Area-preserving generalisation of the GSHHS data

Global Self-consistent, Hierarchical, High-resolution Shoreline data (GSHHS) is a shoreline dataset amalgamated from two databases in the public domain. GSHHS is developed and maintained by Dr. Paul Wessel, SOEST, University of Hawai'i and Dr. Walter H. F. Smith, NOAA Laboratory for Satelitte Altimetry.

GSHHS is distributed as original dataset (full resolution) along with different levels of details (high, intermediate, low and crude). The generalisation was performed by Douglas-Peucker algorithm adapted for sphere.

Generalised versions include known disadvanteges of Douglas-Peucker algorithm. These are visible angularity in polylines and collapsing of polygons into the lines (see figure below).

The data on this site was generalised by our algorithm with area-preservation property (algorithm is implemented as GRASS GIS module, more information on module on v.cartographic.line.generalization).

Global datasets aimed for distribution as geographical coordinates should be generalised by taking into account deformations to ensure same degree of generalisation over the whole dataset.

The generalisation was performed in sinusoidal equalarea map projection (Sanson). Lengths were calculated as geodesics on spheroid. Differences of angles were calculated in plane. Angle errors in map projection have limited imapct on angle differences over small area which is taken into account during local modifications of polylines. Generalised lines were repojected back into geographic coordinates.

Even though the algorithm preserves areas in plane, mapping of polylines between spheroid and plane involves some changes of areas.

The generalisation is performed for number of scales and the parameters were adjusted for printed maps in that scale with lot of details as it is found on topographical maps or charts.

IMPORTANT: No topological or geometrical pre- or post processing was performed in this first version of data. Therefore data contains self-intersections, ovelapping and topological inconsistences as a result of automatic generalisation process.

Download

Data is available as compressed shapefiles.

The data is generalised for scales:

Web Processing Service

The generalisation algorithm used is available as OGC Web Processing Service (URL: http://wps.kartografija.hr/pywps/cgi-bin/pywps.cgi).

WPS uses GML as input and output data format. Requests are limited to 50 MB.

One can use it to perform generalisation of its own data for arbitrary map scales. The description of usage can be found on v.cartographic.line.generalization).

Examples

Phillipines for
paper map at
the scale of
1:100 000 000

Phillipines for paper map at
the scale of 1:50 000 000

Phillipines for paper map at the scale of 1:20 000 000

View more examples ...

If interested in the source code of GRASS module v.cartographic.line.generalization contact the authors.

AUTHORS

Dražen Tutić, Miljenko Lapaine and Martina Triplat Horvat
University of Zagreb, Faculty of Geodesy
Contact e-mail: dtutic@geof.hr

Take a look at the poster presentation on 25th International Cartographic Conference, July 3-8, 2011, Paris.

REFERENCES

Tutić, D., Lapaine, M. (2009): Area Preserving Cartographic Line Generalization, Cartography and Geoinformation, Vol. 8, No. 11, 84-100, http://www.kartografija.hr/kig/upload/clanci/kig11_Tutic2.pdf.

Tutić, D., Lapaine, M. (2010): New Method for Reducing Sharp Corners in Cartographic Lines with Area Preservation Property. Proceedings of the 14th International Conference on Geometry and Graphics, Kyoto, 289-290, http://bib.irb.hr/datoteka/480211.Tutic_Lapaine_full_paper.pdf.

Last changed: 2011-07-05