I spend a great deal of time on the internet trying to hunt
down large scale data in Idrisi format. I
found very little. The fast
majority of data I came across was in ArcInfo or ArcView format.
I eventually found data that met most of my criteria – small/medium
scale land use data from a familiar place.
The Read Island forest inventory data used for this project
came from Jerry Maedel the
GIS/RS Coordinator at
the Faculty of Forestry, UBC. I
first discovered the existence of this data set on one of my many search on the
world wide web at www.interchg.ubc.ca/firms/read/data.htm.
This site was an old window into a project, called the Community Map
Analysis Project, that was utilizing Idrisi to study land use and land cover in
three regions of the province. The
data was part of a project done in 1996 to study biodiversity conservation
planning at a sub regional level However, the web page itself is mostly
inoperable; all of the presumably downloadable data at this site was no longer
available, so I contacted Jerry personally and he sent me the data via email.
The data consisted of three three coverages of the island: roads, contour (elevation) lines, and forest inventory
polygons. All layers were in
ArcInfo (vector) format. While all three layers were imported into Idrisi,
I ended up only using the forest and contours data.
Forest Coverage layer:
The ArcInfo forest cover data consisted of over unique forest 400 polygons. The accompanying attribute file for the forest coverage had
over 50 different fields of attributes, attributes
that are used by the logging companies and the Ministry of Forests
(Mof) with
regards to forest management. However,
may of the fields were empty of attribute data, such as breast height, site
index, etc, all important factors in determining forest health, management, and
future growth. The attributes I
utilized for this project from the forest cover attribute files were:
tree species (Douglas fir, Western Red cedar, Lodgepole Pine, Sitka
Spruce, Balsam Fir) , age class, height class, crown class, and environmentally
sensitive areas (ESA). The
remainder of the data was either incomplete or I was unable to decipher what it
pertained to! Click here
to view more information on the meta data.
Contours layer The contours image and
associated attribute file was much more simple than the forest coverage
data. This layer consisted of elevation contours spaced approximately
at 20 foot intervals.
Getting the data into Idrisi was done by
converting the three coverages into shape files in ArcView, and then
importing them into Idrisi using SHAPEIDRISI import function. This was very time consuming and involved a number of hit an misses.
The initial process required importing the forest layer as a vector file
and entering in the essential projection and co-ordinate data. At
first I set my resolution far too high. The initial new x and y
values was approximately 7000 x 12000 respectively. This amounted to
nearly 100 million pixels in my coverage and a VERY high resolution.
High enough to make out the the whites of the eyes of the inhabitants of
the island! (I clued into this only after crashing the computer several
times and creating coverages of nearly 400 megabytes each.) Needless
to say, I lowered my resolution to 1 pixel to every 14 meters.
The process of
creating the different coverages was fairly straight forward after I had projected
and rasterized my 'base' coverage. I simply created an attributes
values file from a specific attribute such as tree age and then assigned
this new attribute file to my base map.
The poor and fragmentary nature of the Metadata meant that had to do a bit of research on my own. I utilized a couple of forestry books in the library along
with the sparse metadata to decipher the meanings of the various fields in the
forest cover attribute table. The
more I read, the more I realized how complex and broad forest management is!
