|
Dealing with Polygon / Polyline Data This page is superceded by New Polygon Page |
Contents
| General Preparing the Mid/Mif files Running the rendering program Combining with point data |
dbSEABED seeks to preserve the spatial resolution of the original input data, especially where that data is point sampling data. But polygons (also polylines and grids, henceforth all in term "poly coverages") are summaries of sampling data with decreased spatial resolution.
Polygon data is also difficult because, if it was to be accepted wholly as areas: (i) it would cover over point samplings in its area, (ii) it cannot contribute to interpolations of point data, and (iii) overlapping polygons of different class (eg, "sand' and "bryozoal sand", perhaps from separate studies) cannot be reconciled on one map.
To solve these problems, dbSEABED transforms poly coverages to point
representations, area griddings, or outline nodes by generating points
which have the spatial distribution of the polygons/grids/polylines and
their geological attributes. There are a number of choices available to
the user:
(i) render the outside perimeter or the inside area of the object
(ii) use rectilinear or random point gridding,
(iii) with or without a blank buffer just inside the perimeter.
The conversion to points is performed at approximately the same scale
as the spacings between the nodes that define the polygon/polyline or
the
cells of a gridding.
An illustration of the methods is shown below. Once points are
created
from the polygon, polyline and grid, that data can be interpolated with
actual sampling data to produce an integrated map. Notice the
complication
of salients and islands on the polygon; these are fully dealt with in
the
dbSEABED software.
a. Poly coverage elements; b. Conversion to points and combination with other point samplings. Letters represented the generated points: r – points for rock area polygon; s – points for seagrass gridding; w – points representing sandwave crest. X – represent nearby actual sampling sites. |
dbSEABED uses the ASCII MapINFO(R) Mid/Mif format to store original information on polygons, polylines and grids. The format is easily readable and understood, it is widely used and can be easily generated from many GIS including MapINFO(R) and ArcVIEW(R).
The Mid/Mif data files are placed in the subfolder "_Data\_MifData". It is important that they be PC format rather than MAC or UNIX: it affects the line lengths. To transform, import them into MapINFO(R) (<Table/Import/MapInfo Interchange format>), open the resulting table (<File/OpenTable>), then save it over the original Mid/Mif by a re-export (see above).
While some coverages have individually attributed objects, many are of just one seabed type.
The program "db8_render_poly.bas" is run from its shortcut in the Host Folder. It works through all MidMif files pointed to in the data file "us8_POLY.asc" (or "au8_POLY.asc") which is held in the "_Data" subfolder (document version held in "_Data\_Documents") and which in structure is like a normal dbSEABED ASCII Data Resource File.
No latitude/longitude limits are applied at this stage.
The outputs from "db8_render_polys" are placed automatically in the folder "_Data" where they are available for processing by the main dbSEABED program "db8_process". From there the points generated from the poly coverages are represented in the EXT/PRS/CLC and FAC/CMP files alongside actual sampling points. The generated points are named for the poly coverage file they originated from, for example "coffs_rck.mif".
Latitude/Longitude and Water Depth limits can be placed on the points at this stage using the usual facility for all points in "UserFile.txt".
Using the EXT/PRS/CLC and FAC/CMP files, the points may be graphed, mapped and interpolated in the same way as directly sampled points.
Chris Jenkins (Email)
INSTAAR, University of Colorado
5-Feb-2002