Thursday, January 11, 2018

as another aside, the fact that climate scientists use averages at all is kind of...it's not mathematically sound.

averages are good when there's a fixed variable of some sort. you can take an individual's average over a fixed task (exam scores, track times, etc), or you can take a fixed task like exam scores and then average it out over various individuals (joe, sally, etc). what you're really doing with an average is repeating trials over and over, and trying to get a guess on a "test statistic" that exists in some platonic reality - the idea is that the average exists in some cloud somewhere, and if you repeat the trial often enough then you'll reveal it. i'm actually not a platonist at all, but you'd be surprised by the things you hear from grown men with math degrees, behind closed doors.

what the hell are you even trying to do by averaging out temperatures over the entire earth, in the first place? there's no test statistic to arrive at. you're not finding some ideal concept of earthly temperature readings. once you get a sequence of ratios in place, you can find the test statistic for the average of that sequence, but what does that mean if the "average temperature of the earth" is a wonky concept in the first place?  it's not devoid of meaning at all, but it's more of a contrived ratio to determine policy (like the cpi, or the unemployment rate) than it is any kind of reflection of anything meaningful. it only make any sense in the context of itself.

consider the following ten data points...

toronto: -25
moscow: -20
stockholm: -15
london: -10 
paris: -8
riyadh: 45
singapore: 46
calcultta: 47
cairo: 51
tehran: 52

my understanding of things suggests that that could very well be a typical january, mid-century.

average temperature: 16.3 degrees. of course, this is a crappy data set, i'm just making a point. but, that's completely fucking worthless as any descriptive measure - it's only useful in comparison to the next data point.

now, suppose that the readings for these cities in 1975 was as follows:

toronto: -13
moscow: -8
stockholm: -5
london: -2 
paris: 0
riyadh: 35
singapore: 36
calcultta: 37
cairo: 41
tehran:42

that's reasonable for 1975, huh? i'm not looking it up, i'm making a point; i should have looked this one up. and bullshitted the other. whatever. the average temperature of this data set is also 16.3 degrees

therefore, there was no climate change over these years? eh...

i should be offering a mathematical solution right now, but i'm not entirely convinced that the idea of modelling the earth in this way makes sense at all.

you hear this push-back: weather is not the same as climate, weather is not the same as climate. i end up doing it myself sometimes. it's an easy way to explain away the fluctuations.

i'm not really convinced that you can talk about a planet's climate at all. i mean, the ratio has a purpose, but it doesn't actually physically mean anything. there is no "earth's climate", there is a collection of overlapping systems, and really several different climates that develop where these systems intersect.

and, right now, it looks like the north and south are moving in opposite directions, as a consequence of opposite causes.

jagmeet singh must cut his beard.