Prediciting 2016 - May update.

Now that we have all the main temperature sets for May 2016, it's time to update my forecasts for the year. See the previous post for details - I'm not making any claims about the actual probability of new records for 2016, I'm just performing a simple linear extrapolation from previous years.

May across the various data sets was sharply down on the high point of the El Niño, and some of the alarmists on places such as Watts Up With , this will inevitably lead to a new ice-age. This has had some impact on the projections for the final temperature of 2016. All data sets show a drop in the forecast of around 0.01 - 0.03 °C, compared with the forecast from April.

Here's the graph showing this for each data set I'm tracking. In case it isn't clear the y-axis is showing the forecast for 2016 compared with the previous record for that data set. In the case of ground based data sets, (GISS, NOAA and HadCRUT) this shows how much warmer 2016 is projected to be over 2015, in the case of the satellite data the comparison is with 1998. The x-axis shows the forecast that would have been made at each month of the year - so the left most edge is showing the forecasts made after the January temperatures are known, the rightmost side are the forecasts based on temperature from January to May.

However this has had practically no effect on the calculated probability of any data set breaking the record. In fact to the nearest percentage point, only RSS 3.3 has changed, from 94% to 93%.

I should stress, that I don't believe these probabilities, the real probability should be reduced to take into account unknowns, and the unusual start to the year. All the above is just for fun, and I take no responsibility for gambling losses.

Footnote

As I write this, Roy Spencer has released the June figures for UAH beta 6, and this shows another sharp drop. I'll look at this in more detail when more results are known, but this will have a bigger effect on the forecast. Including the result for June pushes the UAH beta 6 probability of 2016 being a record down to around 80%, with the expected value now only 0.05 °C above 1998.