home

sitemap

contact

principle investigator profiles

research highlights:

modeling ocean behavior

western tropical Pacific

modeling ocean behavior

Amy Clement RSMAS/MPO University of Miami 4600 Rickenbacker Causeway Miami, FL 33149 305-361-4846 email website

Modeling the Behavior of El Niño in the Past

by Amy Clement

A great deal has been learned about El Niño over the past few decades. There is now a relatively complete observing system in the tropical Pacific which has lead to significant progress in predicting El Niño several seasons in advance. While much is known about the seasonal and interannual timescale behavior of El Niño, there are some important questions about its long-term variability that have yet to be answered. Some unusual behavior in the 1990’s along with the fact that two largest events occurred toward the end of the 20 th century in 1982-83 and 1997-98 has sparked some debate as to whether El Niño is being influenced by increasing levels of greenhouse gases. Looking back over the entire century, however, it is apparent that no two decades of El Niño variability have been the same, making it difficult to characterize the ‘natural’ variability against which to test whether there is an anthropogenic influence.

Some aspects of these issues can be addressed by looking into El Niño’s past. A wide range of proxies are now available that can extend the record back in time to better characterize the inherent variability of El Niño. These data can also provide information about the behavior of El Niño when the external forcing and mean climatic state were different from today, covering periods from the last millennium to hundred thousand years before present. Ideally, we could effectively ‘validate’ our current models of El Niño by running them under specific climatic conditions for those times in the past and then comparing results with the available data.

An interesting example of such a validation is for the Holocene. The data available for this time period include archaeological middens, fossil corals, lake sediments, ocean sediments, and pollen records. While the data sources are quite varied, they all tell a consistent story: El Niño variability was weaker prior to 5,000 BP than it is today. Some of these studies also seem to suggest that El Niño variability peaked in strength around 3,000-1,000 BP and has been decreasing since.

Figure Caption: Number of warm El Niño events in 100 year windows. The black line is the data published in Moy et al. (2002) from the Ecuadorian lake. Warm El Niño events are defined as light colored strata in the sediment record, which Moy et al. argue reflect pluvial episodes during large (warm) El Niño events. The green lines show an ensemble of seven simulations with the Zebiak-Cane model with the solar forcing of the last 12,000 years imposed. The blue line is the ensemble mean. Warm El Niño events are defined in the model as years in which the DJF NINO3 anomaly exceeds 3 K. This event index corresponds to the middle of the rainy season in coastal South American during which large SST anomalies associated with El Niño events are capable of causing the ITCZ to move equatorward and bring large precipitation anomalies to the region.

As part of my research, I have performed experiments with a model of El Niño to test the possible causes of such long-term behavior. I used the changes in solar forcing due to variations in the Earth’s orbital parameters over the past 12,000 years to force the model. The following figure shows a comparison between the model results and a proxy-record of El Niño from an Ecuadorian lake published by Moy et al. (Nature, 2002). The agreement between the model and data is remarkable, showing that both the increase in number of warm events over the Holocene as well as the peak around 1,000 BP can be explained as the response to the orbital forcing over the last 12,000. This example shows how models and paleoclimate data can be combined to gain insight into how and why El Niño behavior has changed in the past. What we learn from such studies can also lead to a better understanding of the overall sensitivity of El Niño to external forcing; understanding which can in turn be applied to the current and future climate.

Figure Caption: Number of warm El Niño events in 100 year windows. The black line is the data published in Moy et al. (2002) from the Ecuadorian lake. Warm El Niño events are defined as light colored strata in the sediment record, which Moy et al. argue reflect pluvial episodes during large (warm) El Niño events. The green lines show an ensemble of seven simulations with the Zebiak-Cane model with the solar forcing of the last 12,000 years imposed. The blue line is the ensemble mean. Warm El Niño events are defined in the model as years in which the DJF NINO3 anomaly exceeds 3 K. This event index corresponds to the middle of the rainy season in coastal South American during which large SST anomalies associated with El Niño events are capable of causing the ITCZ to move equatorward and bring large precipitation anomalies to the region.