ATTP "... how our climate will respond to increasing anthropogenic forcings is not going to depend on whether or not the behaviour of some scientists was not what you would have expected"

Quite true. But (surely?) an equally important point is how our society responds now to the threat of increasing anthropogenic forcings? In that debate, how the participants behave IS likely to affect the outcome. So those who wish to influence others should treat them at all times with the utmost courtesy and respect.

osseo,
My personal view is that it would be nice if everyone treated everyone else with courtesy and respect. I don't always achieve this myself.

In that debate, how the participants behave IS likely to affect the outcome.

Yes, this is probably true.

So those who wish to influence others should treat them at all times with the utmost courtesy and respect.

If someone was indeed trying to influence others (car salesman, for example) then this would seem sensible. Scientists, however, are simply trying to present information. What we choose to do with that information is up to us. We can, of course, ignore it if the scientists don't behave as we might like, but that would still be our responsibility, and our fault, if their information suggested we should do something that we decided not to do. Also, if those scientists are regularly accused of lacking in integrity by some who frequent a certain blog, they may find it hard to treat those people with the utmost courtesy and respect; they are only human ;-)

ATTP,
You censor and ban people at your pathetic site..
You are a pedantic blithering hypocrite with no apparent ability for self reflection.

Of course, aTTP talks about anthropogenic “forcings” as if there was no argument as to their existence; sorry, Mr Rice, but they are still theoretical.

There is very well established radiative physics that allows us to determine the radiative influence of changing GHG concentrations in the atmosphere. There are measurements from space of changes to the outgoing spectrum that are consistent with this radiative physics. There are measurements on the ground of increases in downwelling longwavelength flux that is also consistent with our understanding of radiative physics. If you think that they're only "theoretical" then you presumably regard a great deal of well established science/physics as just being theoretical.

You censor and ban people at your pathetic site..

I certainly moderate strongly and do ban some. What of it? It's my site.

You are a pedantic blithering hypocrite with no apparent ability for self reflection.

Okay, but you claimed I was a bully. What has this got to do with bullying and who have I bullied and how? This is a rhetorical question, obviously.

I was trying to get the topic onto information theory. I failed, but i have to say that aTTP did try to address my points, but clearly didn't understand the problems in getting a signal from entropic data. Maybe he missed that class in his undergraduate course, or more realistically maybe he's never heard of Claude Shannon. Anyway I'd like for someone to explain to me why you would expect to get a reliable signal out of a data set that is entropic (30 years average SAT) with another entropic dataset (One year average) when the owners of the datasets say you can't measure them.

I'll leave it there.

Anyway I'd like for someone to explain to me why you would expect to get a reliable signal out of a data set that is entropic (30 years average SAT) with another entropic dataset (One year average) when the owners of the datasets say you can't measure them.

Maybe because they don't have a 30 year average SAT and a one year average SAT. There is only one dataset per station. From that you can determine a 30 year average and then the anomalies.

"I certainly moderate strongly and do ban some. What of it? It's my site.
Apr 12, 2016 at 8:24 PM | ...and Then There's Physics"

Rather sums up the attitude of climate scientists, to climate science, and the peer review process aswell, as demonstrated by the faked-up 97% consensus.

Yes aTTP, your site is your site, and you are allowed to do with it what you like. Isn't it fortunate that this site is not run along the same lines, as yours, and climate science? Imagine how outraged you would be if you were shut-out, censored or banned.

Isn't it fortunate that this site is not run along the same lines, as yours, and climate science?

I don't see why.

Imagine how outraged you would be if you were shut-out, censored or banned.

I couldn't care less.

Maybe because they don't have a 30 year average SAT and a one year average SAT. There is only one dataset per station. From that you can determine a 30 year average and then the anomalies.

We might be getting somewhere, because I don't understand this response, there is clearly something I don't understand.

My understanding of how they calculate the Global Average Temperature Anomaly (GATA) was that in, say 2015, the global average SAT for that year was calculate from the data available in the SAT temperature record which comprises weather stations and presumably the Argo network. That's one dataset's average.

This average temperature is then compared with the thirty year average, say between 1971 and 2000, that's the average temperature gathered from the weather stations and ocean temperature measurements over that period. You will note another possible cause of entropy in that the sea surface temperatures were previously measured with buckets from ships, so we're not comparing like with like. So that's a second data set of the thirty year average global temperatures.

The 2016 average temperature is then compared to the thirty year average of the thirty year average temperatures and an anomaly calculate, from which we get the GISS, HadCRUT, graphs of temperature anomalies, and "highest ever" temperatures of 0.02C above the previous highest, say 1998, which was itself measured with different instrumentation.

Looking at the Q and A above, and using a modicum of common sense, I can't see how anything remotely reliable can come from this process.

Let me try a simple example. Let's say we have 5000 instruments capable of measuring the temperatures over land and sea. We make an algorithm that takes the min and max daily temperature and decide we'll add them together to divide by 2 and call that the "average" temperature at that location. We then adjust the temperatures to take into account the location of the instruments, altitude, UHI etc., and known defects.

We then use this same network for 30 years to decide the 30 year global average temperature.

The assumption is that the instruments remain precisely the same and in the same locations. At the end of the 30 year period we start to calculate the annual global average temperature using the same instruments and algorithms, and compare this temperature to the 30 year average to get a GATA.

The likelihood is there will be a reasonable signal in that data.

The real world is, of course, a lot messier than that, in fact it's a lot more than a lot messier than that, as we can see above, so why are the scientific community sure that the temperatures they're publishing have any signal worthy of note in them.

