The Best Graphics Team in the World

The Straits Times has the best graphics department in the world - probably. The Washington Post and New York Times might be contenders but because of their paywalls most of their work is hidden away from most of the world.

The latest astonishing demonstration of the graphic skills of the Straits Times comes in an article celebrating the 100 year anniversary of the construction of the Johor-Singapore Causeway. I'm not sure what this raised embankment between Johor and Singapore has done to deserve such an incredible historical account of its 100 year existence but it must have a special place in the hearts of the graphics team at the Straits Times.

An incredible interactive 3D model of the causeway drives the engagement in the article Here’s how the Johor-Singapore Causeway evolved over 100 years. I say 3D model but in fact The Straits Times hasn't created just one 3D model of the causeway but two. To show how the causeway has developed over the last 100 years the graphics team have coded one 3D model to illustrate how the causeway looked after its construction in 1924 and another model to visualize how it looks today - one century later.

The article is further enhanced by some incredibly seamless transitions between vintage photographs/videos of the causeway and the 3D models. The speed of these transitions and the fly-overs of the models are truly impressive (in fact I actually had to reduce the speed of the GIFs in this post by half because the original screen captures were so fast).

You can check out more of the incredible work of the graphics team at the Straits Times in:

• Why the Haze has Reached Singapore's Shores Again
• Singapore 2030 - an interactive 3D mapped tour of some of Singapore's development plans 
• Singapore Underwater - a VR simulation of rising sea levels
• The Colourful History of Singapore's Street Names

Rat-Town, Massachusetts

A new interactive map of Boston seems to indicate that Beantown has become Rat-town. Rats! Boston shows rat sightings around the Massachusetts city made by concerned citizens.

Self-appointed Rat Czar Viviano Cantu is using 311 reports to map all the recent sightings of rats around Boston. The sightings are taken from non-emergency 311 calls to the city. The Rats! Boston map also shows the locations of other critters. It doesn't just visualize the locations of rat reports to the city. It also tracks 311 sightings of other animals in Boston. For example on Arlington Street someone seems to have spotted a race between a hare and a tortoise.

It turns out that this isn't an urban re-enactment of Aesop's famous Fable, but is instead two separate reports of a pet turtle spotted in the park and a dead rabbit found on Arlington Street. In fact a lot of the reports to Rats! Boston concern dead animals. On first viewing the map I was intrigued why there were so many reports of pigeon sightings - which is surely not that much of a rare occurrence to necessitate a call to the local authorities. Apparently these are all reports of dead birds that residents want the city to remove.

The law of unintended consequences does suggest that creating a map of 311 reports is not always a good idea. For example Jennifer Wong's map (Human) Wasteland, which tracked 311 calls in San Francisco of 'human waste or urine' quickly became a reference call for loony right-wingers in their irrational hatred of San Francisco. Jennifer's map was created to highlight the issue of homelessness in San Francisco. Unfortunately for many on the right a concern for the plight of the homeless falls a long way below their hatred of liberals. 

It would be a shame if Viviano Cantu's map fell to a similar plight, for example by some idiot re-naming Boston 'Rat-Town, Massachusetts'.

AI-Powered Satellite Search

Clay Explore is an interactive map demo of a new open-source AI Earth observation model. The map allows you to search aerial imagery of Southern California, Seoul and Puerto Rico using machine learning.

Using Clay Explore you can click on any map tile or draw an area to search the map for similar looking areas. Each of the three searchable maps (Southern California, Seoul and Puerto Rico) comes with a relevant example search. In Southern California there is an example of searching for and finding solar farms (as shown in the screenshot above). In Puerto Rico there is an example of looking for tarpaulins and blue roofs (signs of recent hurricane damage) and in Seoul there is an example search of 'green spaces'. However you can also click on any area of the map to carry out your own machine learning search.

The Clay AI is an open-source Earth observation model. It is currently accepting entries for the AI for the Earth Challenge 2024.

