[Petrochemicals (OP)]

This is a graph of historical sea surface temperatures. There are others avaliable but they all show the same anomoly around 1940.


1.jpg (84.61 kB . 491x619 - viewed 1891 times)

As you can see there has not been such a stand out reading since. What causes such a surface temperature spike ?

[evan_au]

I did hear an explanation of this in a podcast, but I would be hard-pressed to find it again (maybe Radiolab?).

As I heard it, Britain was collating and reporting most of the sea-surface measurements before and after WW2.
- Method: Take a bucket, throw it overboard, pull up some water, and measure its temperature.

During WW2, a lot of US ships joined in the reporting
- Method: Measure the temperature of the water inlet pipe used for engine cooling

There is potentially a small difference in temperature between the results of the two methods:
- A bucket of water might experience some evaporative cooling
- An engine inlet pipe might experience some conductive heating

Google found a reference for this discussion; apparently the measurement method is now recorded and used to correct the measured temperature (but if you misclassify the method, you might still get misleading answers)...
https://journals.ametsoc.org/view/journals/clim/33/18/jcliD190972.xml

[Petrochemicals (OP)]

As I heard it, Britain was collating and reporting most of the sea-surface measurements before and after WW2.
- Method: Take a bucket, throw it overboard, pull up some water, and measure its temperature.
 

I cannot believe the data has not been filtered for this, the graph plot is widespread and I cannot find any other. Simply rejecting methods from unknown sources would filter it out, the original method in isolation would give a better reading sort of like the CET record.