CDO

I guess, daypctl is correct, but I'm not an expert in these indices.

In fact, i think that we must use here not daypctl, but timpctl!

timpctl needs 3 inputs:
- ifile1: where we want to compute the Xth percentile
- ifile2: the result of timmin
- ifile3: the result of timmax

But why need to know min and max values (the results of timmin and timmax) to calculate a percentile?..
If anybody has an idea?

Thanks.

Natalia

The percentile algorithm for timpctl uses histograms with minimum and maximum bounds given in ifile2 and ifile3, respectively.
I'm also not an expert in these indices but I think the 90th percentile TNn90 of daily minimum temperatures of the reference period have to be computed with runpctl,90,5.

Hi,

Thank you for your response.

I calculated TNn90 with runpctl.
But after, when i calculate eca_tn90p:
cdo eca_tn90p tnfile tnn90file ofile

(where:
tnfile: time series of daily min (period: 1850-2005)
tnn90file: the result of runpctl for the reference period: 1960-1990 )

I got the following error:
Input streams have different number of time steps!

But how it is possible they have the same number of steps?

So, it is still not clear.
If there are the cdo experts in this kind of indices, could you forward my question, please?

Best regards,
Natalia

Normally, tnn90file must contain TNn90 which is calculated as the 90th percentile of daily minimum
temperatures of a five day window centred on each calendar day of a given climate reference period.
It means that in output TNn90 must have the following dimentions (366, d2, d3).
But rupctl doesn't do it like this...

Natalia

Hi Natalia,

have you tried
cdo ydrunpctl,10,5 BASE -ydrunmin,5 BASE -ydrunmax,5 BASE PCTL10
?

Arne

Hi Arne,

Thank you for response!

Yes, i tried:
cdo ydrunpctl,10,5 BASE -ydrunmin,5 BASE -ydrunmax,5 BASE PCTL10

But after, when i do:
cdo eca_tg10p TG PCTL10 TG10P

I get again:
Input streams have different number of time steps!

Natalia