This is my contribution to the #30DayMapChallenge for November 2023.
About the #30DayMapChallenge , check out https://30daymapchallenge.com/ !
I am a postgraduate student studying GIS and data science at UCL, trying to learn R, python, and making use of GitHub.
Background in planning, new to London from Japan, some knowledge about QGIS, and interested in graphic design (trying to make the maps as aesthetic as possible!)
- where I will be uploading my maps!
- where I will be uploading my code as well
- my first public repository on GitHub
- my first project using R and python
The data itself will not be uploaded due to limitations in file size, but I will try to be as reproducible as possible (but this might happen later.)
Through the #30DayMapChallenge I want to:
- be more familiar with R and python to draw maps using code. I will try to stay away from QGIS and Illustrator, although it is tempting to use them...
- also be familiar with other software that I don't know. Blender might be one of them.
- know more about London, the UK and Europe. The datasets will probably feature London and surrounding areas a lot. Some data might come from Japan.
- know where to get data. Any recommendations are appreciated!
The maps I made for this year's #30DayMapChallenge !!!
- 30 new works of my own!
- Mapping skills using R (especially using
terra,sf,tidyverseandggplot2, andrayshader) - New connections with the mapping community!
- Time. It took time.
- Wanted to explore other tools! Especially 3D stuff like Blender or Unity.
- Should have gone more into interactive works
Anyway, I survived! Thank you everyone for the tremendous support, engagement, and sharing! Not sure whether I will do this again, but gained so much from it!
I made it to the final day. My favorite.
I think my favorite for this is the dataset I started with: Santander Cycles.
Apparently the docks are separated into zones, but which zone are the cycles most used? Which combination? Are they complete within zones, or are they traveling long distances?
This is the map I got. People ride around Hyde Park, Queen Elizabeth Olympic Park, and Marylbone. This was a more obvious pattern than I thought there would be!
I did the data wrangling in Python, and I have uploaded the code for that as well!
Anyway, this is the end of my challenge, and hope you enjoyed!
I always wanted to make the spike map I see in many places, and have seen throughout the challenge.
Now is probably the last chance to make it, and here it is: a spike map of UK and Ireland.
Although I didn't use the render_highquality() function, it still is pleasing!
London might be a massive outlier, but it is definitely not the only place in this country where it's densily inhabited!
One more to go!
Spreading the boundaries of a map for this one. The Cambridge Dictionary defines a map as "a drawing of the earth's surface" showing natural or artificial stuff.
How about taking the cross-section of the surface? I plotted X-Z plane, plotting longitude on one axis and the elevation on the other. The data is purely geographic, but doesn't necessarily draw the earth's surface.
Is this a chart or a map?
We had points in day 1, and now dots. I considered dots as something that do not represent particular points, where points the location itself represents the existence of something.
I represented London using dots today, changing the size according to the density of Points of Interests in that area.
The dot pattern reminded me of the halftone dots in printing, so I tweaked the overall look of the map into a retro color scheme.
For minimal I went for the stereotype minimal: lines only, geometric shape, geometric font.
Map of London centered around Tottenham Court Road Station, very close to UCL. Very straightforward and minimal map!
Antarctica: the windiest, coldest, and driest continent on Earth. This extreme weather has made even reaching a particular point quite an achievement!
I have mapped the monuments that commemorate these achievements that are located in this continent. They do have data for everything.
The most remarkable is the one on the geographic south pole, but there's quite a lot, including tents from expeditions, tractors, abandoned stations, and even cemeteries.
Let's visit Japan for the second day in a row in this challenge.
Now for black and white, I wanted to answer the simple question: where in Japan is it "black" or "white"? This map could be one answer: the places where the names include black or white! To be honest I was expecting a bit more for both; I might have needed to include more places, like names of rivers, mountains, and other natural terrain.
No clear trend for this one, but there might be some interesting trends for some other colors!
I wanted to map something worth mapping in 3D. This map shows the zoning restrictions in Osaka, Japan in terms of density.
The Japanese zoning system restrict the volume allowed in each area, limiting the ratio of the floor area to the land area. (i.e. A 200 % limit means a building is allowed a floor area twice the land area.) An interesting characteristic is that high-density designations are located along main avenues, so in this map the grid pattern of Osaka can be visible, although this is merely mapping the density restrictions!
Data fetched from the Ministry of Land, Infrastructure, Transport and Tourism, whom opened a new website where planning data can be obtained as shapefiles.
