small maps for figures
This library makes it easy to generate small data-driven raster maps. It's a complement to our @carbonplan/maps
library which is for making interactive tiled maps that can be panned and zoomed. The use case here is figures and other static graphics where we don't need interactivity, where we can easily load the entire raster dataset we want to render, and where we often want to show the full globe.
Not relying on tiles frees us up to more easily use custom projections, which we have implemented here performantly via WebGL. We're using a separate package containing the GLSL shader math for inverse geometric map projections.
Create a Minimap
component and include the graphical elements you want. Most maps contain a Raster
(with the data) and a Path
(typically coastlines or similar). In addition, you might want to render Graticule
lines or the Sphere
that bounds the globe.
The source for Raster
can be an image (e.g. png
) or a zarr array or group. It must contain a complete data grid in the equirectangular projection (spanning -180,180 lon and -90,90 lat).
Here's a simple example where we use a json
file for the Path
and the "Blue Marble" image in png
format for the Raster
. We specify that the format is rgb
because each pixel of the image has three color values that we want to render as rgb.
import { Minimap, Raster, Path, Sphere, Graticule } from '@carbonplan/minimaps'
import { naturalEarth1 } from '@carbonplan/minimaps/projections'
return (
<Minimap projection={naturalEarth1}>
<Path
stroke={'white'}
source={'https://cdn.jsdelivr.net/npm/world-atlas@2/land-50m.json'}
feature={'land'}
/>
<Graticule stroke={'white'} />
<Sphere fill={'black'} />
<Raster
source={
'https://upload.wikimedia.org/wikipedia/commons/thumb/2/23/Blue_Marble_2002.png/2880px-Blue_Marble_2002.png'
}
format={'rgb'}
transpose
/>
</Minimap>
)
Here's a slightly more complicated example where we use a zarr
file for the Raster
. The data here are one-dimensional, so we specify a colormap
and clim
and the lut
mode.
import { Minimap, Raster, Path, Sphere, Graticule } from '@carbonplan/minimaps'
import { naturalEarth1 } from '@carbonplan/minimaps/projections'
import { useColormap } from '@carbonplan/colormaps'
const colormap = useColormap('fire')
return (
<Minimap projection={naturalEarth1}>
<Path
stroke={'white'}
source={'https://cdn.jsdelivr.net/npm/world-atlas@2/land-50m.json'}
feature={'land'}
/>
<Graticule stroke={'white'} />
<Sphere fill={'black'} />
<Raster
clim={[0, 50000000]}
mode='lut'
nullValue={9.969209968386869e36}
source={
'https://carbonplan-climatetrace.s3.us-west-2.amazonaws.com/v0.4/blog/total_emissions.zarr'
}
variable={'emissions'}
colormap={colormap}
/>
</Minimap>
)
You can specify the source
as either the URL of an image resource (e.g. .png
or .jpeg
) or a Zarr store (.zarr
). In the case of using a Zarr store, you can either point to an array, or to a group an additionally specify the variable
.
For either type of raster data source, we can use any of the map projections in @carbonplan/minimaps/projections
. Here's how the result would look for the blue marble image.
# naturalEarth1
# orthographic
# equirectangular
# mercator
All the original code in this repository is MIT-licensed. We request that you please provide attribution if reusing any of our digital content (graphics, logo, copy, etc.).
CarbonPlan is a nonprofit organization that uses data and science for climate action. We aim to improve the transparency and scientific integrity of climate solutions with open data and tools. Find out more at carbonplan.org or get in touch by opening an issue or sending us an email.