Any estimate of temperature trends that have their endpoint on the uptick of the El Nino curve will give a misleadingly high trend. A new paper falls into this trap, claiming the global temperature hiatus never existed.
The headlines say there is fresh doubt over the so-called global warming “hiatus” or “pause” – a period since around the turn of the century when many scientists believe that global surface temperatures have increased at a lesser rate than before. This is because a new study suggests that the temperature of the oceans were being underestimated in the past 20 years or so because the ocean buoys used to measure sea temperatures were recording slightly cooler temperatures that the older ship’s intake systems, not a particularly new or surprising finding. It is an interesting study in that it produces individual ocean temperature data for buoys, satellites as well as for different ways of combining ship-intake methods and buoys.
Clearly different instruments measuring ocean temperatures will have different properties and such effects need to be studied so that data sets from each instrument can be combined to provide ocean temperature measurements over longer periods than any one method has achieved. So far, so good. But then the paper makes a gross error for the calibration of instrumental effects is one thing, what those instruments show is another. So the temperature measured by the buoys is cooler than ships, but what measurements are they producing, how is the ocean temperature changing, especially during the hiatus period?
The fundamental error made in the paper is in calculating the temperature trends in recent years – the period of the so-called global temperature hiatus. One has to be very careful about estimating temperature trends as they depend strongly on start and end years and changes of a year or two in them can produce very different results. One also has to be aware of the structure in the temperature data of the past 20-years or so as it is dominated not by long-term warming but by natural inter-annual events that are much stronger. There is the very strong 1998 El Nino that elevates temperatures, the much cooler La Nina years of 1999 and 2000, the El Ninos of 2010 and 2015 as well as smaller El Ninos and La Nina effects.
The 2015-16 El Nino has been one of the strongest on record temporarily elevating global temperatures by a significant margin. 2016 was the warmest year of the instrumental period (post 1850) but in the later part of the year global surface temperatures started to decline markedly. This means that any estimate of temperature trends that have their endpoint on the uptick of the El Nino curve will give a misleadingly high trend. It is obvious that a better trend will be obtained after the natural El Nino has ended. Likewise care must be taken if the start point is near the La Nina years of 1999-2000.
The temperature trends of the oceans estimated by the paper fall into this trap. Technically the trends calculated are accurate for the start and end points used, but they are unwise start and end points which are, to use a frequently misunderstood term, cherry-picked.
Consider the graph of ocean temperature data produced by the paper’s lead author. The El Nino at the end of the data is obvious. Like every other El Nino it will come down again, probably altering the conclusions of the paper that the pause never existed. Click on image to enlarge.