By Andy May
The simple answer is that the part of the ocean, roughly 54% that has HadSST values, is getting cooler. But Nick Stokes doesn't think so. His idea is that the coverage of the polar regions is increasing fast enough that the additional cooler cells from year to year cause the "true" upward trend in sea temperature to decrease. I decided to study the data further to see if this made sense.
First, let's look at the diagram in question in Figure 1. It's from our previous post.
Figure 1. This post shows the global mean temperature for the part of the ocean covered by HadSST this year. Both HadSST and ERSST temperatures are displayed, but the ERSST grid values are limited to the HadSST covered area.
As we discussed in our previous posts, the populated HadSST grid cells (the cells with values) have the best data. The cells are 5 ° latitude and longitude. At the equator, these cells are over 300,000 km² and larger than the state of Colorado. If Nick's idea were correct, we would expect the cell population at both poles to increase. Figure 2 shows the percentage of the global network (including the 29% with land) that is covered with populated SST cells (sea surface temperature) per year. The number of missing cells does not vary much, the minimum is 44% and the maximum is 48%. There could be a downward trend from 2001 to 2008, but not after that. While Figure 1 will flatten out after 2008, you'll have to work hard to see an increase from 2008 to 2018.
Figure 2. The number of monthly null cells in the HadSST dataset as a percentage of total global monthly cells per year (72 x 36 x 12 = 31,104).
So mixed message from this plot. Let's look at the zeros by year and latitude in Figure 3.
Figure 3. The number of monthly zeros by year and latitude.
Figure 3 shows that the zeros in the polar regions are fairly constant over the period 2001 to 2018. I made 2018 a thick black line and 2001 a thick red line so you can see the beginning and the end of the series more clearly. The actual variability in the South Indian, Pacific and Atlantic Oceans is between 55 ° S and 30 ° S. These are mid latitudes, not polar latitudes. Neither 2018 nor 2001 are outliers.
The same pattern can be seen when we watch a movie with the zero cells changing from 2001 to 2018. Click the map below to view the movie.
Figure 4. Map of the number of monthly null cells in the HadSST dataset for 2001. To see where the null cells are in all years through 2018, click the map and a movie will play. As before, the white areas in the specified year have no zero months, the blue color is either one or two zero months and the other colors are more than two zero months. Red means the entire cell is zero.
The number of null cells in the polar regions does not seem to change significantly from 2001 to 2018. The changes occur in the southern mid-latitudes. The number of zero cells in percent of the globe decreased slightly from 2001 to 2008, but only from 48% to 44%, which is not enough to reverse a trend. After 2008 there is no trend in null cells. From 2008 to 2018 the temperature trend is flat and not decreasing. However, given the changing number of cells, it is difficult to say that this is due to the number of colonized cells in the polar regions.
The reader is free to choose, but in my view we still have no idea what the global ocean surface temperature is, or whether it is warming or cooling.