The use of artificial intelligence and machine learning is revolutionizing the analysis of satellite and aerial imagery. The ability of machine learning algorithms to automatically detect and identify objects in satellite images has enhanced our abilities to monitor deforestation and illegal fishing, track wildlife, and implement disaster response.

Some other recent examples of machine learning being used to search aerial imagery include OneSoil (which uses AI to detect the types of crops being grown), Земляна проказа (identifying illegal amber mines) and Curio Canopy (identifying tree canopy cover in European cities).

Robin Wilson also recently  released an impressive Aerial Image Search Demo which allows you to explore an AI image search of aerial imagery for yourself. Using the map you can search an image of Southampton for any object that you want. 

The map says that under the hood the AI uses "the SkyCLIP vector embedding model, and Pinecone vector DB". If you are interested in developing your own AI satellite image search there are links to both models. Meanwhile you can have fun searching Southampton for shipping containers, churches, crops, car parks, swimming pools - or anything else that you can think of. 

Via: WeeklyOSM

The US Road Fatality Map

Last week I posted a link to the NYC Congestion Zone Live Crash Tracker, an interactive map of car crashes in New York City. If you live outside of New York then you might prefer Roadway Report instead, which is a visualization of American roadway fatalities in the 21st Century.

The Roadway Report map uses road traffic accident data from the National Highway Traffic Safety Administration's Fatality Analysis Reporting System database, to show road fatalities since 2001 across the whole United States. The map's 'About' page warns however that 'There are four unmapped deaths for every death on the map'. 

The map itself comes with a limited number of options to filter the fatalities shown on the map by date range and by 'driver / passenger' deaths or by 'pedestrian / bike / other deaths'. If you zoom in your neighborhood on the map you quickly get a sense about where the accident black spots in your town are located. You can even click on the map markers to view an incredible number of details about each individual crash, including (but not limited to) the date and time, the number and type of vehicles involved, the number of fatalities, and the weather. 

The extent of the details in the database for Roadway Report means that a much better map could be created using the Roadway Report API. For example you could use the API to create a map which showed local vehicle fatalities by time of day and day of the week, or all road fatalities which involved alcohol and/or drug use. 

Mapping Power Outages in Kiev

The Map of Power Outages in Kiev visualizes power outage schedules in the Ukrainian capital. Due to Russia's ongoing attacks on power stations in Ukraine the electric power company Yasno has to schedule times of planned power outages. The Map of Power Outages in Kiev uses this schedule to provide an interactive map at the individual building level of these scheduled power outages.

It is important to note that this is not a real-time map of actual power outages in Kiev. The map is for informational purposes only, and shows only scheduled power outages, not the possible actual situation at individual addresses. However it is possible to use the day and time controls provided on the map to see when the power is scheduled to be on or off at individual addresses throughout Kiev.

The Map of Power Outages in Kiev was created by Vadym Klymenko. His blog post on the Kyiv Electricity Map provides an interesting account of how he was able to get the individual building address data from OpenStreetMap using Geofabrik. The address data used by the power supplier Yasno comes in a different format than that used by OSM and Vadym also provides an interesting description of the problems that this caused in creating the Map of Power Outages in Kiev.

Your Daily Map Trivia Game

TripGeo Trivia is a new daily geography quiz which requires you to identify ten cities based on a number of clues. Every day ten new random cities from around the world need to be identified. To help you in this task you can view three clues as to the identity of each city.

Every day you get to identify ten new cities. For each city you get three clues and a choice of possible answers. Using the clues you simply need to click on the correct city's marker on an interactive map. There are 30 points in total to be won in TripGeo Trivia. If you guess a city correctly on your first attempt you score 3 points. If you require two clues then you score two points. If it takes you three clues to identify the city then you only score 1 point.

Every day there are 30 points to be won and 10 new cities to be identified on a Google Map. The clues for each city have been generated by ChatGPT. At the end of the game you can explore each of that day's ten cities in more detail. Select a city that you are interested in and you can learn about a number of the city's most popular landmarks and venues, as chosen by ChatGPT, and view their locations on a Google Map.

