/sci/ - Science & Math

>>10047477
>I know for a fact that weather forecasting is unreliable as fuck beyond 24 hours because we just don't have good time series predictors to deal with such chaotic systems yet.

This is literally the stupidest thing I've read today

>>10047497

Weather models were some of the first strange attractors studied. I also encourage and challenge you to take literally any weather related time series and try to fit any kind of model into it.

If you want I can look up a dataset and give you all the time you want until you shut up. Not even the best predictors can reliably do shit with weather data beyond the scope of a few hours.

t. statistician

Bump.
Why is this board so scared of actual discussion?

>>10047539
Nobody actually knows any of this shit and just argues from authority. That’s by design though because only climate scientists can really parse the data, so you have to take it on faith that they know what they are doing.

It’s also why the only “solution” to climate change is taxes, as anything (read: geoengineering) that could blow up in their face as a result of piss poor understanding is dismissed out of hand.

>>10047477
>It's another /pol/tard doesn't understand the difference between weather and climate episode

>>10047576

>I use the words weather and climate properly on my post.
>Say no time series prediction works for weather.
>Ask what time series prediction works for climate.

Stop evading. Give me an answer. I want to become rich and apply the same model to forecast the price of stable stocks in 10 years from now.

>>10047595
Dude, it’s literally extrapolation of historical data with fudge factors thrown in. There is no “model”.

>>10047606

Do you even read the shit you type?
Since you don't like the word "model": What "extrapolation" do they use?

>>10047611
>plot temperature data points
>get a ruler
>place ruler in the middle of data points
>draw a line that goes through and past plotted points

That’s it.

>>10047629

Unironically least squares regression? How the fuck is that acceptable.

>>10047645
>How the fuck is that acceptable
It's settled science!
t. 98% of climate "scientists"

>>10047684

I can understand using least squares regression or some shit like this to say that there is an uptrend. I just really want to understand how they arrive at shit like "7 degrees in 100 years!".

That is just ridiculous.

>>10047690
>I can understand using least squares regression or some shit like this to say that there is an uptrend

The uptrend is what the %97 all agree upon, the magnitude has no consensus at all. The “predictions” for actual temperature rise are fear mongering for the sole purpose of acquiring funds.

>>10047690
>That is just ridiculous.
Well, it is. In practice, they more likely fit a [S]ARIMA[X] model, calculate the conditional mean and variance of the process 100 years ahead, and take the upper 95% cofidence bounds of that as the maximum reasonable temp increase - which then gets reinterpreted as a forecast of the future mean.

High forecasts are politically safe as well.

As an aside, I hat a look at some climate science curricula a while back. They stidy a diversity of subjects, and do not get much hardcore modelling.

>>10047645
>He believed a random dude on /sci/ when he could have just spent 10 seconds googling 'Climate models' and seen for himself what the different models are
https://en.wikipedia.org/wiki/Climate_model
https://en.wikipedia.org/wiki/Geophysical_Fluid_Dynamics_Laboratory_Coupled_Model

>>10047709
Back to /x/ retard

>>10047477
>climate change is like the stock market
L0Lno fgt pls

>>10047477
>>10047539
>>10047556
>>10047595
>>10047606
>>10047645
>>10047684
The rising temperature is literally just an average. You average all the weather station measurements for a given range of dates (could be each week, or the whole year) after adjusting for systematic errors, and you plot them. That's why the plots showing the upward temperature trend are the "global mean surface temperature." Turns out averages of enormous numbers are much less chaotic than local weather fluctuations.

If your theory is that anthropogenic emissions are the primary cause, we can expect the trend to continue in proportion with that activity, which is expected to increase due to increases in affluence and the total population. There are several different projections of "optimistic" and "pessimistic" trajectories, partly because it is very difficult to predict the changes in affluence and population other than "a steady increase," and because some things like political action and changing societal views of the issue are literally impossible to predict.

>>10048308
enormous data sets*