Contour

1. Overview

MapServer (since version 6.4) can compute and render a contour layer on the fly from a raster source. The raster source is one band of raster data, which represents a digital elevation model (DEM). More info about DEMs at: http://en.wikipedia.org/wiki/Digital_elevation_model

2. How it works

CONNECTIONTYPE CONTOUR. The new (to MapServer 6.4) type is a hybrid layer, which has a raster data source as input and vector features as output. Initially, only the line representation of those vector features will be supported. More detailed information about the original implementation can be found in RFC 85.

Since the internal layer is of type vector, queries will be supported and operate on the vectors (not on the raw raster source). In the future we might see a need to add a query mode that queries the raster source, but this is not included in this phase of work.

To render a contour layer, we need to define a layer in the mapfile with the following options:

  • Set the layer TYPE to LINE.

  • Set CONNECTIONTYPE to CONTOUR.

  • Set the DATA to the raster file that contains the elevation band. Starting with MapServer 7.0.1, a TILEINDEX can also be used, together with WMS Time to specify which raster of the tileindex must be used.

  • Specify the band to use as elevation using PROCESSING «BANDS», same as regular raster.

  • Specify one or more classes and styles to render the line features.

PROCESSING settings:

These options should be specified at layer level:

  • CONTOUR_INTERVAL: elevation interval between contours

  • CONTOUR_LEVELS: comma-separated list of one or more “fixed levels” to extract

  • CONTOUR_ITEM: provides a name for the item (attribute) in which to put the elevation. (optional)

You can also provide explicit min/max scaledenom in the CONTOUR_iNTERVAL or CONTOUR_LEVELS values if you wish to use scale-dependent contour spacing. This is done by adding an optional «miscaledenom,maxscaledenom:» prefix to the value or list of values. See the example below.

Προειδοποίηση

If you only set a single CONTOUR_INTERVAL for all scale ranges, MapServer will likely crash at outer scale ranges (first example below). More recent GDAL versions (such as 3.2.0) will output an error message such as: «Input values and/or interval settings would lead to too many levels». To avoid this, set multiple CONTOUR_INTERVAL lines to handle various scale ranges (see the working Copernicus DEM example below, with output map images).

Example of a simple layer definition:

LAYER
  NAME "my_contour_layer"
  TYPE LINE
  STATUS DEFAULT
  CONNECTIONTYPE CONTOUR
  DATA "/mnt/data/raster/grib/dem.grib"
  PROCESSING "BANDS=1"
  PROCESSING "CONTOUR_ITEM=elevation"
  PROCESSING "CONTOUR_INTERVAL=10"
  CLASS
    STYLE
      WIDTH 2
      COLOR 255 0 0
    END
  END
END #layer

Example of a layer definition with scale-dependent contour ranges:

LAYER
  NAME "my_contour_layer"
  TYPE LINE
  STATUS DEFAULT
  CONNECTIONTYPE CONTOUR
  DATA "/mnt/data/raster/grib/dem.grib"
  PROCESSING "BANDS=1"
  PROCESSING "CONTOUR_ITEM=elevation"
  PROCESSING "CONTOUR_INTERVAL=0,25000:5" # interval of 5 for scales of 25000 or less
  PROCESSING "CONTOUR_INTERVAL=25000,500000:10" # interval of 10 for scales in the 25000 to 500000 range
  PROCESSING "CONTOUR_LEVELS=500000,0:10,25,50,100" # explicit list of levels for scales of 500000 and up
  LABELITEM "elevation"
  CLASS
    STYLE
      WIDTH 2
      COLOR 255 0 0
    END
    LABEL
      ...
    END
  END
END #layer

Example of a layer definition with a tile index and WMS TIME support:

LAYER
  NAME "my_contour_layer"
  TYPE LINE
  STATUS OFF
  CONNECTIONTYPE CONTOUR
  TILEINDEX "data/contour_ti.shp"
  TILEITEM  "location"
  PROCESSING "BANDS=1"
  PROCESSING "CONTOUR_ITEM=elevation"
  PROCESSING "CONTOUR_INTERVAL=20"
  CLASS
    STYLE
      WIDTH 1
      COLOR 255 0 0
    END # STYLE
  END # CLASS

  METADATA
    "DESCRIPTION" "contour"
    "wms_title" "contour"
    "wms_timeitem"  "TIME"
    "wms_timeextent"  "2004-01-01/2004-02-01"
  END # METADATA
END # LAYER

Example of a layer definition with scale-dependent contour ranges & Copernicus EU DEM data:

LAYER
  NAME "dem-contours"
  TYPE LINE
  STATUS ON
  CONNECTIONTYPE CONTOUR
  DATA "eu_dem_v11_E40N20_clip.TIF"
  PROCESSING "BANDS=1"
  PROCESSING "CONTOUR_ITEM=Band_1"
  PROCESSING "CONTOUR_INTERVAL=0,25000:5" # interval of 5 for scales of 25000 or less
  PROCESSING "CONTOUR_INTERVAL=25000,500000:10" # interval of 10 for scales in the 25000 to 500000 range
  PROCESSING "CONTOUR_LEVELS=500000,0:10,25,50,100,200,300,400,500,600,700,800,900,1000" # explicit list of levels for scales of 500000 and up
  CLASS
    NAME "Contours"
    STYLE
      WIDTH 1
      COLOR 139 69 19
    END
  END
END # layer

Figure: contours zoomed-out to coastline (Corsica island) generated from Copernicus EU DEM

../../_images/contour-coastline.png

Figure: contours zoomed-in (Corsica island) generated from Copernicus EU DEM

../../_images/contour.png

2.1 Data cellsize

The data produced by the gdal contour algorithm are generally in high resolution. A lot of point are used to generated contours with precision. You might want to generalize/simplify the line in some cases (ie. Shape Smoothing). The [data_cellsize] attribute binding represents the cellsize of the extend fetched from the raster file. This is different than the map cellsize.

In the following example, I generalize my shape with a tolerance of 25% of the data cellsize to produce smooth contours at all scales:

LAYER
  NAME "MyContourLayer"
  STATUS DEFAULT
  DATA "wind.tif"
  TYPE LINE
  CONNECTIONTYPE CONTOUR
  PROJECTION AUTO END
  PROCESSING "BANDS=1"
  PROCESSING "CONTOUR_ITEM=elevation"
  PROCESSING "CONTOUR_INTERVAL=0,0:1"
  GEOMTRANSFORM (smoothsia(generalize([shape], 0.25*[data_cellsize])))
  CLASS
    EXPRESSION ([elevation] >= 0)
    STYLE
      COLOR 0 0 255
    END # STYLE
  END # CLASS
END # LAYER