The Catalan GeoGuessing Game

Developer Toni Vidal has released a new GeoGuessr inspired game featuring photographs of the stunning and diverse landscapes of Catalonia. His Geoendevina game simply requires you to guess the locations of a series of photos taken in the Catalonia region of Spain.

The rules of Geoendevina are very simple. In each round of the game you are presented with a different photograph, each of which depicts a unique location somewhere in Catalonia. From the bustling streets of Barcelona to the serene beaches of Costa Brava, or the majestic peaks of the Pyrenees to the historic landmarks in Girona, each image is a window into the rich tapestry of Catalonia.

Your task is to pinpoint the exact location of each photograph by clicking on the map within 60 seconds. The more accurately you identify the location, the better your score. At the end of each game, you will receive a summary showing the total distance between your guessed locations and the correct positions. The closer your guesses, the better your score.

Toni's game is loosely based on my own Backdrop game, which requires players to identify the locations in famous paintings. Toni has adapted the code so that the game works a lot better on mobile devices and introduced a 60 second time limit for each round.

Tracking American Spies in Germany

Bayerischer Rundfunk and netzpolitik have carried out a joint investigation into how our location data is for sale across the world. These days nearly everyone voluntarily carries around their own personal tracking device in the form of a smartphone. These devices record our movements all day long. 

What most people don't know is that their location data is openly being sold by global data brokers.

Journalist Sebastian Meineck from netzpolitik was given 3.6 billion datapoints, containing location data on around 11 million people, free of charge by a a U.S. data vendor called Datarade. The data was given as just a sample of the data that Datarade was willing to sell to the journalist.

Bayerischer Rundfunk and netzpolitik both carried out their own investigations into the personal information which was contained in the Datarade data. In Under Surveillance Bayerischer Rundfunk reveals how the data includes location data of US spies working in Germany. By exploring the location data of people who spend most of their working hours at a US military base in Germany the broadcaster was able to identify an individual working in the 'Tin Can' - a building used by the US for internet surveillance. Using the location data provided by Datarade it was then just a simple task to determine this US spy's home address in Germany. They are also able to work out other personal details of the individual from the data "including family relationships, preferred supermarket and weekend activities".

This is just one example from the 11 million people's location data provided by Datarade to netzpolitik. According to netzpolitik the "location data comes from mobile phone apps that pass on GPS data for advertising purposes". This data is most often provided by popular apps such as weather, navigation or dating apps.

Both Bayerischer Rundfunk and netzpolitik have written a series of articles on their investigations into the Datatrade location data. In How data brokers sell our location data netzpolitik lists and links to 17 of these articles.

Via: Quantum of Sollazzo

The Retro Gamer's Map

The Retro.Directory is an interactive map which shows the locations of venues related to retro gaming. These include gaming museums, arcades, cafes, bars, clubs and repair services. The map is designed to help retro gaming enthusiasts discover retro-themed locations nearby and around the world.

I am so old that I can actually remember a time before computer games. I can remember the amazing excitement of playing Pong for the first time on a television set (via an Atari games console) and the sudden appearance of arcade games cabinets such as Space Invaders and Galaxians in arcades, bowling alleys and cafes.

I get a visceral thrill every time I stumble upon a retro computer game from my youth. So an interactive map which lists the locations of venues devoted to retro video games ticks a lot of my geek-gamer boxes. What makes the Retro.Directory even more appealing is its deliberate retro computer aesthetic, redolent of an early Yahoo. The sidebar collapsible menus for example allow you to filter the results shown on the map (by events and different types of retro gaming venue).

If there is one thing I enjoy more than playing computer games it is playing pinball. To this day one of my proudest lifetime achievements was clocking the pinball machine in the student union bar at Reading University.

Thanks to Pinball Map it is very easy to track down the locations of nearby pinball machines. Pinball Map has been mapping the locations of pinball machines around the world since 2008. From Soho down to Brighton the Pinball Map must have mapped them all. It currently shows the location of over 43,000 machines. You can even use Pinball Map to find the locations of your favorite pinball machines by name.