In order to make a flat map from a spherical Earth, we have to distort some of the characteristics in one way or another. In today's map I have distorted it in a way that we usually do not see: transforming the whole world into a British National Grid!
I thought about making this sort of map for the "Bad Map", but I felt there is no significant difference between this map and the Mercator projection that we usually see. The BNG is a transverse Mercator projection with 2°West as the central meridian instead of having the equator the origin in an "usual" Mercator scale. Hey, what difference does that make?
I wanted to explore new datasets today, and found an interesting dataset by the JAXA Global ALOS project of Japan, where it determines the forest/non-forest-ness of the world.
I needed the low-res version of their data to map the whole world, and found out that the latest version they had was from 2010. Using high-res data might be better for detailed analysis, but that will not happen today.
I have re-used the template from day 18, showing that a lot of the area in the world is completely non-forest.
Today I visited the data for the city of Bristol: data was collected for part of my assignment, although never visited the city in person.
I have mapped the height data for Bristol, and thought about the topic of "outdoors" and decided to un-map where the buildings are, leaving the outdoors to be mapped.
Ordinance Survey has a detailed map on heights as well, which is what I decided to map! I was not familiar with the .asc file type, but worked out just fine.
London road and rail networks, filtered the main ones, plotted on QGIS, period. It took 6 minutes and 24 seconds. I took too much time tweaking the aesthetics.
I'm not going to spend more time explaining the map than making it, but I am truly surprised you can make a decent map in this limited amount of time!
Atmosphere: I mapped the temperature range showing the annual temperature difference between the maximum and minimum.
Today's significance for me is creating a "legendgram", although this is just an inset density plot using the same dataset and coloring. Does this help understand the whole picture better?
For flows, I came back to the dataset that I used for Day 1 of the challenge: Santander Cycles. I really do think I've come very far! One of my favorite visualisations is the map with airline routes, and I was thinking I could do something similar! I am so happy I managed to do so.
In addition to the point data that I mapped in day 1, TfL has loads of data for their bike share scheme, including journey details for every journey. I have mapped some 10,000 trips made by the Santander Cycles in their latest dataset available, which accounts for roughly 1/3 of a day.
I did some basic analysis to their data today. Their journey data did not include any spatial data, only the name and the ID of their start and end points. I joined this with the spatial data with the points to draw the line, and calculated the speed for each trip. I added a Scripts_Data folder to save the Jupiter Notebook I have worked on for today's (and Day 1's) data wrangling. I stopped here today, but am thinking about further analysis on this to see if any interesting data pops out!
The first day of the second half: Oceania! This was a difficult area to map.
I ended up with a simple map of the annual average temperature in Australia, hence I couldn't quite manage to do the following:
- Precisely extract all the countries and territories that are recognized as "Oceania"
- Find a way to map areas too small if you just mapped the rasters
- Find a nice way to center around the International Date Line, not the Prime Meridian
I couldn't manage to change the font for the labels, which is very annoying.
Still, it's interesting to find how cold Australia actually is. My northern hemisphere self refuses to recognize that a country in the south is much colder than where I come from.
Sources: WorldClim, ArcHub
Halfway through the #30DayMapChallenge !
OpenStreetMap has a lot of points of interests in their data, and I found one that interested me: surveillance cameras. (I personally won't define this as a point of interes, though.) Today, I used the DBSCAN analysis to find out whether there are any notable clusters that appear in the London area.
- Central London: Well, everything is clustered here. No surprise.
- Around Heathrow and the RAF: this makes sense to me as well.
- Wembley: Is this because of the stadium?
- Wimbledon: Is this too? Although the tennis courts seem to be apart from the station.
- Twickenham, Brentford, Wandsworth, and some other neighborhoods in the south: Is there a particular reason?
Please let me know if you have any thoughts!
Europe. I probably should have mapped something more European, but the fire alarm that rang last Sunday night was so irritating I decided to search how often this kind of event occurs.
Found a data on London Data Store on fires by borough, mapping it, voila!
......would have been too easy.
I wanted to try the 3D rendering made available by the Rayshader package.
Today's map is my first 3D version, featuring my home borough of Camden successfully hiding itself behind the fire-filled City of Westminster. Okay, at least it's good to know I don't need to expect too many weekend evenings disturbed!
Choropleths are the most basic of maps, and now they're starting to get bivariate.
How about a trivariate map?
I mapped each country's GDP, annual precipitation and the occurances of hydrological disasters per area. (The latter two might have made a better map, but I wanted to give it a shot.) It was too difficult to code the whole thing; data-wrangling in R and edited on QGIS.
