[ncl-talk] Question on specx_ci confidence limits

[Dennis Shea]

There is no official reference. It was something from an applied math class
.

It requires the degrees of freedom and uses the inverse chi-distribution to
get the upper and lower bounds.
I am not sure where in the statistical literature this is described.

Sorry

On Mon, Sep 20, 2021 at 11:59 AM Lyndz via ncl-talk <
ncl-talk at mailman.ucar.edu> wrote:

> Dear NCL-experts,
>
> I would like to know the reference used for estimating the lower and upper
> bounds of the red noise spectrum. These are computed using the specx_ci
> function:
>
> https://www.ncl.ucar.edu/Document/Functions/Shea_util/specx_ci.shtml
>
> According to the description, the lag-1 autocorrelation was used to
> estimate these lines.
>
> I'll appreciate any help on this.
>
> -Lyndz
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at mailman.ucar.edu
> List instructions, subscriber options, unsubscribe:
> https://mailman.ucar.edu/mailman/listinfo/ncl-talk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.ucar.edu/pipermail/ncl-talk/attachments/20210920/5401f669/attachment.html>