Lucia at Rank Exploits poses this thought:
“I always think it’s best to ask yourself: Do I really think a particular method of looking at data is meaningful? Would I still believe this if the answers turned out “wrong” from my POV? Or, will I eventually find myself explaining my own method gives uncertainty bounds that are “too small”, when my method suddenly gives “wrong” (in my eyes) answers?
So, in this regard, I need to ask Tilo: Why throw out GISS Temp”?
Thanks, Lucia, for taking a different look at this. It is true that one can make arguments for and against GISTEMP as a valid/not valid data set.
My view is that this is a lot like voting for a company wide policy change on a company board of directors.
Let’s say this issue was on a company board of directors decision to choose to make changes to policy related to employee comfort. Some employees complain that the work environment is too hot and they have been suffering a long term effect. The board decides to hire four consultants with the mandate: “tell us if we should expend the money to replace all of our a/c units company wide in all of our world locations. The cost will be huge, so we need to know before we make a policy change to do this.”
One of the consultants to members of the board who strongly advocates the policy change also has been lobbying company staff worldwide and other board members with the data he has collected and collated that shows that the trend is shifting in the direction that he advocates. As consultant, he is also the creator of one of the datasets used to evaluate the policy change.
Now when the time comes to make the decision, the board brings in all the data sets from consultants. They look at each one and see that the majority of them have no change in the last 11 years that supports the policy change to put in new a/c units. Yet the one consultant that has been pushing this policy change gives an impassioned speech that his data set tells a story that the others do not. Some of the board members who are skeptical of this person and his data that supports the policy change do some research of their own. They discover that the dataset created by the consultant who advocates the policy change has been adjusted at many data points, almost without exception in favor of the policy change. Some board members also learn of some math errors in the data, point out the math errors, and also some of the questionable ways individual data points have been adjusted. the consultant shrugs and retorts “you’re just a bunch of court jesters”.
Meanwhile, it has been discovered that one of the business friends of the consultant who has been lobbying board members and staff has a company that trades in air conditioner systems. That person has been traveling to all of the worldwide offices of the company and lobbying the employees to tell them that their work environment is indeed getting hotter, and that the data from his friend the consultant proves it beyond the shadow of a doubt. He tells them that his friend the consultant uses special techniques to find the “real” trend in the data and that the other datasets aren’t as valuable as this one. He urges the employees to form pacts and unions to lobby the corporate board to make a change. The employees do just that.
The board looks at the data, they listen to the impassioned pleas of the employees, and they also listen to the one consultant who calls them “court jesters”, and his friend the a/c salesman, who says “the time is now, you must act now’. But a few employees that are concerned that the expense the company may be about to undertake is unwarranted, they think the work environment is just fine, and the “solution” may hurt the company more than help it.
One of the employees finds that in the largest company facility, 78% of the temperature sensors in used to collect environmental data in that facility have been installed incorrectly, and shows that they are too close to equipment that produces waste heat. They also discover that two of the consultants use IR sensors to get the data, but that the other two consultants are using the direct measure environmental sensors, 78% percent of which in the largest company facility are installed incorrectly. These few employees that discover this also lobby the board by pointing out some of these issues with the datasets.
So it is time for the board to vote. The one consultant who has lobbied the board most heavily says “don’t worry about that 78% of the problematic environmental sensors in the biggest facility, I can adjust for that.”
But then one of the other skeptical board members says: “The employee that found this says “How can you adjust for these if you’ve never seen or visited them? How can you know they are all equally biased or not?”. And, isn’t it true that in some of the data you presented, there were no sensors present, and some of the data was interpolated by you, particularly at the far ends of the building?”
The consultant says: “I stand by my data and methods, and if you don’t do something soon, your facility may reach a tipping point where you can no longer keep it cool enough to work in, your company productivity will tank.”
One of the board members says, “Ok lets stop and look at this differently”. “What if we simply ignore the dataset from the consultant who calls us “court jesters” and has the buddy who’s the air conditioner salesman?. Look, now there’s no trend in the last 11 years”.
So 3 of 4 datasets, each presented by independent consultants are in front of them and show no change in the past 11 years. The one that does show a trend has been heavily lobbied and has been shown to have errors in measurement by environmental sensors and questionable data adjustment methods applied. Plus the consultant who prepared it has insulted those members who dared to question his data and methods, and he is the only one of the four consultants who has links to the air conditioner salesman, as it was discovered that the air conditioner salesman invited the consultant to speak at one of his employee rallies.
How do you think the board of directors will vote on this policy change?