Virtual Reality OpenStreetMap

osm4vr is a virtual reality world built using OpenStreetMap map tiles and building footprints. Using osm4vr with a VR headset you can explore the world in virtual reality. Alternatively, if you don't have access to a headset you can simply fly around the world in your browser instead.

Most of the 3d buildings are created using OSM building footprints with building heights, so the graphics can be a little basic. In VR you can flap your arms and fly around the world. In the browser you can use your arrow keys and mouse to navigate around instead. osm4vr also has a search box so you can quickly transport yourself to any location in the world.

You can also explore OpenStreetMap in virtual reality using VR Map. VR Map displays OpenStreetMap data as a virtual reality environment with 3D buildings and trees. You can move around this demo VR Map using your keyboard's arrow keys (or the W,A,S,D keys) and by using your mouse to pan around the scene.

Currently VR Map allows you to explore only a few select locations around the world (including Vienna, San Francisco and New York). However if you fork VR Map on GitHub you can change the initial map coordinates to any location in the world. The virtual reality environment uses an OpenStreetMap tile layer as the ground layer. Buildings tagged in OSM are extruded by their height tags (or to 15 metres if the building doesn't have a height defined). Trees are rendered using the OSM tags for height, circumference and diameter_crown.

osm4vr was inspired by VR Map. However, while VR Map works only in limited areas (thanks to geographical data pre-downloaded from OSM), osm4vr loads the OSM map tiles and buildings dynamically so you can freely move around the whole world.

10 Million Street Views

Street-level imagery such as Google Maps Street View panoramas has become a pivotal resource for many researchers as it can provide a unique perspective on built environments. The ability to access and analyse comprehensive street-level imagery provides researchers with a powerful tool for exploring and understanding urban environments. 

Accessing comprehensive street level imagery at scale can be difficult, expensive and time consuming. Which is why the Urban Analytics Lab at the National University of Singapore (NUS) has introduced the Global Streetscapes project. The NUS Global Streetscapes project provides extensive coverage of urban street level imagery, with 10 million street-level images across 688 cities worldwide, enriched with over 300 attributes. 

The dataset includes images from both Mapillary and KartaView, two crowdsourced 'street view' platforms that offer a diverse range of street-level imagery. Each image in the dataset is annotated with attributes, such as the type of road, weather conditions, and the mode of transportation used to capture the image. These annotations enable researchers to filter and select the images that are the most relevant to their specific studies, for example for evaluating walkability or mapping the levels of urban greenery.

The Global Streetscapes project has even pre-computed some of the evaluations of street view imagery that researchers commonly use, such as the green view index, which ranks the ratio of vegetation pixels in an image to the total number of pixels. 

The NUS Global Streetscapes project is free to use. The project provides open access to a comprehensive dataset of 10 million street-level images enriched with extensive metadata, as well as the code and documentation necessary for using and extending the dataset. All the code and documentation for the project can be found on the Global Streetscape GitHub page and the dataset itself can be accessed on the project's Hugging Face page.

Hat-tip: Map Channels

30 Days of Crashes in New York City

Between June 16th and July 15th, 149 people were injured by cars in the planned New York congestion zone and 4 people were killed.

At the beginning of June New York Governor Kathy Hochul canceled New York City’s planned congestion zone. Under the planned congestion zone vehicles traveling into or within the central business district of Manhattan would have been charged a fee. 

In response to Huchul's cancellation of the congestion zone scheme Transpo Maps has begun mapping all the crashes in the planned congestion zone area. The map plots the locations of vehicle crashes over the last 30 days using data from NYC’s open data portal. Apparently the map is updated as new data is made available on the data portal.

The NYC Congestion Zone Live Crash Tracker shows the locations of crashes where people were injured using orange markers. Pink markers are used to show crashes where there was a fatality. If you click on a census tract on the map you can view data on the number of crashes in the block during 2023, and data on the number of cyclists, pedestrians and car drivers injured and killed in car crashes last year. You can also view data on the number of 'vulnerable residents' and car-free households in the selected census tract.

