[gradsusr] Probability precipitation map using ensemble forecast

[Hosmay Lopez]

Hi Miguel,

This is straightforward in GrADS if you use a combination of functions. It goes like this:

If you have a variable for precipitation named precip and let say, 100 ensemble members. You do:

set e 1 100
define prec5mm = const( const( maskout( precip, precip - 5), 1), 0.0, -u)

The maskout function sets the values to undefined where precip < 5
The inner const function sets the remaining defined values to ones.
And the outer const function sets the undefined values to zeros.
So the new variable prec5mm is made of 1 wherever precip > 5 and 0 otherwise.

Now you average for all your ensemble members and multiply the results by 100 to get % of ensemble members with precipitation above 5mm .

set e 1
define percentage5mm = 100 * ave( prec5mm, e=1, e=100)

Hope this helps.

Hosmay Lopez
UM/CIMAS/RSMAS NOAA/AOML/PHOD
4301 Rickenbacker Causeway,
Miami, FL 33149, USA
Phone: (305) 361-4329

________________________________
From: gradsusr-bounces at gradsusr.org [gradsusr-bounces at gradsusr.org] on behalf of Miquel Bernis [bernisfeina at gmail.com]
Sent: Friday, June 19, 2015 3:26 AM
To: GrADS Users Forum
Subject: [gradsusr] Probability precipitation map using ensemble forecast

Hi everyone,

Do you think is there any way to plot a forecast probability precipitation map, using an ensemble forecast? I mean, calculating which forecasts of the ensemble give, for example, an amount of rain > 5 mm and which not, for an specific moment of time, and displaying a contour map of the % you can take from that. Watching to the grads functions it seems not possible, but maybe some of you, with more experience on grads than me, can imagine a way to do that.

Thanks a lot!
Miquel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20150619/97935f45/attachment.html