Steele, you're a laugh a minute. So because he once used an alias he's not to be believed? So on your own assessment you are not to be believed. All we have for you is an alias. We don't know your real name even, although we do know his. So in "Steele world" he has to be more credible than you. I, however, am not using an alias, so in Steele world have a heap more credibility than either of you.

Of course in logic world, it doesn't matter whether you use an alias or not. What matters is whether you are right.

And are you seriously asserting that they actually have reliable temperature measurements in the interior of Africa? They don't. Same thing happens in Antarctica. They only have three recording stations from memory there, and they just infer the rest of continental temperature from that.

You can't make a bad argument good, by making a good reputation bad. All you do is ruin whatever crediblility you still retain.

re the great Central Africa debate. One graph showed that they had no data for the grid area in question. The second graph showed that, despite having no data they nonetheless determined that the region had record warm temperatures.

How can that happen?

When these agencies don't have data for a particular location for a particular period they have two main ways to fill the gaps:

1) they use other data from the location. So let's say they have no data for September but data for August. They'll look at other years where they have data for both, determine the average difference between each and apply that difference to the missing month. Clearly each year that average will change and so the data in the missing month will change.
2) they compare the data from adjacent locations which do have data for the missing month. So as in our example above, they'll find near-by location(s) and using the ratios between the two, determine the missing data.

Generally they use both methods and give a weighting to each to average the calculated data.

But giving a weighting means that the data is open to inadvertent or deliberate fudging. If the weighting results in a cooling, then it is assumed to be wrong and the weighting changed.

The other interesting aspect to this is that newly calculated data for location A which is based on location B data can then be used to find missing data for location C. And so we have a series of estimated data rather than actual readings. Even worse, you can end up with location D having actual data but having it changed because it doesn't agree with A, B or C.

Its this type of data manipulation which has caused things like the absurdity over 1934. Once 1998 and 1934 had the same temperature. But today, 1934 is determined to be over half a degree cooler than 1998. And each year the difference between the two changes as new data causes old data to be recalculated.

And this is called science!

GrahamY,
Seriously? You think antarctic temperatures are extrapolated from just three weather stations?
That may have ben the case fifty years ago when we didn't understand the significance of antarctic weather. But those days are long gone.
To discover the current situation, see http://amrc.ssec.wisc.edu/aws/index.html

Similarly Africa isn't as backward as you think. They do have a related problem though: red tape sometimes makes the data difficult to access.

It is worrying that you're so ignorant of the science that you base your opinions on the assumption that it's not even being done!

Aidan,

There's a world of difference between having a weather station and having a site that is acceptable to the GHCN. There are a range of criteria that must be met before, for example, GISTEMP will accept that station as having good enough data to be used. Issues such as siting, maintenance etc make many stations unacceptable.

The last time I looked at this there were 11 antarctic stations that were considered of high-quality and were used by GHCN. Of those 7 were in the west antarctic and so close together that they were in the same grid and therefore treated, essentially, as one site.

The vast majority of the antarctic is not covered by the GHCN and therefore is inferred from related sites.

As to Africa, the graph shown in the video originally referenced, showed vast areas of Africa with no data. Why? Well there are myriad reasons for a site to be rejected in any particular month. You only need to make an error in or miss one day, to have the site marked as no data.

There is this assumption that the global temperature-station data network is pristine, closely monitored and professional maintained. The truth is very different. That's why I prefer the satellite data.

I don't know why mhaze, recent checking of the BOMs stations has shown that more than 50% of don't comply to their own published standard.

Unsurprisingly, they are happy to use all these non complying stations in all their official records & record braking reports.

Dear GrahamY,

You asked; “So because he once used an alias he's not to be believed?”

Nope. It is just combined with no “climate qualifications and a history of having to retract bogus climate claims” he would “not ever be my go to source”.

And to be clear I went and deconstructed two sections of his video which showed instances of him being less than honest before raising the issue of credibility.

Here is the time stamped link to the part of the video where he plays silly buggers with the graphs.

http://youtu.be/mGe9JO58Uc8?t=167

So to your baseless claim that “And are you seriously asserting that they actually have reliable temperature measurements in the interior of Africa? They don't.”

Here is a link to the Global Historical Climatology Network-Monthly (GHCN-M) Mean Temperature (Version 4) documentation. In it you will find a map on page 8 titled “Figure 1: Location of the ~26,000 stations in the GHCNm v4 inventory.”
ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v4/documentation/CDRP-ATBD-0859%20Rev%201%20GHCN-M%20Mean%20Temperature-v4.pdf

You will see from the map that there most certainly are stations in countries like Chad which are in the area you and your mate are referring to.

Further you will see they are colour coded as to the length of years of available data. Those within the area we are discussing are within the last 50 years and while obviously not able to give a robust enough data set to give a “Temperature Departure from Average Sep 2016 (with respect to a 1981-2010 base period)” they most certainly would allow a determination of “Temperature Percentiles”.

This is all stuff someone who was even half way serious about getting to the bottom of an issue could have easily found. Tony Heller aka Steve Goddard either didn't bother or didn't want it to spoil a good story.

In my opinion he is a lightweight with an agenda and there are far more honest, more respected, and better qualified contrarians you could be putting up.