The 2024 European Election Map

Zeit has created an interactive map which visualizes the results in the 2024 European Union elections in 83,000 municipalities. The map in Explore Europe's Most Detailed Electoral Map colors each electoral area in Europe based on the politics of the leading candidate in the election.

The map allows you to compare the 2024 European Union election results with the results from 2014 and 2019. By switching between the 2014 and 2024 results you can see that there has been a general shift across much of Europe to more right-wing parties. The far-right in particular have made huge gains in France, Austria, the Netherlands, the Czech Republic, Latvia and Germany.

Zeit's accompanying article (in German) explores some of the most interesting results to emerge in June's election. For example the article explores a pattern across much of Europe where inner-city areas tend to vote for more left-wing parties than those in rural areas, who tend to favor parties of the right. 

The newspaper also notes that Poland once again seems politically split along a 100 year old border. In August, 1772, Russia, Prussia, and Austria signed a treaty that partitioned Poland. Poland regained its independence and reunited as the Second Polish Republic in 1918 after World War I. However, the pre-1918 geographical divide seems to re-emerge in the results of every Polish election. Zeit notices that in this European election 'In the western and northern regions that were once German, people vote differently (usually more liberally) than in the former Russian regions.'

The German newspaper also points out the political differences between the Wallonia and Flemish regions in Belgium and the split in Germany along the old West and East German border.

Battles of World War II & American Wars

HistoryMaps has been very busy in the last few weeks, releasing new interactive maps visualizing the:

• Battles of World War II
• Battles of the American Revolution
• Battles of the American Civil War

Nono Umasy's HistoryMaps website is a fantastic resource for anyone interested in world history, offering hundreds of interactive timelines and maps that explore historical events across the scope of human civilization.

The latest three interactive maps on HistoryMaps provide a mapped guide to the battles of World War II, the American Revolution and the American Civil War. On each of these individual maps the locations of important and significant battles are displayed on an interactive map and listed chronologically in the map sidebar.

You can click on the individual battle markers on the map to read the battle's synopsis in the sidebar timeline (each of which includes a link to the battle's Wikipedia entry). The map sidebar also includes a timeline control which allows you to filter the battles shown on the map by year. Other controls allow you to filter the battles by 'Naval', 'Land' and 'Sieges' (and on the World War II map by 'Aerial').

On the cartography side the underlying maps seem to use chronologically accurate borders. For example the country borders for the World War II map includes Czechoslovakia, Yugoslavia, East Prussia and other historical geopolitical entities which no longer exist.

Your Urban Heat Island Score

Climate Central has mapped out the urban heat island hot-spots in 65 major U.S. cities. Each city map on Climate Central's Urban Heat Hot Spots shows an Urban Heat Island (UHI) Index score for each census tract, revealing where UHI boosts temperatures the most and least in each city.

As well as providing individual UHI maps for 65 cities Climate Central has also released a national interactive map which shows how much additional heat communities across the country face due to living in a built environment. The UHI Index score for each city census tract is based on an estimate of how much the urban environment increases temperatures. The estimate is based on local factors such as land cover types (green spaces, paved etc), building height and population density. 

 

Urban heat islands are areas of towns and cities which can become unbearably hot, especially on days with extreme heat. These areas can often become 10-20 degrees warmer than other areas in the very same city. A lack of trees and tree cover in cities is one of the biggest causes of urban heat islands. Urban heat islands tend to occur in areas with the densest built environments and with very little shade provided by tree canopy cover.

The US Tree Equity Score map can help you to quickly determine those neighborhoods in your city that are most in need of increased tree canopy cover. If you click on a neighborhood on the Tree Equity Score map you can discover its 'tree equity score', the current percentage of tree canopy cover, and the local levels of air pollution. The 'tree equity score' uses a range of factors to calculate "how well the benefits of trees are reaching communities living on low incomes and others disproportionately impacted by extreme heat, pollution and other environmental hazards."

