Alias           Type             Width     Decimal	
-----------------------------------------------
Shape           FIELD_SHAPEPOLY     80
Ufi             FIELD_DECIMAL        6     0   - unique number
Area            FIELD_DECIMAL       18     5
Perimeter       FIELD_DECIMAL       18     5
Code            FIELD_CHAR          80     0 - Vegetation unit (map symbol)
Descript        FIELD_CHAR          64     0 - Description of unit
Categories      FIELD_CHAR          50     0 - vegetation categories
Feature         FIELD_CHAR          12     0 - (GA field)
Plotrank        FIELD_DECIMAL        1     0 - (GA field)
          

The scanned map was included in the Christmas Island GIS. It was digitised by Geoscience Australia in mid October 2001 on behalf of Parks Australia North Christmas Island.

Newveg.shp is the revised version of Vegetation.shp. Dr. Pete Green a Research Fellow from EA created this .shp file by taking the original vegetation file, making all the vegetation types semi-transparent, and superimposing this theme over a composite aerial photograph of the island (Christmas.ers, generated from the 1987 aerial photos). Dr. Green then adjusted the position of the clearing boundaries (category SV) around all previously mined fields, expanded these polygons to include all major roads, and added and subtracted polygons as required. In all cases Dr. Green tried to pick the original boundary of the clearing - there has been substantial regrowth around the borders of some mined fields, but this regrowth can be distinguished from the adjacent forest by its more regular canopy. Dr. Green made a concerted effort to ONLY adjust the position of forest/clearing boundaries, and NOT forest/forest boundaries. The aerial photos are excellent for judging the position of the former, but we have no prior evidence for adjusting the latter - as far as Dr. Green knows, the positions of forest/forest boundaries have never been ground-truthed.

Where boundaries were indistinct on Christmas.ers, Dr. Green occasionally made reference to two sets of black and white aerial photographs of the island taken in 1976 and 1981.  In September 2003 this dataset was converted to the MGA(GDA94) projection using an x y shift  (550015 = x, 8780001 = y) on the CIG85 projection.

1. Major roads were not considered - forest is often drawn as continuous across roads.

2. The boundaries of clearings are inaccurate, sometimes up to one or two hundred meters.

3. Some small, isolated forest blocks have been omitted.

In addition, the accuracy of the forest/clearing boundaries of the newveg.shp will depend on how closely Dr. Green zoomed in on Christmas.ers. Dr. Green adjusted most boundaries when working at a scale of 1;1000 to 1:1500, but occasionally had to work at greater scales than this. In some areas (e.g. Field 25, and the area around the sports ground) the 1987 aerial photos are of particularly poor quality (shadows, distortion etc) and it was sometimes difficult to distinguish boundaries.