Making nice maps for posters with Python 🗺️+🐍

Making nice maps for posters with Python 🗺️+🐍¶

To communicate your results effectively to people 🧑‍🤝‍🧑, you may come to a point where making maps are needed.

These maps could be created for a conference poster, a presentation, or even for a social media 🐦 post!

In this tutorial 🧑‍🏫, we’ll focus on making basic 2D maps, and by the end of this lesson, you should be able to:

Set up basic map elements - basemap, overview map, title and axis annotations 🌐
Plot raster data (images/grids) and choose a Scientific Colour Map 🌈
Plot vector data (points/lines/polygons) with different styles 🗠
Save and export your map into a suitable format for your audience 😎

🎉 Getting started¶

Once you have an idea for what to map, you will need a way to draw it 🖌️.

There are plenty of ways to make maps 🗾, from pen and paper to Photoshop.

We’ll start by loading some of these tools, that help us to process and visualize our data 📊.

import icepyx as ipx  # for downloading and loading ICESat-2 data
import pygmt  # for making geographical maps and figures
import rioxarray  # for performing geospatial operations like reprojection
import xarray as xr  # for working with n-dimensional data

Just to make sure we’re on the same page, let’s check that we’ve got the same versions.

print(f"icepyx version: {ipx.__version__}")
pygmt.show_versions()

icepyx version: 0.6.2
PyGMT information:
  version: v0.6.0
System information:
  python: 3.9.10 | packaged by conda-forge | (main, Feb  1 2022, 21:24:11)  [GCC 9.4.0]
  executable: /usr/share/miniconda3/envs/hackweek/bin/python
  machine: Linux-5.13.0-1021-azure-x86_64-with-glibc2.31
Dependency information:
  numpy: 1.22.3
  pandas: 1.4.1
  xarray: 0.21.1
  netCDF4: 1.5.8
  packaging: 21.3
  ghostscript: 9.54.0
  gmt: 6.3.0
GMT library information:
  binary dir: /usr/share/miniconda3/envs/hackweek/bin
  cores: 2
  grid layout: rows
  library path: /usr/share/miniconda3/envs/hackweek/lib/libgmt.so
  padding: 2
  plugin dir: /usr/share/miniconda3/envs/hackweek/lib/gmt/plugins
  share dir: /usr/share/miniconda3/envs/hackweek/share/gmt
  version: 6.3.0

A note about layers 🍰¶

What do you do when you want to plot several datasets overlapping the same geographical area? 🤔 A general rule of thumb is to have the raster images on the ‘bottom’ 👎🏽, and the vector data plotted on ‘top’ 👍🏽. Think of it like making a fancy birthday cake 🎂, starting with the dense cake flour (raster), and decorating the colourful icing on top!

1️⃣ The raster basemap 🌈¶

Making the first figure! 🎬¶

Colours are easier to visualize than numbers. Let’s begin with just three lines of code 🤹

We’ll use PyGMT’s pygmt.Figure.grdimage to make this plot.

fig = pygmt.Figure()  # start a new instance of a blank Figure canvas
fig.grdimage(grid=ds_iceland["h"], frame=True)  # plot the height variable
fig.show()  # display the map as a jupyter notebook cell output

Already we’re seeing 👀 some rainbow colors and a lot of gray.

Let’s add some axis labels and a title so people know what we’re looking at 😉

Previously we used frame=True to do this automatically, but let’s customize it a bit more!

fig.grdimage(
    grid=ds_iceland["h"],
    frame=[
        'xaf+l"Longitude"',  # x-axis, (a)nnotate, (f)ine ticks, +(l)abel
        'yaf+l"Latitude"',  # y-axis, (a)nnotate, (f)ine ticks, +(l)abel
        '+t"ATL14 ice surface height over Iceland"',  # map title
    ],
)
fig.show()

Now we’ve got some x and y axis labels, and a plot title 🥳

Still, it’s hard to know what the map colors represent, so let’s add ➕ some extra context.

Adding a colorbar 🍫¶

A color scalebar helps us to link the colors on a map with some actual numbers 🔢

Let’s use pygmt.Figure.colorbar to add this to our existing map 🔲

fig.colorbar()  # just plot the default color scalebar on the bottom
fig.show()

Now this isn’t too bad, but we can definitely improve it more!