Trains, Balloons and Automobiles

I have died and arrived in train spotting heaven.

The Train Positions map combines a live real-time map of Dutch trains with the locations of traffic webcams. The result is that you can track the positions of trains in real-time and actually see them pass locations on live webcams.

Unfortunately a lot of the webcams featured in the Train Positions map are currently offline or have poor views of the passing trains. Nevertheless there are still some gems to be found. My favorite is the webcam of a level crossing at Mierlo-Hout, east of Eindhoven (shown in the animated GIF above). This provides a great view of the train line, with the added bonus that you know when a train is coming because of the level crossing warning lights and chimes (obviously you can also see it arriving on the map).

If train-spotting is not nerdy enough for you then you might prefer the SondeHub Tracker live map, which allows you to follow the paths of weather balloons in real-time. 

A radiosonde is a meteorological instrument used to measure and transmit atmospheric data. The instrument is carried aboard a weather balloon and sends its data to a ground station in real-time by radio signal. The SondeHub Tracker is a live real-time map of active radiosondes (an historical view of past flights is also available), which is used to assist with the tracking, recovery and re-use of meteorological radiosondes.

As well as showing the real-time position of active radiosondes the map also shows the past flight-path of the radiosonde, its predicted forward path, and (where in operation) the real-time position of the chase car.

The D-Day Map Room

The Map Room at Southwick House in Portsmouth was where Allied Supreme Commander General Eisenhower and General Montgomery spent much of early 1944 planning for D-Day. The walls of the Map Room were hung with huge maps of the English Channel. Maps that are still in place in the Map Room at Southwick House to this day.

In particular one wall of the Map Room is covered by a very large map of southern England and the west coast of Europe - including Normandy.  On this map each of the landing beaches (Utah, Omaha, Gold, Juno and Sword), are clearly marked as well as the routes that the Royal Navy ships would take across the English Channel after mine-sweeping ships had successfully cleared safe routes.

Richard Osgood, Senior Archaeologist at the Military of Defence says that the map played a key role in the preparations for the Normandy Landings, “D-Day was a massive event, it changed history. And all of it was meticulously planned on this very map. That’s how important this map truly was, and still is today.”

The National Museum of the Royal Navy has now released an impressive interactive presentation which allows you to explore the map in detail. The D-Day Map - Operation Neptune is not only an interactive online version of the huge D-Day wall map at Southwick House, it is also as part of a chronological guide to the planning, preparation, operation and outcomes of D-Day itself.

The map is part of a guided history of D-Day and is used to explain how the Allies planned and prepared for the massive Operation Neptune. During this guided history of D-Day little model ships and planes are animated on the map to help explain troop movements, landings and the other important events of Operation Neptune.

Via: Webcurios

Your Future Climate Twin

If you want to know how climate change will affect New York City in 60 years time you just need to travel to Arkansas today. In the town of Ola, Arkansas you can experience today the climate that New York is expected to have in 2080, when summer temperatures will be 12.1°F warmer and 5.3% wetter.

The concept of climate analogs is often used in climate science to describe a location whose current climate is similar to the projected future climate of another location. The idea is employed to help people understand the potential impacts of climate change by comparing familiar climates to projected ones. A new interactive web map developed by the University of Maryland Center for Environmental Science enables you to discover the climate analogs of 40,581 locations and 5,323 metro areas worldwide.

Enter your location into Future Urban Climates and you can discover your climate twin. The location today that has a climate which is the most similar to the climate in your town in 2080. The future climate analogs generated by the map are based on "high and reduced emissions scenarios, as well as for several different climate forecast models".

The use of climate analogs has become an established tool to help explain the dangers of global heating. Just last month The Pudding published its Climate Zones map. The Pudding's Climate Zones - How Will Your City Feel in the Future? explains the current climate zones of 70 global cities and the climate zones that they will experience after global heating.  The map allows you to select a city, observe it moving into its future climate zone and learn how average temperatures in the city will be changed by global heating.

