The Dog Ate Global Warming - Interpreting climate data can be hard enough. What if some key data...

National Review Online ^ | September 23, 2009 | Patrick J. Michaels

[justa-hairyape]

Good points. But what I never understood is why just use the temperature readings ? Without knowing the humidity or the density of the air you are measuring, the temperature information cannot give you the actual Heat Content of the atmosphere. For example, yesterday it got up to about 90 degrees F in my shop, but I was comfortable because the air was so dry (less dense). So 90 degree air in the humid tropics is typically holding much more energy then 90 degree air in the southwest US. All temperature readings should also have density readings associated with them to make them useful to determining any recent atmospheric energy trends. Now you could use historical humidity readings, but that wont be accurate in areas where the density dramatically changes.

[r-q-tek86]

ping

[AFPhys]

“So in the mid-80’s, all the data from what time period was destroyed?”

-—<>-—<>-—<>-—<>-—<>-—

I have been following this specific issue for quite some time on the heavy, professional, science boards. Many researchers have gotten into the swing of requesting (and being denied) data from East Anglia (CRU).

CRU DID accept and accumulate the data from nations and stations world wide. They did NOT ever gather any information as to the “quality” of that data to begin with. Many countries and stations simply funneled the data collected over to CRU, and have completely depended on CRU to maintain the master database. The data in some cases has been collected for many centuries- since the invention of the thermometer. Other stations collected for only a few decades. Many many stations have been disbanded in the last 15 years, the demise of the USSR... as well as for economic reasons. In some cases that collection station was “moved” instead, and became the “official” record for that area.

What CRU has apparently been doing is the following, though it is speculation as they really are saying very little to anyone. Every time data gets dumped to them, CRU integrates it with the historical data from that area.

For example, a station is “moved”. The two stations running in parallel show some type of clear difference in behavior. CRU has been adjusting the historical data for the whole area according to their new “belief” as to what the new station is telling about the inaccuracies of the previous record. That adjustment may affect not only that specific pair of stations, but other stations as well, in the same area.

They have apparently been completely tossing the historical data, and retaining ONLY the “value-added” data, which has been adjusted according to their secret sauce. They have probably not even kept accurate records of the reasons for adjustments. For many areas of the world, the ONLY records that have been in existence since the 1980’s is the CRU database. All the original paper records and other records were long ago turned over to them for safekeeping.

Many times during the last decade (the period I’ve been following this closely) it has been noted that the “historical” temperatures given by CRU for a given area has been adjusted. I would guess that on average these data releases occur about annually. The odd thing is that the whole track was not simply adjusted up or down, but the shape would change, too. Many skeptics (that means “good scientists” noted this and expressed the why question). In recent years, this question has come more and more and finally has broken through to this point.

My example is only the “value adjustment” for a changed station. Presumably, CRU did adjustment for many other reasons, also. Again, the reasons are by now probably lost to antiquity since they seem now to reside only in the minds of the CRU trustees and employees. All that remains is the “value-adjusted” history.

No good scientist is now confident about anything in the CRU data. We don’t know what stations were used any more, or their quality. We don’t know whether the temperatures were properly adjusted for “urban heat island” effect, or for any other effect for that matter. (Some of the initial queries to CRU were intended to gain a handle on how well their UHI adjustments were done, I believe.) It seems that we may even have lost such simple data as how many stations temperatures were being collected for in the 1980’s versus the number now. We know it is far less now, but no one claims to know the number.

CRU has conglomerated all their raw data into their “value added” temperatures, and we really don’t know anything about the usefulness of the data now. We can collect in various publications over the years certain specific information, for example, the CRU temperature record for the Gold Coast 1880-1985 as seen in a 1989 research paper, and compare it to the G.C. CRU temperature record for 1880-2002 in a 2005 paper. It is clear in such comparisons that the the data from 1880-1985 is wildly different in some cases...

It appears to skeptics like me that the net effect of whatever CRU has done to the data has tended to:
1. Lower older temperatures.
2. Smooth out peaks and valleys.
3. Raise recent temperatures.

