Actually jopo, if you have another look at this you'll recognise that you have a large misunderstanding there and an incorrect conclusion. And jantimot's claim that they were incompetent in the past is also a long way off the truth.
- They now have enough overlapping buoy and ship data to statistically reliably determine the difference in readings they are getting from the two methods at the same time/location. The ship data is relatively sparse so they have needed time to get enough overlapping data.
- As a result of recognising the size of the difference they can see that buoy data is consistently reading cooler than ship data at the same time/location.
- This is just unavoidable tricky instrument differences and differences in the methods used to take readings using the two approaches.
- It would not matter if we had a long historical set of data from each source; we'd just compare old buoy data with more recent buoy data and old ship data with recent ship data to determine the warming. But we have relatively sparse (old and recent) ship data and more recent and much more geographically widespread buoy data.
- Because all the historical data is ship data and because buoys read cool relative to ship data, the historical data is therefore consistently warmer than what buoys would have given us if they had been deployed historically.
- We are now comparing a mixed dataset that is proportionally becoming dominated by buoy data. The older total dataset is only ship measurements and the recent dataset is some ship measurements and mostly buoy measurements. In between times the dataset is a varying mixture of the two sources, increasingly dominated by buoy data as the ARGO network was deployed.
- In order to use all the available data, and because of the temperature reading differences between the ship and buoy data, and the changing mix of the two, it is therefore necessary to adjust the readings from the two sources so they are directly comparable.
The four key points are:
- There is no "calibration" of buoy data to ship data as you claim. They are not calibrating for absolute temperature. They are determining the anomaly between past and current readings, allowing for the different instruments used in each period and the affect of that on results.
- They have only recently gathered enough overlapping ship and buoy data to complete a detailed study and statistically reliably recognise the differences in measurements from the two sources, for the same time and location.
- So they have only just been able to conduct the re-analysis project to correctly ensure the comparison allows for that.
And they have found what they have reported.
- And on top of that we have had the last couple of years of record temps, which are included in that reanalysis, so, along with the sea temp record corrections, that just further emphasises that the "hiatus" actually never was.
In addition you don't want to use only the relatively sparse ship data. By merging the two datasets correctly in this way we get the benefit of both the recent much more geographically widespread ARGO buoy over all oceans and the longer but much more sparse and northern hemisphere centric history of the ship measurements.
No past incompetence. No rigging data. Just sufficient overlapping data from two sources to recognise an issue and correct for it.
I have my doubts about your claim that the satellite data differs from this. You've not stated what satellite data you are talking about, so it is impossible to say, but I suspect you are comparing apples and oranges.
