grd2kml(cmd0::String="", arg1=nothing, kwargs...)

keywords: GMT, Julia, grid KML

Create KML image quadtree from single grid


grd2kml reads a 2-D grid and makes a quadtree of PNG or JPG images and KML wrappers for Google Earth using the selected tile size. We downsample the grid depending on the viewing level in the quadtree using a Gaussian filter, but other filters can be selected as well. Optionally, illumination may be added by providing a grid file with intensities in the (-1,+1) range or by giving instructions to derive intensities from the input data grid automatically (see shading). Values outside the (-1,+1) intensity range will be clipped. Map colors are specified via a color palette lookup table. Contour overlays are optional. If plain tiles are selected (i.e., no contours specified) then the PNG tiles are written directly from grdimage. Otherwise, we must first make a PostScript plot that is then converted to raster image via psconvert.

Required Arguments

The 2-D gridded data set.

  • N or prefix : – prefix=name
    Sets a unique name prefixed used for the top-level KML filename and the directory where all referenced KML files and raster images will be written [GMT_Quadtree].

Optional Arguments

  • A or mode : – mode=:a|:g|s | mode="a|g|saltitude"
    Select one of three altitude modes recognized by Google Earth that determines the altitude (in m) of the tile layer: a absolute altitude, g altitude relative to sea surface or ground, s altitude relative to seafloor or ground. To plot the tiles at a fixed altitude, append an altitude altitude (in m). Use 0 to clamp the features to the chosen reference surface. [By default the tiles are clamped to the sea surface or ground].

  • C or color or cmap or colorap or colorscale : – color=cpt
    Where cpt is a GMTcpt type or a cpt file name (for grd_z only). Alternatively, supply the name of a GMT color master dynamic CPT [turbo] to automatically determine a continuous CPT from the grid's z-range; you may round up/down the z-range by adding +i zinc. Yet another option is to specify color="color1,color2 [,color3 ,...]" or color=((r1,g1,b1),(r2,g2,b2),...) to build a linear continuous CPT from those colors automatically. In this case color1 etc can be a (r,g,b) triplet, a color name, or an HTML hexadecimal color (e.g. #aabbcc ) (see Setting color). When not explicitly set, but a color map is needed, we will either use the current color map, if available (set by a previous call to makecpt), or the default turbo color map.

  • E or url : – url=URL
    Instead of hosting all files on your computer, you may prepend a remote site URL. Then, the top-level prefix.kml file will use this URL to find all other files it references. After building completes you must place the entire prefix directory at the remote location pointed to by the URL [local files only]. With this arrangement you can share the prefix.kml with others (say, via email or for download) and users can open the file in their Google Earth and access the remote files from your server as needed.

  • F or filter : – filter=type
    Specifies the filter to use for the downsampling of the grid for more distant viewing. Choose among boxcar, cosine arch, gaussian, or median [Gaussian]. The filter width is set automatically depending on the level.

  • H or subpixel : – subpixel=scale
    Improve the quality of rasterization by passing the sub-pixel smoothing scale to psconvert (same as H option in psconvert) [no sub-pixel smoothing]. Ignored when |-W| is not used.

  • I or shade or shading or intensity : – shade=grid | shade=azim | shade=(azimuth=az, norm=params, auto=true)
    Gives the name of a grid with intensities in the (-1,+1) range, or a constant intensity to apply everywhere (affects the ambient light). Alternatively, derive an intensity grid from the input data grid grd_z via a call to grdgradient; use shade=az or shade=(azimuth=az, norm=params) to specify azimuth and intensity arguments for that module or just give shade=true to select the default arguments (azim=-45,nom=:t1). If you want a more specific intensity scenario then run grdgradient separately first.

  • L or tilesize : – tilesize=size
    Sets the fixed size of the image building blocks. Must be an integer that is radix 2. Typical values are 256 or 512 [256]. Note: For global grids (here meaning 360-degree longitude range), we will select a tilesize of 360 if tilesize is not specified.

  • S or extralayers or extra_layers : – extralayers=true | extralayers=extra
    Add extra layers beyond that necessary to capture the full resolution of the data [none]. This will let GMT interpolate your grid and make more tiles, versus letting Google Earth interpolate the last resolution raster images.

  • T or title : –title=XXX
    Sets the title of the top-level document (i.e., its description).

  • V or verbose : – verbose=true | verbose=level
    Select verbosity level. More at verbose

  • W or contours : – contours="cfile|pen[+sscale/limit]"
    Supply a file with records each holding a contour value and a contour pen. We then overlay the selected contour lines on top of the image [no contours]. Consequently, contours triggers the tile creation via PostScript and thus is slower. If cfile is not a valid file we assume you instead gave a pen and want to draw all the contours implied by the cpt specified in cmap. The contours are overlain via calls to grdcontour. Note: The contour pen width(s) refer to the highest tile level and are reduced by a factor of scale [sqrt(2)] for each lower level. Contours with scaled pen widths < limit [0.1] points are skipped (except for pen widths that exactly equal 0 or "faint"). Use +s to change these values.

  • f or colinfo : – colinfo=??
    Specify the data types of input and/or output columns (time or geographical data). More at

  • n or interp or interpol : – interp=params
    Select interpolation mode for grids. More at interp

Quadtree building

We extend the input grid vi grdcut to obtain a square dimension that can be repeatedly divided by 2 until we arrive at tiles with the original grid increments. For global grids this mean we extend the grid to a 360 x 360 Cartesian region and an initial grid increment of one degree. This is the first global tile. As the quartering of tiles and halving of grid increment continue we may not end exactly at the original grid spacing but at the largest increment less than or equal to the original increment. For non-global grids, e.g., smaller (local or regional) grids, we extend the domain to a radix-2 multiple of the tilesize times the grid increment. This initial tile is then quartered and the grid increment halved until we reach the original grid increment. Tiles that have all NaNs are not produced. THe tiles are inherently pixel-registered. Thus, if a global grid has gridline-registration then we are down-sampling the extended grid onto a pixel-registered coarser grid. Because these nodes do not coincide with the original nodes we widen the filter width by a factor of sqrt(2). We detect if NaNs are present in any tile and if so produce a transparent PNG tile; otherwise we make an opaque JPG tile.

Contour overlays

Because each tile is a fixed size image (e.g., 512x512 pixels) but the amount of data represented changes by factors of 4 for each new level, we cannot use a constant thickness contour pen for all levels. Thus, the pen you supply must be considered the final pen applied to the highest resolution map overlays. Furthermore, because the dpi here is very small compared to regular GMT plots, it is important to improve the appearance of the contours by using sub-pixel smoothing (subpixel). Both generating PostScript tiles and using sub-pixel smoothing adds considerable processing time over plain tiles.


The intensity grid can be created from the data grid using grdgradient and, optionally, modified by grdmath or grdhisteq. Custom intensity grids built with several different illumination angles can be combined with grdmath. For a single illumination angle the automatic illumination can be used instead.


To test a quadtree image representation of the coarse topography grid earthrelief06m, using the optimally determined tile size, auto color, and supplying a suitable title, try::

grd2kml("@earth_relief_06m", prefix="Earth6m", title="Earth Relief 6x6 arc minutes", cmap=:earth)

To make a quadtree image representation of the large topography grid file, supplying automatic shading based on the topography, and using 512x512 tiles, supplying a suitable title, and using color masking for unmapped area, try::

grd2kml("", shade="+d", prefix="ellice", -L512, title="Ellice Basin Bathymetry")

See Also

gmt2kml, grdgradient, grdhisteq, grdimage, grdmath, kml2gmt, psconvert