El Niño and La Niña Redefined
This is the week to publish lots of interesting events and articles apparently. I have a number of things I would love to post about, but only so much time. Here is one that relates directly to something I posted on earlier: warmest La Niña years. Just a few short weeks after NOAA operations wrote that 2012’s La Niña was the warmest on records, NOAA researchers announced they recalculated historical La Niñas because of warming global temperatures. NOAA confirmed something that occurred to me while I was writing that post: eventually, historical El Niños will be cooler than future La Niñas. How then will we compare events across time as the climate evolves? The answer is simple: redefine El Niño and La Niña. Instead of one climate period of record, compare historical ENSO events to their contemporary climate. In other words, “each five-year period in the historical record now has its own 30-year average centered on the first year in the period”: compare 1950-1955 to the 1936-1965 average climate; compare 1956-1960 to the 1941-1970 average. This is different from the previous practice in which NOAA compared 1950-1955 to 1981-2010 and compared 2013 to 1981-2010. The 1950-1955 period existed in a different enough climate that it cannot be equitably compared to the most recent climatological period.
Figure 1. “The average monthly temperatures in the central tropical Pacific have been increasing. This graph shows the new 30-year averages that NOAA is using to calculate the relative strength of historic El Niño and La Niña events.”
I want to point out something on this graph. Is long-term warming evident in this graph? Yes, there is. But note they plot the breakdown by month. The difference between 1936-1965 and 1981-2010 in October is >1°F. Meanwhile, the same difference in May is ~0.5°F.
Here is the effect of NOAA’s change:
Figure 2. 3-month temperature anomalies in the Nino-3.4 region. (Top) Characterization of ENSO using 1971-2000 data. (Bottom) Same as top, but using 1981-2010 data.
NOAA’s updated methodology resulted in the identification of two new La Niñas: 2005-06 and 2008-09. The reason is warmer temperatures in the most recent decade than the 1970s (it sounds obvious when you say it like that). That warming masked La Niñas with the old methodology. It also means that the 2012 La Niña is no longer the warmest La Niña, as I related from the National Climatic Data Center last month:
Figure 3. Anomalies of annual global temperature as measured by NOAA. Blue bars represent La Niña years, red bars represent El Niño years, and gray bars represent ENSO-neutral years.
That record will now go down as a tie between 2006 and 2009, with 2012 coming in a close third. This situation is analogous to the different methodologies that NOAA and NASA use to compute global temperatures and where they rank individual years. Records might differ because of methodological differences, but the larger picture remains intact: the globe warmed in the 20th and so far in the 21st centuries. That signal is apparent in many datasets. Within the week, I’m sure we’ll hear from GW skeptics that La Niña years have been getting cooler since 2006. Here is what is most important: 2000s La Niñas were warmer than 1990 Niñas, which were warmer than 1980 Niñas, etc.