Here are some ways to further customize the colorbar 📊:

Justify the colorbar position to the Middle Right ➡️
Add a box representing NaN values using +n ◾
Add labels to the colorbar frame to say that this represents Elevation in metres 🇲

🔖 References:

fig.colorbar(position="JMR+n", frame=["x+lElevation", "y+lm"])
fig.show()

Now we’ve got a map that makes more sense 😁

Notice however, that there are two colorbars - our original horizontal 🚥 one and the new vertical 🚦 one.

Recall back to what was said about ‘layers’ 🍰. Every time you call fig.something, you will be ‘drawing’ on top of the existing canvas.

‼️ To start from a blank canvas 📄 again, make a new figure by calling fig = pygmt.Figure()‼️

Choosing a different colormap 🏳️‍🌈¶

Do you have a favourite colourmap❓

When making maps, we need to be mindful 😶‍🌫️ of how we represent data.

Take some time ⏱️ to consider what is the most suitable type of colormap for this case.

What type of colourmap to choose?

Done? Now let’s use pygmt.makecpt to change our map’s color!!

🔖 References:

Crameri, F., Shephard, G.E. & Heron, P.J. The misuse of colour in science communication. Nat Commun 11, 5444 (2020). https://doi.org/10.1038/s41467-020-19160-7
List of built-in GMT color palette tables: https://docs.generic-mapping-tools.org/6.3/cookbook/cpts.html#id3

fig = pygmt.Figure()  # start a new blank figure!

pygmt.makecpt(
    cmap="fes",  # insert your colormap's name here
    series=[-200, 2500],  # min an max values to stretch the colormap
)

fig.grdimage(
    grid=ds_iceland["h"],  # plot the xarray.Dataset's height variable
    cmap=True,  # setting this as True tells pygmt to use the colormap from makecpt
    frame=True,  # have automatic map frames
)

fig.show()

Once again, we’ll add a colorbar on the right for completeness 🎓

fig.colorbar(position="JMR+n", frame=["x+lElevation", "y+lm"])
fig.show()

(Optional) Advanced basemap customization 😎¶

If you have time, try playing 🛝 with the pygmt.Figure.basemap method to customize your map even more.

Do so by calling fig.basemap(), which has options to do things like:

Adding graticules/gridlines using frame="g" 🌐
Adding a North arrow (compass rose) using rose="jTL+w2c" 🔝
Adding a kilometer scalebar using something like map_scale="jBL+w3k+o1" 📏

🔖 References:

# Code block to play with
fig = pygmt.Figure()  # start a new figure

# Plot grid as a background
fig.grdimage(
    grid=ds_iceland["h"],
    cmap="oleron",
    shading=True,  # hillshading to make it look fancy
)

# Customize your basemap here!!
fig.basemap(
    frame="afg",
    rose="jTL+w2c",
    map_scale="jBL+w3k+o1"
    # Add more options here!!
)

fig.show()  # show the map

3️⃣ Saving the map 💾¶

Now put it all together, like mixing the dry and wet ingredients of a cake 🍰

We’ll start with the raster basemap 🌈, and then plot the vector features 🚏 on top.

fig = pygmt.Figure()  # Create blank new figure

### 1. Raster layers

## 1.1 - Plot the ICESat-2 ATL14 height grid
pygmt.makecpt(
    cmap="fes",  # insert your colormap's name here
    series=[-200, 2500],  # min an max values to stretch the colormap
)
fig.grdimage(
    grid=ds_iceland["h"],
    frame=[
        'xaf+l"Longitude"',  # x-axis, (a)nnotate, (f)ine ticks, +(l)abel
        'yaf+l"Latitude"',  # y-axis, (a)nnotate, (f)ine ticks, +(l)abel
        '+t"ATL14 & ATL11 ice surface height over Iceland"',  # map title
    ],
    cmap=True,  # use colormap from makecpt
    shading=True,  # add hillshading
)