You can also explore climate analogs in the National Geographic's Your Climate Changed (an interactive map showing the future climate analogs of 2,500 cities around the world), on the Analog Atlas (2050 climate analogs based on two different climate change predictions), and on the Summer of 2080 Will Be This Warm map (using global heating scenarios of 4.2 degrees or 1.8 degrees centigrade).

Europeans Profiting from American Deaths

Around a third of the guns used in mass shootings in the United States were manufactured in Europe. In European Weapons, American Victims the German newspaper Tagesspiegel investigates how German and Austrian arms manufacturers are profiting from exploiting America's love affair with guns.

Guns are the major cause of children's deaths in the United States, ahead of both traffic accidents and cancer. These deaths are a lucrative business for gun manufacturers. Two of the most successful gun brands in the US are the German SIG Sauer and the Austrian Glock. In an in-depth investigation Taggespiegel explores how both SIG Sauer and Glock are profiting from American deaths.

In an 8 part series the newspaper examines the 'US gun cult and its European profiteers, concentrating on the role of SIG Sauer and Glock. The investigation uses data from the Mother Jones Mass Shootings Database. You can also explore US gun crime on the Mass Killing Database, a collaboration between Northeastern University and the Associated Press. This database keeps track of all multiple homicides in the United States from 2006 with four or more victims.

The Gun Violence Archive also collects data on gun-related violence in the USA (the Mass Killing Database isn't limited to mass killings involving only guns). The Gun Violence Archive reports that there have already been 285 mass shootings in the USA so far this year. To date 126 children have been killed by guns in 2024 and 312 children have been injured by guns.

John Snow's Proximity Mapping

I've probably seen over 100 modern interactive visualizations of John Snow's famous map of cholera victims during the 1854 cholera outbreak in Soho. John Snow's map shows the locations of cholera deaths clustered around a water pump in Broad Street. The map helped to disprove the prevailing miasma theory on the spread of diseases and establish that cholera was actually spread by contaminated water.

Despite the central importance of proximity to John Snow's interpretation of the Soho cholera outbreak I think that Carto's new 1854 Soho cholera outbreak map is the first of the 100+ reinterpretations of the data to actually visualize which water pump the victims lived closest to. 

On the Carto map straight lines are drawn from John Snow's recorded victim's homes to the nearest pump. The purple lines indicate all those victims which lived closest to the Broad Street Pump. As you can see a number of victims actually lived closer to other pumps. John Snow wrote in a letter to the 'Medical Times and Gazette', "In five of these cases the families of the deceased persons informed me that they always sent to the pump in Broad Street, as they preferred the water to that of the pumps which were nearer. In three other cases, the deceased were children who went to school near the pump in Broad Street."

You can view an interactive version of John Snow's original map on the Wellcome website. The map, was originally published in 1855 in Dr. John Snow's 'Report on the cholera outbreak in the Parish of St. James, Westminster, during the autumn of 1854'. 

AI Satellite Search

boats identified by AI

In recent years machine learning and artificial intelligence has revolutionized the way we analyze the world from above. The ability to automatically detect and identify objects in satellite images has unlocked a myriad of possibilities, including monitoring deforestation, tracking wildlife, and enhancing disaster response.

Some recent examples of machine learning being used to search aerial imagery include OneSoil (which uses AI to detect the types of crops being grown), Земляна проказа (identifying illegal amber mines) and Curio Canopy (identifying tree canopy cover in European cities).

Robin Wilson has released an impressive Aerial Image Search Demo which allows you to explore an AI image search of aerial imagery for yourself. Using the map you can search an image of Southampton for any object that you want. For example in the screenshot at the top of this post you can see the results of my search for 'boats'. 

The map says that under the hood it uses "the SkyCLIP vector embedding model, and Pinecone vector DB". If you are interested in developing your own AI satellite image search there are links to both models. Meanwhile you can have fun searching Southampton for shipping containers, churches, crops, car parks, swimming pools - or anything else that you can think of. 

