[Met_help] [rt.rap.ucar.edu #40169] History for about the calculation of intensity in the mode step

[RAL HelpDesk {for John Halley Gotway}]

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Dear sir:

     Could you tell me the specifc formula of how to calculate the intensity in the mode step? I have checked the user guider and the paper 'objected-based verification of precipitation forecast' published in Monthly Weather Review,2006. But still could not figure out how? Does the high percentiles represent heavy rainfall?

     Thanks!

Kefeng

2010-08-18 



zhucoffee 


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Re: [rt.rap.ucar.edu #40169] about the calculation of intensity in the mode step
From: John Halley Gotway
Time: Wed Aug 18 16:23:36 2010

Kefeng,

Yes, that's basically correct.  First, MODE can be run on any pair of
gridded data fields in GRIB format.  However, it's typically been run
on precipitation data.

Objects are defined in each field using a convolution and thresholding
process.  Each of the defined objects is really just a bunch of
contiguous grid cells.  The percentiles of intensity for each
object are computed as follows:
- Inside each object, MODE looks at the RAW data value for each grid
cell inside of it.  For an object with an area of 100 grid cells,
they're be 100 corresponding raw data values.
- MODE sorts those raw data values, and extracts the percentiles from
that distribution of values.  The 10th, 25th, 50th, 75th, and 90th
percentiles are written to the MODE output file.
- Additionally, MODE writes out a "INTENSITY_SUM" column which is the
sum of all the raw data values.  For precip, you can think of this as
a volume of rain in the object.

Therefore, when comparing two objects, if one has a higher value for
the 90th percentile of intensity, then that object contains more
intense precipitation.

Hope that helps.

John Halley Gotway
met_help at ucar.edu

RAL HelpDesk {for zhucoffee} wrote:
> Wed Aug 18 16:00:03 2010: Request 40169 was acted upon.
> Transaction: Ticket created by zhucoffee at 163.com
>        Queue: met_help
>      Subject: about the calculation of intensity in the mode step
>        Owner: Nobody
>   Requestors: zhucoffee at 163.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40169 >
>
>
> Dear sir:
>
>      Could you tell me the specifc formula of how to calculate the
intensity in the mode step? I have checked the user guider and the
paper 'objected-based verification of precipitation forecast'
published in Monthly Weather Review,2006. But still could not figure
out how? Does the high percentiles represent heavy rainfall?
>
>      Thanks!
>
> Kefeng
>
> 2010-08-18
>
>
>
> zhucoffee

------------------------------------------------