0
$\begingroup$

I would like to ask you what is the best (or the right way) to calculate the rate of change in air temperature over a period of time and then see how this rate changes over time. Daily air temperature data do not follow a line, but it goes up and down.

So, to estimate the seasonal rate of change in temperature: Can I use simple linear regression and take the slope as the value of change or should I calculate the average rate of change between my first and last date for each season? or between the first day and the highest point? or what is the right way to estimate it?

Thank you for your comments, advices and answers!

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

First plot the data. Such data would probably be cyclic. Options for probably modeling include:

  • autoregressive integrated moving average (ARIMA),
  • partial differential equations.
Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.