I’ve not seen many scientists claiming that there was anything malicious in way CRU’s adjustments were made. They simply are asking what was done, why, and how. Of course it can’t help but become a alarmist vs. skeptics battle for those who have high investment in the CRU record. As the questioning, and now I would even consider that it has become “investigation”, has accelerated, the CRU scientists are increasingly defensive. It now seems clear that they don’t have the raw data anywhere, and indeed that they are now unlikely to have any idea how whatever automated processes they used to “add value” did the adjustments.

A LOT of data was collected by and entrusted to CRU. Some seems to be completely gone now. Scientifically, I have to say that this really is a data disaster. It is even tempting to say something about the Library at Alexandria.

[donna]

The liberals are doing the same for cost of living, unemployment, poverty, marriage/divorce, abortion, ACT - any record they need to rig for their propaganda.

[AFPhys]

I’m not accusing CRU of being liberal, having an agenda, or anything at all, donna. I don’t know, and it really doesn’t matter to the science loss. They may simply have not been thinking about future possibilities when they discarded the raw data, among them that the data would be adjusted over and over. By now, though, it is clear that is a disaster.

I won’t even accuse them of being inept scientists. Unquestionably, there was a heckuva lot of data, and keeping every iteration of data and every record of the way adjustments were made is certainly daunting. They may well have believed that they didn’t have the ability to keep all those records, and believed no one would ever look at them if they were kept.

The worst thing I am willing to say about them at this time is that they should have been transparent all along, and let people know what they were doing. If others disagreed with their judgment, they could propose alternatives. Now, though, it is clear that their judgment proved not to be adequate, and much that might have been saved had this been known a decade ago is probably irretrievable.

It is never a good idea to have all your eggs in one basket... a complaint that I have had many times over the years when I’ve seen “cost cutting consolidation” in many arenas.

[donna]

You have a more forgiving outlook than I. I’m willing to hammer them as guilty-as-sin. I cannot believe that a scientist would ever let primary data be lost.

I worked at a bank during the years of conversion to personal computers and attended many meetings to discuss retention schedules for original source documents. It was serious business.

After CRU mishandled the source data, they randomly manipulation the data. It is the last nail in the coffin. They should be ridiculed as the religious cultists that they are!

[SunkenCiv]

Thanks neverdem.

· join · view topics · view or post blog · bookmark · post new topic · subscribe ·

[75thOVI]

As an old Boomer Quartermaster I can assure you that weather observations by a sub on patrol was limited to the observation on the rare occasions we were at periscope depth as to whether or not it was raining. At QM A school in Orlando we received training on the weather observations that surface ships make at regular intervals. The manual said that “All ships, (except submarines)”..... and I stopped paying attention after that. When we went outside for some cloud identification training, the Chief pointed to me and said, “Seaman Nigh, what kind of cloud is that?” I looked up and said, “Big, white, fluffy, might rain iffin it don’t clear up.” He looked at me, looked at my Dolphin insignia, (he had a pair just like them,) and said, “Yep, close enough.” Of course in Florida, weather observations at any given point in time have no relevance on the state of the weather in 30 minutes.

[roaddog727]

For both my Bachelors (Economics) and Master’s Degree’s (Finanace/Quantitative analysis) I did quite a bit of Econometric modeling. What was pounded into our heads from day one of each and every class was “You MUST preserve the SANCTITY of the Data”. From the raw data, all analysis flows.

So, when you come up with a new analysis and refer back to your original data source, your analysis can undergo peer review, and as such your findings/conclusions can be repeadted based on the original dat and your transfomations performed on the original source data.

That’s the entire point of source data. Recreatability of analysis. Also, if you get the model right, you should have an expected forecating result +/- some margin of error (standard error of estimate or some such) Typically, the models I ran had an R^2 of somewhere around .990 to .996. I would not even consider results of less than .970.

[dirtboy]

Ping your attention to a CRU article from before Climategate...

[Tunehead54]

AGW/EPA Fraud Bump! ;-)

[FBD]

bookmark

[listenhillary]

BTTT