Query on Resampling of Maps and SMCE

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Query on Resampling of Maps and SMCE

Newbie
To Whom it May Concern

I have created a number of rasters for use in a SMCE analysis. Typically, the coordinate boundaries of the various maps differ depending on the extent of the data, so I have had to resample the maps to a common georeference entitled "Study Area". To illustrate my point I have included two maps below as an example. Map 1 represents slope shading in a study area represented by a red circular boundary. Map 2 represents protected areas within the same study area.


Map 1


Map 2

As can be seen from the above images, a large portion of the study area for Map 2 has no data and when resampled, displays a question mark ("?"). It is only the rectangular area which encompasses the "protected area" data which has info.  My classification within this rectangular area in the class domain created is: red = "protected area"; and white = "developable area". Beyond this there is no data and question marks appear when selected.

My concern with the above is that when standardizing the criteria, a large portion of Map 2 will not contribute to the aggregated scoring of the pixels. In other words, the developable areas beyond the protected area will not be accounted for as no scoring will be assigned to these pixels. Is there any way to deal with this? Any assistance will be much appreciated!

Thank you for your time.

Regards,

Newbie
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Query on Resampling of Maps and SMCE

Bas Retsios
Hello,

First of all, ensure that all your maps are "cropped" to the actual
study-area. SMCE decides the study-area based on the first map that is
added to the criteria tree, and during computation later on, all pixels
in the selected study-area are candidates.
Then, most likely you would want to use Map 2 as a "constraint",
allowing pixels that are "developable area", and blocking pixels that
are "protected area" (or the inverse, depending on the objective). In
this respect, consider that ILWIS outputs a "?" (nodata), if one of the
inputs of a mathematical expression is "?" (thus 1 + "?" = "?", "?" * 5
= "?" etc.). As a consequence, if one of the input maps that are used in
SMCE has nodata pixels, those propagate all the way to the suitability map.
I assume that this is unwanted in your case. Then you would want to
compute a new map from Map 2, with "developed area" and "protected area"
pixels, instead of "protected area" and "?" pixels. This is a simple
formula, to be entered in the ILWIS command-line (the white space
towards the top of the ILWIS window, just under the catalog filtering
buttons).
Two options:
1. Generate an output map that uses a "Class" Domain of two items:
"developed area" and "protected area"
Areas := ifundef(Protected_Areas, "developed area", "protected area")
2. Generate a boolean map, whereby True would represent the one class,
False the other. The best is to assign "True" to the allowed pixels,
however during the Standardization stage in SMCE you can still decide
whether a boolean map as an indicator is used as-is ("True=pass"), or
inverted ("False=pass").
Allowed_Areas := ifundef(Protected_Areas, True, False)
Press <ENTER> to execute the command-line.
It is recommended to name maps to what they represent, to prevent confusion.
Note that you can also use the newly computed map (Areas or
Allowed_Areas) as a factor, instead of a constraint. Then the two
classes, or the boolean values, would get a score in the Standardization
stage. It all depends on your objective.

Best regards,

Bas Retsios
ILWIS team

On 9/3/2016 3:44 PM, Newbie wrote:

> To Whom it May Concern
>
> I have created a number of rasters for use in a SMCE analysis. Typically,
> the coordinate boundaries of the various maps differ depending on the extent
> of the data, so I have had to resample the maps to a common georeference
> entitled "Study Area". To illustrate my point I have included two maps below
> as an example. Map 1 represents slope shading in a study area represented by
> a red circular boundary. Map 2 represents protected areas within the same
> study area.
>
> <http://ilwis.forum.52north.org/file/n4025863/Slope_Shading.jpg>
> Map 1
>
> <http://ilwis.forum.52north.org/file/n4025863/Protected_Areas.jpg>
> Map 2
>
> As can be seen from the above images, a large portion of the study area for
> Map 2 has no data and when resampled, displays a question mark ("?"). It is
> only the rectangular area which encompasses the "protected area" data which
> has info.  My classification within this rectangular area in the class
> domain created is: red = "protected area"; and white = "developable area".
> Beyond this there is no data and question marks appear when selected.
>
> My concern with the above is that when standardizing the criteria, a large
> portion of Map 2 will not contribute to the aggregated scoring of the
> pixels. In other words, the developable areas beyond the protected area will
> not be accounted for as no scoring will be assigned to these pixels. Is
> there any way to deal with this? Any assistance will be much appreciated!
>
> Thank you for your time.
>
> Regards,
>
> Newbie
>
>
>
> --
> View this message in context: http://ilwis.forum.52north.org/Query-on-Resampling-of-Maps-and-SMCE-tp4025863.html
> Sent from the 52° North - ILWIS Community Forum mailing list archive at Nabble.com.
> _______________________________________________
> Ilwis mailing list
> [hidden email]
> http://list.52north.org/mailman/listinfo/ilwis
> http://ilwis.forum.52north.org
> Please respect our mailing list guidelines:
> http://52north.org/resources/mailing-lists-and-forums/guidelines

_______________________________________________
Ilwis mailing list
[hidden email]
http://list.52north.org/mailman/listinfo/ilwis
http://ilwis.forum.52north.org
Please respect our mailing list guidelines:
http://52north.org/resources/mailing-lists-and-forums/guidelines
Loading...