To make it clear, I'm dealing only with the issue of the seeming confidence the scientific community in data, which to a layman, looks decidedly dodgy, not with any conspiracies, or accusations of malpractice, I'm assuming good faith from all parties. Furthermore, I'm not challenging the fact that the Earth has warmed. It's solely to do with the confidence the scientific community, from which you'd expect a more trenchant view on accuracy than your average layman, or woman has in the temperature data they putting into policy decisions.

If I have the process wrong I'd appreciate being corrected.

Okay, I think it's not quite as you say. You don't use all the data to produce a single 30-year average, and then use that to determine the anomalies. The anomalies are calculated with respect to the 30 year average for each station and at each time. For example, you have monthly data for 5000 stations. You calculate a 30 year average (over some base period) for each station and for each month. You then compute the anomaly for each station for each month. In other words, the anomaly for Feb 2015 for a particular station is the difference between the Feb 2015 value at that station, and the 30 year average for that station for February. You then average these anomaly values to get the overall anomaly. It's, of course, a bit more complicated than that, but I think that is the general idea.

"Isn't it fortunate that this site is not run along the same lines, as yours, and climate science? 

I don't see why.

Imagine how outraged you would be if you were shut-out, censored or banned.

I couldn't care less.

Apr 13, 2016 at 7:27 AM | ...and Then There's Physics"

People have been driven out of academic environments for those sort of views, but you are immune because of your Green connections. What a supreme display of arrogance.

What a bizarre response.

People have been driven out of academic environments for those sort of views,

Huh? What's this got to do about academic environments. I'm talking about how I choose to run my blog and how others choose to run theirs. Also, if any of my colleagues conducted themselves in a manner comparable to how some conduct themselves here, I doubt they would do so without being encouraged to stop behaving that way. Academic freedom doesn't quite mean "say whatever you want, whenever you want, wherever you want".

but you are immune because of your Green connections.

Maybe you can tell me what Green connections I have. I'm not aware of any.

What a supreme display of arrogance.

Again, I'm talking about how I run my blog and how others can run their's. If you think it's argument to somehow decide to run something that is yours, you've got a strange view of arrogance. I might even argue that it's rather arrogant of you to think that you can tell me how I should be running my blog.

Lets be clear about something. I've had comments deleted here. I've had most of my comments deleted on Kevin Marshall's site (despite them being on posts in which I'm mentioned) and may be effectively banned. Tom Fuller has banned me and deleted all my comments on the post in which he did so, despite me being mentioned in that post. Paul Homewood recently failed to post my response to his comment and appears to suggest that I can't comment there until I satisfy some condition that he has set.

I'm not complaining; they can run their sites as they wish. However, maybe you should consider that your pearl clutching about how I run my site is based more on my willingness to admit how I do it, than because what I do is somehow wildly different to what is done on sites that probably promote views with which you agree.

My far from unique rule of thumb is that where you don't have data then you don't have data. "Oh, it's quite OK to interpolate data up to 1200 miles away. Trust me; I'm a climate scientist" doesn't quite hack it, I'm afraid.

Sparse coverage is a fact of life in some parts of the world; what do you do for the purpose of calculating an average for areas where there are no thermometers? If you simply miss out the area, you are effectively infilling with the average for the dataset, which may be inacurate say in the Arctic, where temperatures for the region are rising faster han the global average, or you can choose to infill with a value, say the regional or hemispheric value, which will give another answer, or you can extrapolate form the nearest known values. Each choice is defensible and each of the published surface estimates does this slightly differently, which may account for some of the (minor) differences between them.

Do you refer to the 1200 km distance that GISTEMP use when calculating trends? If so, it is described on the GISTEMP website and in a peer-reviewed paper detailed there.

Thought experiment: Take 2 stations on the same site. The temperature rises by 0.1C over a period. Clearly they should both register the same delta-T. Now move one away by a given distance, they will now probably observe a different but still correlated delta-T. Now move it again, and measure the correlation.

What Hansen et al found, by an analysis of the station data, is that the correlation drops to 50% at a distance of 1200 km, and so they chose this as a threshold when extrapolating across sparsely-covered regions. Note that it is temperature change, not absolute temperatures that were found to correlate across distance. [Clearly, that's a simplification, but it captures the gist, read the paper for the detail].

Hansen, J.E., and S. Lebedeff, 1987: Global trends of measured surface air temperature. J. Geophys. Res., 92, 13345-13372, doi:10.1029/JD092iD11p13345.

You're welcome.

RR,

If you don’t have data, you don’t have data; what is it about that simple statement that you do not understand?

It's not that people don't understand it, it's that it's silly. Extrapolating into areas without data is standard practice and almost always unavoidable. What - in your view - is the minimum area a thermometer should cover before it's acceptable to produce some kind of global temperature anomaly? Or, what is the minimum number of thermometers necessary to do so. There is an actual answer to this, I think.

Or, what is the minimum number of thermometers necessary to do so. There is an actual answer to this, I think

Sixty.

To repeat, if you're calculating a global average and you have areas of 'missing' coverage than you either give up and go home or calculate the average with the data you have, which effectively means you're assuming the missing region is at the average for the globe. That's one choice - there are others.

Thanks. I thought it was about 100. We'll see if RR agrees? ;-)

Sixty per hemisphere would be ample.

http://journals.ametsoc.org/doi/full/10.1175/1520-0442(1999)012%3C1280%3AEOSDOF%3E2.0.CO%3B2

Either way, the dataset is massively oversampled.