### 1.2 - Add a colorbar
fig.colorbar(position="JMR+n", frame=["x+lElevation", "y+lm"])

## 1.4 - Advanced basemap customization (gridlines, north arrow, scalebar)
fig.basemap(
    frame="af",
    rose="jTL+w2c",
    map_scale="jBL+w3k+o1"
    # Add more options here!!
)

fig.show()

### 2. Vector layers

## 2.1 Coastline
fig.coast(shorelines=True, resolution="h")  # Plot high resolution shoreline

## 2.2 Plot ICESat-2 ATL11 point track
fig.plot(
    x=dataset.longitude,
    y=dataset.latitude,
    color="lightgreen",
    style="c0.1c",  # circle of size 0.1 cm
    label="Track 1358 pt2",  # Label this ICESat-2 track in the legend
)
fig.legend()  # Default legend position is on the top-right corner

## 2.3 Text annotations
# Start off with labelling the capital Reykjavík!
fig.text(x=-23.2, y=64.2, text="Reykjavík", font="16p")
# Add a red square of size 0.2 cm at the capital
fig.plot(x=-21.89541, y=64.13548, style="s0.2c", color="red")
# A-B transect labels
fig.text(x=-20.5, y=65.4, text="A", font="15p,Helvetica-Bold,purple")
fig.text(x=-19.5, y=63.4, text="B", font="15p,Helvetica-Bold,purple")
# Label the oceans
fig.text(
    x=[-19, -18],
    y=[62.8, 66.8],
    text=["Atlantic Ocean", "Arctic Ocean"],
    font="24p,ZapfChancery-MediumItalic,blue",
)
# Label top 3 largest ice caps/glaciers
fig.text(
    x=[-16.5, -21.1, -18.6],
    y=[64.5, 64.8, 65.0],
    text=["Vatnajökull", "Langjökull", "Hofsjökull"],
    font="9p,Times-Italic,blue",
)

## 2.4 Overview map
with fig.inset(position="jBR+w3.5c+o0.2c"):
    # All plotting here in the with-context manager will
    # be in the inset cut-out. This example uses the
    # azimuthal orthogonal projection centered at 10W, 60N.
    fig.coast(
        region="g",
        projection="G-10/60/?",
        land="darkgray",  # land color as darkgray
        water="lightgray",  # water color as lightgray
        dcw="IS+gorange",  # highlight Iceland in orange color
    )

fig.show()

To save ⬇️ the figure, use pygmt.Figure.savefig.

The format 💽 you save it in will depend on where you want to display it.

As a general guide:

Social media 🐦 or Presentations 🧑‍🏫
- PNG or JPG (raster formats)
- Use about 150dpi or 300dpi
Posters 🪧 or Publications 📜
- PDF or EPS (vector formats)
- Use about 300dpi or 600dpi

fig.savefig(fname="iceland_map.png", dpi=300)
fig.savefig(fname="iceland_map.pdf", dpi=600)

That’s all 🎉! For more information on how to customize your map 🗺️, check out:

Tutorials at https://www.pygmt.org/v0.6.0/tutorials/index.html
Gallery examples at https://www.pygmt.org/v0.6.0/gallery/index.html

If you have any questions 🙋, feel free to visit the PyGMT forum at https://forum.generic-mapping-tools.org/c/questions/pygmt-q-a/11.

Submit any ✨ feature requests/bug reports to the GitHub repo at https://github.com/GenericMappingTools/pygmt

Cheers!

Cloud Computing Tutorial

List of Projects

ICESat-2 Hackweek 2022

Making nice maps for posters with Python 🗺️+🐍

Contents

Making nice maps for posters with Python 🗺️+🐍¶

🎉 Getting started¶

A note about layers 🍰¶

0️⃣ The data¶

Download ATL14 Gridded Land Ice Height 🏔️¶

🧊 Load the grid data into an xarray.Dataset¶

1️⃣ The raster basemap 🌈¶

Making the first figure! 🎬¶

Adding a colorbar 🍫¶

Choosing a different colormap 🏳️‍🌈¶

(Optional) Advanced basemap customization 😎¶

2️⃣ The vector features 🚏¶

Coastlines for context ⛱️¶

Overlay ICESat-2 ATL11 point track 🐧¶

Text annotations 💬¶

(Optional) adding an overview map 📍¶

3️⃣ Saving the map 💾¶