The data for the disasters are acquired from EM-DAT, an international database for disasters worldwide. They have data for all sorts of disasters, and I decided to filter with hydrological ones, which I expected a correlation with precipitation. I have counted on a country-basis and standardised it by the area.
From the map, I was suprised you can actually sort of see the green areas with high precipitation, the yellow countries with high GDPs, the purple (relatively small) countries having a lot of disasters. The correlations are difficult to make out, but comparing Europe and Asia, there seems to be a different trend already.
Source: World Bank, EM-DAT.
South America is the continent that observes the highest annual precipitation. This is a map of where the precipitation occurs in this area.
I was quite surprised to see two things from the map I made:
- The continent does rain a lot. Even the light-colored areas see around 2,000 mm of precipitation per year, which is almost triple of that of the UK. (Japan has an average of 1,700 mm which is still a lot, but the continent as a whole having this much is impressive.)
- The extremes are really extreme. Although I was prepared when Wikipedia told me Colombia has an average of 3,200 mm, I did not expect over 7,000 mm in certain areas. I also expected the highest point will be somewhere in the Amazon!
It is Day 12 already, and noticed I was not using raster data at all! For me this is a practice making use of raster data as well. I'm unsure whether tidyterra is the way to go, but at least it worked for me.
Source: WorldClim
For retro, I remade the map I made in Day 2: with aesthetics in retro style.
That is to say, I just changed the colors and the font from what I did earlier.
Which, still makes a big difference!
Doing things completely in code makes these adjustments so easy; I don't want to imagine if I had a more complex map to change the aesthetics...
Sources: TfL Unified API, London Datastore
When I was in elementary school, I used to live in Westchester County just north of New York City. Back then I didn't know about this field of GIS or geography (well I did spend some time playing SimCity, if that counts), so I decided to revisit New York with a geospatial lens!
The grid in Manhattan is famous for its one-way throughfare, and I mapped the rare exceptions that did carry both ways of traffic. (Yes I did not drive back then as well.)
Seems like NYC has a decent set of open geospatial data to explore...
Sources: NYC Open Data
Analysis using hexagons is totally new to me. When I was doing my undergraduate in Japan, the only grids we used were the square ones.
As explained here there seems to be advantages, so I decided to give a shot.
I mapped the number of AirBNBs in each hexagonal grid of 350 meters across. Did some basic spatial joins, but did not expect over 200 listings in a single cell!
Just having a dark background makes the maps a lot cooler, but my interest toward graphic design makes me want to make cool maps using a lighter theme....
Sources: London Datastore, AirBNB
I have never engaged with African Geospatial Data before, so I tried mapping the population of major African cities.
ArcGIS Hub does have a lot of data, which is appreciated a lot!
Now that I have a full week of mapping code ready, it is starting to get easier to create maps using the data I already have. Now this is reproducibility!
Sources: ArcGIS Hub
Wayfinding was always my interest, and this seemed like a perfect time to try!
I have created my version of the Legible London signs with a focus on cycle maps!
I have made this map with QGIS for the first time in this challenge. (No codes today!)
After a week of tackling data with R, I now really appreciate the GUI of QGIS.
(Even simple things like rotating views was hard to figure out in R.)
I also tried a new basemap: Carto tiles!
Sources: TfL, CartoDB
When I think about Asia, what comes to my mind are the huge airports competing against each other for size, efficiency, and the status as the world hub.
I have mapped the 50 largest airport by annual passengers, with labeling the top 5.
Dealing with text using ggplot2 is still difficult for me.
I sourced Wikipedia since no others came handy, but felt like I was betraying all my past teachers! Sorry prof!
Sources: Wikipedia, World Bank
For an analog map, I made a map of London - in a way I understand it!
Inspired by "The Image of the City" I tried to draw what I know about the city, and how each component affects how I perceive it.
Few things I noticed:
- Inside or outside of the Ring Road is a big deal to me, crossing it is another big deal (every day I do!)
- The roads and rail do help me a lot when it comes to recognition
- Cycling around the city did help me a lot!
- I know too little about the east
- I know too little about the south
I hope I get to know this city a little better in the near future!
Source: Me!
Okay, a bad map.
I decided to make a map with the wrong information: fake must-see spots within London!
I asked ChatGPt to make up some spots to recommend to an innocent tourist!
It must believe that steampunk is something really big in London.
