[ncl-talk] diurnal cycle of precip plot

[Guido Cioni]

Hey Adam,
I saw those plots already more than once. Yes, they are pretty nice :) 

I would say that making the plot itself is no hassle when you already have the data as a 2-D array: they are normal contour plots with raster fill and a user-defined color scale.

What I think is a little bit more “difficult" is doing the data analysis BEFORE. I never managed to read the method part of papers presenting such results but I believe they explained in detail what they did. So this should be more a “method” question and not an “NCL” question since the algorithm can then be coded without problems once you have it.

I don’t know what they use for the analysis but I believe rain rate from TRMM as function of time and space. Then what you have to do for every grid point is:
extract the JJA average as function of time (I would do that with CDO) 
compute the UTC time relative to the maximum of precipitation over a 24 hours window (this can be done easily with maxind in NCL when you have the JJA averages) 
convert the time of the precipitation peak to local time: that’s the tricky part. Depending on the time variable that you have I would create an array which holds all the local time shifts and then use it to correct the time for every grid point by checking whether it is into a determined area. Note that in Python there are some ready-to-use modules to use for this task (e.g. pyephem).

Hope that helps.


> Il giorno 11 nov 2017, alle ore 03:11, Adam Herrington <adam.herrington at stonybrook.edu> ha scritto:
> 
> I am trying to compute the mean local time of maximum in precipitation, (as a map plot). These plots are becoming more common these days (attached is an example from Bechtold et al 2014), and would appreciate any discussion on what methods are used to create these plots.
> 
> Thanks,
> 
> Adam
> 
> <Screen Shot 2017-11-10 at 7.09.46 PM.png>_______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20171111/337b4642/attachment.html>