[Qgis-user] combining fragmented forests (RMG)

[Falk Huettmann]

Hi there,

greetings;
I have seen and done such things, a lot,
and I would like to give you a message of caution about it.

One serious problem is here that it includes error inconsistencies (sheet
edges vs sheet centers),
another one is about classes and patches that are incorrect to start with
(all sheets; merged or not).

In other words,
there are several types of errors in such 'technical artefacts' aka  merged
forest patch maps;
they "cannot be used for navigation", to put it politely.
It's a blurr of things, and a major problem with forest maps and forest
inventory maps out there.

I know of forest maps and regions that have accuracies of 40%, if
ground-truthed.
So yes, they look great on the screen and in GIS but are close to random
(50% accuracy)
if ever used and interpreted for applications.
Please be aware; happy to hear about a fix; beyond technical GIS ones, or
rubber sheeting.

Keep me posted please; thanks.
Very best
   Falk

PS There is currently no single accepted map of global forest cover;
FAO, IUFRO etc cannot really agree on such details whatsoever.
And yes, Remote Sensing products, LIDAR and air photo interpretation
can be rather poor for quality, usually they are.











On Wed, May 6, 2020 at 2:50 PM Michael Harte <mharte at ozemail.com.au> wrote:

> Reiko,
>
> You could try something like this:
>
>
>    1. Merge all forests into a single file (if you haven't already);
>    2. Create a new field in the forests table called something like
>    "border"
>    3. Create lines that trace the panel boundaries, that separate the
>    forests you want to join;
>    4. Buffer each of those lines by some small distance  (try 1 meter);
>    5. For each buffer "Select by location" all features in forests that
>    intersect buffer;
>    6. Use the field calculator in the table to enter a unique identifier
>    for the selected features in the "border" field;
>    7. Do this for each buffer;
>    8. Dissolve the forest layer on the border field.
>
> Feel free to get in touch.
>
> Cheers,
>
> Mike
>
>
>
> Message: 1
> Date: Wed, 6 May 2020 14:55:30 -0400
> From: RMG <reikogoodwin at gmail.com> <reikogoodwin at gmail.com>
> To: qgis-user at lists.osgeo.org
> Subject: [Qgis-user] combining fragmented forests
> Message-ID:
> 	<CALwVOrJq9njBAdtvVeXLVsVT8SEjcDfp5mp=AYdoOi93w3TWxA at mail.gmail.com> <CALwVOrJq9njBAdtvVeXLVsVT8SEjcDfp5mp=AYdoOi93w3TWxA at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Hello,
>
> I have five shapefiles that contain thousands of fragmented forests that I
> created from a NDVI raster of a large protected area in West Africa, which
> I separated into five panels. Now, how do I stitch thousands of the
> fragmented forests at the borders?
>
> Each shapefile's attributes have ID, count (number of units), area, and
> perimeter, so I can visually recognize which fragments need stitching, but
> because there are so many, I don't think I can do it manually.
>
> Any help would be appreciated.
>
> Best wishes,
>
>
> Reiko Matsuda Goodwin
> Comoé Monkey Project <https://www.facebook.com/ComoeMonkeyProject/> <https://www.facebook.com/ComoeMonkeyProject/>
> Guenon Conservation Community<http://facebook.com/GuenonConservationCommunity/> <http://facebook.com/GuenonConservationCommunity/>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20200506/c823b462/attachment-0001.html> <http://lists.osgeo.org/pipermail/qgis-user/attachments/20200506/c823b462/attachment-0001.html>
>
> ------------------------------
>
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20200506/f5df9e17/attachment.html>