Hacker News Reader

Comment by pfarnsworth | Hacker News Reader

There is a 10F difference in temperature between my upstairs and downstairs without any heat turned on. There is a 4F difference in temperature between the front of the house and the back of my house.

There are so many factors as to how temperature is calculated, how exactly can we take a temperature today and compare it to values from years ago? I'll buy that satellite imaging will produce values that are useful over large areas, but those only go back 20-30 years. To compare against values before that seems like pseudoscience at best.

View on HN

Do you think people have systematically moved the locations of thermometers to warmer places? There's a thing called averages, medians, and other statistics that can remove or put error bars on that kind of uncertainty.

Please don't conflate some sort of political viewpoint to my my strictly scientific criticisms. I didn't say anything except that the idea of comparing data today to many decades ago seems ridiculous when there are so many factors involved with what the temperature is, including location, sun, wind, etc. I'm more interested in accuracy rather than politics or news-grabbing headlines.

Regardless, your point doesn't have any validity. There is a systemic difference between the front of my house and the back of my house in terms of temperature. My thermometers are extremely accurate, with 0.5F if I remember correctly. And the average/median would still show a net 4F difference. So no, your point is invalid, there is still a problem with the data due to location.

It's not a political criticism. For there to be a steady, statistical rise in temperature over time due to how people measure temperatures, you have to come up with some mechanism for how that could possibly happen. It would have to be a deliberate effort to game temperature readings so they look warmer slowly over the course of decades. Someone would have to come up with a rate of change such that thermometers don't all go from a shady place to direct sunlight in one year, so it looks all gradual.

Your argument holds no water. We're not talking about one person with a thermometer they've been moving to increasingly warmer parts of the house for a hundred years. If you don't trust the data sets we have, then you might as well throw almost all science away because most of it is built on shakier data sets than this.

[Not the op] I neither think that someone moved a thermometer to a warmer place. The problem is that some weather stations were originally placed outside a city many years ago, but the city grow until the weather station is surrounded by the city. Cities are 2°F or 3°F warmer than the countryside, so this creates a fake temperature. Obviously, this changes are not ignored, so the temperature tables have corrections for this kind of problems.

Some of the weather stations were moved the countryside, to avoid the problem or because the office they were placed was relocated. So when the place changes, it necessary to adjust the temperature tables to compensate any differences between the locations.

So, there are a lot of corrections that have to be calculated correctly to have reliable temperature values.

That's a good point about cities growing to encompass weather stations, and it would be a slow steady change. But like you said, that can be corrected for and I presume you or the gp are not the first to have thought of it. So I still strongly disagree with the assertion that using this data is pseudoscience. Almost everything in science has to compensate measurements for known anomalies or biases.

Especially after climate gate where it was revealed that a warming factor was applied to the raw data, forcing a hockey stick....

Please don't spread this nonsense. "Climate gate" was a nonevent. https://en.wikipedia.org/wiki/Climatic_Research_Unit_email_c...

Reply on news.ycombinator.com