Information on temporal departures

[gkb999]

Greetings
From the page : https://xcdat.readthedocs.io/en/stable/generated/xarray.Dataset.temporal.departures.html
Departures calculated say: if no Reference_period is provided, the climatological reference period will be the full period covered by the dataset. In that case, the anomaly should be zero, right?
I used that option, where I left xcdat to use the default way to calculate anomalies and I don't get zero yet.
Any explanation would be appreciated

[pochedls]

@gkb999 - I don't think the anomaly should necessarily be zero. If I have 6 temperature values: [255, 260, 265, 265, 260, 255] and I use the full period (all values) then the mean is 260. I subtract the mean from each value to get the departures: [-5, 0, 5, 5, 0, -5]. So there are zero values, but the anomalies are not always zero.

[gkb999]

Thanks a lot!! It answers what I want

[gkb999]

Also, does adding weighted to "True" strengthens the non-zero justification?

[tomvothecoder]

I converted this from a GH Issue to a Discussion Q&A post.

[gkb999]

@pochedls
Sorry to bring this back after a while.
Yet, again, if I change the baseline/reference period, I get different results.
Like my data was from 2002-2022
I used 2002-2012, got a different one, 2005-2015, got a different one and just used as below:
The one without any baseline/reference period: like: s_depart = ds_climo.temporal.departures("ts", "month") and got a different result.
The last default one, uses some baseline (climatology) to compute right?

Like I'm using 20 years of data, just curious if it uses all the 20 years as climatological average? if yes, then I should get 'zero'.
It might be using a default baseline based on the no.of samples?
Can you please explain?

[pochedls]

@gkb999 - please look at the previous post – you should not expect to get anomalies of zero if you use all data. For the following time series of pretend annual values:

[255, 260, 265, 265, 260, 255]

Base period	Climatological Mean	Annual Anomalies
All years	260.	[-5., 0., 5., 5., 0., -5.]
First 3 years	260.	[-6., -1., 4., 4., -1., -6.]
First 5 years	261.	[-5., 0., 5., 5., 0., -5.]

Note the departures are different if you use three versus five years as the climatological period.