peak positive ENSO phase, El Niño quickly decays. As ENSO neutralizes, the dataset results begin to converge closer to an agreeable state. However, this process takes longer than El Niño to neutralize. The dataset difference decay is even slower after a strong El Niño compared to a weaker El Niño.
Although a typical El Niño event decays quickly after its peak phase, its subsurface signal lingers within the region. This leads to a slow signal strength decay during the neutralizing phase of El Niño events. As a result, the decay rate of the difference among the ocean reanalysis data sets also slows down. Because of this, researchers determined that equatorial Pacific Ocean signal strength is strongly correlated to the dissimilarity among the ocean reanalysis datasets.
The asymmetry in growth rate and decay rate of the difference among the analyzed datasets leads to lower data consistency during El Niño. This makes analyzing El Niño mechanics during its decay phase inherently more challenging than in its developmental phase.
Advances in Atmospheric Sciences
Article Publication Date