[GRASSLIST:7028] Re: calculating center of intensity

[Charles Ehlschlaeger]

Tom,

I purposely threw in that caveat because I assumed that a human would 
have to decide what a storm pattern was. If an algorithm automatically 
found storm cells of contiguous positive rainfall values, there might be 
splinter cells a meteorologist would consider part of the same pattern. 
Ergo, I would expect the user to zoom to a region with a single rain 
event and mask out any areas they didn't consider part of that storm. 
Then, a variation of the procedure below would do the trick. As Hamish 
suggests, r.univar is probably better than r.volume.

sincerely, chuck

Thomas Adams wrote:

> Chuck,
>
> Does this assume you only have one raster map cell shape? What if 
> there are multiples, which is what I think Jason has? I think, then, 
> you would get a single center of mass for the cluster of separate cells.
>
> Tom
>
>
> Charles Ehlschlaeger wrote:
>
>> I don't have GRASS working on my computer to check this out, however, 
>> here is a procedure to calculating the center of a shape:
>>
>> Assume you have a raster map "shape" where the value of 0 is at all 
>> places w/out the shape and a positive value where the shape is 
>> located at.
>>
>> 1) Find the total volume of the shape w/:
>>
>> 1a) r.mapcalc command function:
>> one = "1"
>>
>> 1b) r.volume d=shape clump=one
>>
>> (There should be a better way of doing this.)
>>
>> 2) Calculate volumeOnAxis of shape in North-South dimension with the 
>> r.mapcalc command:
>>
>> NSvolumeOn = shape * y()
>>
>> 3) Calculate massOnAxis of shape in East-West dimension with the 
>> r.mapcalc command:
>>
>> EWvolumeOn = shape * x()
>>
>> 4) Find the total volume of NSvolumeOn with:
>>
>> r.volume d= NSvolumeOn clump=one
>>
>> 5) Find the total volume of EWvolumeOn with:
>>
>> r.volume d= NSvolumeOn clump=one
>>
>> 6) The coordinates of the center will be
>> east = total volume of EWvolumeOn / volume of shape
>> north = total volume of NSvolumeOn / volume of shape
>>
>> A little csh and awk programming could replicate Idrisi's center program.
>>
>> <INSERT DISCLAIMERS ABOUT USING AT OWN RISK...>
>>
>> sincerely, chuck
>>
>> Markus Neteler wrote:
>>
>>>On Thu, Jun 02, 2005 at 03:42:36PM -0400, Jason Horn wrote:
>>>  
>>>
>>>>Does anyone know of a way to calculate the center of a shape in a  
>>>>raster file based on shape and pixel values?  For example, imagine a  
>>>>NEXRAD RADAR image showing a storm cloud.  Could GRASS be used to  
>>>>calculate the center of the cloud?  I'm not having any luck finding a  
>>>>grass raster function to do this.
>>>>    
>>>>
>>>
>>>I did a quick google search for
>>> nexrad storm cell identification 
>>>or
>>> nexrad  SCIT
>>>which a couple of hits. Some programming might
>>>be needed. I don't know if these algorithms are freely
>>>available.
>>>
>>>Markus
>>>
>>>  
>>>
>>
>>
>>-- 
>>Chuck Ehlschlaeger, Associate Professor & GIS Center Director
>>Department  of  Geography,      Western  Illinois  University
>>306C Tillman Hall,  1 University Circle,  Macomb,  IL   61455
>>cre111 at wiu.edu,    phone: 309-298-1841,     fax: 309-298-3003
>>
>
>-- 
>Thomas E Adams
>National Weather Service
>Ohio River Forecast Center
>1901 South State Route 134
>Wilmington, OH 45177
>
>EMAIL:	thomas.adams at noaa.gov
>
>VOICE:	937-383-0528
>FAX:	937-383-0033
>


-- 
Chuck Ehlschlaeger, Associate Professor & GIS Center Director
Department  of  Geography,      Western  Illinois  University
306C Tillman Hall,  1 University Circle,  Macomb,  IL   61455
cre111 at wiu.edu,    phone: 309-298-1841,     fax: 309-298-3003

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/grass-user/attachments/20050606/20f1d470/attachment.html