From the Watts Up With That? website
The blue line shows the number of stations that are taking recordings. The red line shows where there is data missing. Usually this data is between 1 and 9 days. So in theory ALL of them could be one day missing something. Making assumptions that the temperature in Singapore was 35C degrees every day of the month except the one day they forgot to measure because it was at -50C degrees simply isn't going to happen.
All in all, if you look at the last 15 years you're looking at less than 1/4 of the data is missing possibly 1 day, and not more than 9 days.
Now this chart, they've gone from less than 25% where they have to fill in data because it's not been recorded, for whatever reason, such as electricity failure in third world countries etc. And they've got the purple line where the records are "estimated", as far as I can make out because a little bit of the information is missing.
This would be potentially 1/30th of 1/4 of the information. So, 95% of the data isn't 100% complete.
And again, how wild are the temperatures we're looking at. If you have temperatures for the month that range between a few degrees, and then they have to estimate a temperature one day, and it's the same as the 10 days before, chances are the temperature is going to be pretty close.
So say 95% of the data is wrong is misleading. 95% of the data, based on 1/120th of the information being estimated, which might be out by a small amount, shows how you can make statistics say what you like.
Is 95% of the data wrong or is less than 1% of the data slightly out? There's a big difference between the two.
Their conclusions is:
"The US accounts for 6.62% of the land area on Earth, but accounts for 39% of the data in the GHCN network. Overall, from 1880 to the present, approximately 99% of the temperature data in the USHCN homogenized output has been estimated (differs from the original raw data). Approximately 92% of the temperature data in the USHCN TOB output has been estimated. The GHCN adjustment models estimate approximately 92% of the US temperatures, but those estimates do not match either the USHCN TOB or homogenized estimates."
So, the US makes up 39% of the data means what? Does it mean that 39% of the data is being used equally with the other data, or does it mean that they adjust the data to make it fit? Well they sourced their own previous article and it doesn't say. So.... they're making a claim and not backing it up.
The point here should be that many third world nations are third world nations and can't always collect data as well as a first world country. Their data might not be 100% perfect, that doesn't mean that all the data around the world is wrong. It also doesn't mean that 95% of data, or 99% of data is irrelevant.
Approximately 66% of global surface temperature data consists of estimated values
The previous article suggests that 66% of data contains some estimates. So they can't even decide if it's 66% of 99%.
Even the 66% of data with estimates doesn't mean it's wildly wrong. It means that some of the data is estimated. It's not hard to estimate the data in many cases. If you look at the weather for 4 days and then not the fifth and the temperature feels about the same, then look on the 6th day and it's about the same and feels the same, you could pretty much guess what the temperature on the 5th day was.
So while they make valid points that the data is 100% reliable, there is nothing to suggest the data is too unreliable in order to change global temperatures massively. You're talking about 1% of data being estimated and it's probably not being estimated outside of a 1C degree of inaccuracy.