ChatGPT has made up a description for me as well. Everything is so WRONG!
Enjoy your time in imaginary London!
Sources: ChatGPT, OpenStreetMap
Keep in mind, this is imaginary!
| Name | Description |
|---|---|
| Chrono Square | A mysterious square near the River Thames, where time seems to stand still, and every clock displays a different era. |
| Luminous Alley | A narrow, cobblestone alley in Soho that glows softly at night, rumored to be haunted by friendly ghosts that guide lost travelers. |
| Mystic Bazaar | A bustling market in Camden, where travelers can find exotic artifacts, magical trinkets, and fortune tellers predicting the future. |
| Steampunk Square | A square in Greenwich adorned with intricate steampunk sculptures and gadgets, reminiscent of a bygone industrial era. |
| Moonlit Masquerade | An elegant ballroom in Westminster, hosting enchanting masquerade parties, live orchestras, and dance performances under the shimmering moonlight. |
| Theater of Legends | A historic theater in Camden, where legendary performances from Shakespearean plays to modern dramas captivate audiences, preserving the city's theatrical heritage. |
| Steampunk Emporium | A steampunk-themed marketplace in Westminster, showcasing intricate gadgets, retro-futuristic fashion, and unique artifacts from the Victorian era. |
| The Whimsical Workshop | An imaginative workshop in Soho, where artists, inventors, and dreamers collaborate to create fantastical sculptures, whimsical gadgets, and interactive installations. |
| Name | Description |
|---|---|
| Ethereal Gardens | A magical garden hidden behind the façade of an old bookstore in the heart of Covent Garden, where rare and mythical plants flourish. |
| Aquaville | An underwater city beneath the Thames, accessible only through secret tunnels, where merfolk and humans coexist in harmony. |
| Whispering Willow Park | A secluded park in Kensington where the wind whispers secrets to those who sit beneath the ancient willow tree. |
| Verdant Vale Gardens | A lush garden oasis in Wandsworth, where rare plants, cascading waterfalls, and colorful butterflies create a picturesque natural retreat. |
| Elysian Park | A serene park in Lambeth, known for its blooming flowers, tranquil ponds, and a statue of a mythical goddess, creating an idyllic atmosphere. |
| Glowing Lotus Garden | A serene garden in Lambeth illuminated by radiant lotus flowers at night, offering a meditative atmosphere for quiet reflection and contemplation. |
| Tranquil Treetop Retreat | A secluded treetop sanctuary in Greenwich, featuring cozy treehouse cabins, hammocks, and meditation areas, offering a peaceful escape amidst nature. |
| Enchanted Forest | A magical forest in Southwark, home to mythical creatures, sparkling waterfalls, and ancient ruins, inviting visitors on a whimsical adventure. |
| Royal Rose Gardens | A regal garden in Kensington, adorned with blooming roses of various colors, where visitors can indulge in afternoon tea and enjoy classical music concerts. |
| Name | Description |
|---|---|
| Time Traveler's Tavern | A historic tavern in Islington, where time travelers from different eras gather, sharing stories over ale and traditional British fare. |
| Historical Harbor | A historic harbor in Tower Hamlets, once bustling with trade ships and pirate legends, now transformed into a museum preserving maritime history. |
| Theater of Legends | A historic theater in Camden, where legendary performances from Shakespearean plays to modern dramas captivate audiences, preserving the city's theatrical heritage. |
Polygons lets you do everything, so I chose this theme because wanted the geometry of the polygon to speak for itself.
The Green Belts in England is definitely something that characterizes its urban structure, and I mapped it! I didn't know there are so many cities with green belts around them.
I also tried out the ggplot2 package to map today. I found it difficult to show texts the way I want it to, but maybe practice makes better!
Sources: data.gov.uk, gadm.org
The map for lines - the routes for the iconic double-decker (and single-decker) bus routes in London!
The shaded area show the area of the 33 London Boroughs.
The vast network covers most of London, and some routes go beyond to Slough and Dorking as well.
Credits for Transport for London again!
I wanted to make a route diagram of the whole bus network (something like the tube network map) with all the routes drawn explicitly, but couldn't find a way to deal with the overlaps. Still illustrates the complexity of the network!
Sources: TfL Unified API, London Datastore
For the very first day of my very first #30DayMapChallenge, I have mapped the docking stations for the Santander Cycle, a bike-sharing scheme in London.
The points are colored according to the area name within the docking station name.
Sources: TfL Unified API, OpenStreetMap