[ncl-talk] seasonality remove and trend

[Debasish Hazra]

Thanks Dennis.

Deba

On Mon, May 30, 2022 at 9:39 AM Dennis Shea <shea at ucar.edu> wrote:

> One approach:
>
> [1] compute the climatological monthly means and then calculate the
> monthly anomalies: x(time,lat,lon)
>
>   xMonClm = *clmMonTLL* <https://www.ncl.ucar.edu/Document/Functions/Contributed/clmMonTLL.shtml> (x)                  ; monthly climatology   (12,lat,lon)
>   xMonAnom = *calcMonAnomTLL* <https://www.ncl.ucar.edu/Document/Functions/Contributed/calcMonAnomTLL.shtml> (x,xMonClm)    ; (time,lat,lon);  (0,1,2)
>
>
> [2]* xDtrend = dtrend_n
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/dtrend_n.shtml>*(xMonAnom,True,0)
> ; trend returned as attribute
>      printVarSummary(xDtrend)
>
> ======
> NCL is not a statistics package. It does have some simple regression
> functions.
>
> *regline*
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/regline.shtml>,
> *regline_stats*
> <https://www.ncl.ucar.edu/Document/Functions/Contributed/regline_stats.shtml>,
> *regCoef*
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/regCoef.shtml>,
> *regCoef_n*
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/regCoef_n.shtml>,
> *reg_multlin_stats*
> <https://www.ncl.ucar.edu/Document/Functions/Contributed/reg_multlin_stats.shtml>
>
> *bootstrap_regcoef*
> <https://www.ncl.ucar.edu/Document/Functions/Bootstrap/bootstrap_regcoef.shtml>
>
>
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/equiv_sample_size.shtml>
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/rtest.shtml>
>
> On Sun, May 29, 2022 at 2:30 PM Debasish Hazra via ncl-talk <
> ncl-talk at mailman.ucar.edu> wrote:
>
>> Hi,
>> I am analyzing 25 years monthly data (300 months) of a variable and would
>> like to measure the trend of the variable from beginning to the end of the
>> time. Questions are :
>> 1. How do I remove the seasonal variability by using a 12-month running
>> average prior to computing the linear trend. What function in NCL to be
>> used.
>> 2. Is there any weighted least square linear regression available in NCL
>> to be used for regression and trend analysis ?
>>
>> Appreciate your suggestions.
>> Thanks,
>> Deba
>> _______________________________________________
>> 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/20220601/b6cc74fb/attachment.htm>