Unfortunately my own searches of the map reveal that Southampton may be harboring a large number of illegal amber mines and nuclear weapon silos.

Via: Weekly OSM

France Defeats the Far-Right

first round (left) and second round (right) election results

The left of center New Popular Front has emerged as the most popular party after France's second round of voting in its national parliamentary election. Le Pen's National Rally party has fallen from its first round triumph to a third placed position, behind both the New Popular Front and President Macron's center-right coalition. 

Comparing Le Monde's second round election map with the national picture after the first round of elections reveals the scale in the drop in support for Le Pen's National Rally party when the voters were given a clear choice between a far-right candidate and a candidate from the center-left or center-right. The brown of the National Rally (in the map's above) is far less prominent after the second and final round of voting, particularly when you consider that the scale of the National Rally's support is skewed on Le Monde's election map by its popularity in the geographically larger rural election areas.

Le Parisien's second round results map includes controls which allow you to filter the results by party. The map of National Rally's winning seats (shown in blue in the map above) reveals that the far-right party swept nearly all of the seats on the Mediterranean coast and that the party also has a lot of regional support in the north-east of the country. 

Interestingly every UK election map I saw after Thursday's election included a hexmap cartogram view and none of the French election maps I've seen have offered anything over than a strict geographical view. Providing an equal representation view of electoral areas, for example by using a hexmap layer is one way of ensuring that every electoral district is given equal visual weight, regardless of its geographic size.

After the success of the National Rally in the French European elections in June Neocarto did create a map which used proportional symbols to try and give more of an equal visual weight to the different sized electoral areas. The comparison above of the geographical view of the results¹ with a proportional symbol map shows how a strictly geographical electoral map can skew the visual picture. The proportional symbol map (on the right) shows in red the areas where the New Popular Front's combined votes could have won in the European elections. 

Not only did the Neocarto proportional symbol map show a more balanced picture it also hinted at how the unification of left-wing parties after the European elections in the New Popular Front (NFP) provided a clear challenge to the supremacy of the RN in many electoral constituencies. It will be interesting to see if Neocarto creates a similar proportional symbol map for France's parliamentary election results.

^{1. these maps don't show the actual European election results but how the results would have looked had the left-wing parties  already joined together into the New Popular Front. The New Popular Front was formed on the 10th June, in response to the electoral success of the National Rally in the European elections on the 9th June.} 

The Car Free Cities Atlas

The cyclists of Helsinki have the best access to protected bike lanes of any citizens in the world. 94% of people in the city live within 300m of a protected bikeway. This means that Helsinki scores number one out of the 1095 cities ranked in the Atlas of Sustainable City Transport.

The new Atlas of Sustainable City Transport ranks and maps cities around the world based on how easy residents can access public transit, protected bike lanes and car-free places. The map allows you to explore the sustainable transport options in over 1,000 cities around the world.

The Atlas provides a global data dashboard that shows how cities are performing in terms of sustainable transportation by tracking nine indicators. These indicators focus on factors such as how close people live to sustainable transportation options like bike lanes and public transit and how many people have access to essential services without needing a car.

Using the map it is possible to select from 1095 cities around the world and explore how each individual city performs under each of the nine sustainable city transport indicators. The map also provides a ranking for each city under each of the nine indicators allowing you to see which cities around the world have the best and worst sustainable transport options.

The residents of Minneapolis have the best access to protected bike lanes in the United States. In Minneapolis 40% of people live within 300m of a protected bikeway. It is a long way short of the 94% in Helsinki but is way better than the situation facing the residents of Bakersfield, Louisville and the five other U.S. cities where 0 citizens live within 300m of a protected bikeway.

24 cities in the United States have no rapid transit infrastructure. The best performing city in terms of access to rapid transit is New York, where 43% of residents 'live within 1km of high-capacity public transport running on a dedicated right-of-way.' Even that doesn't quite live up to Leipzig, where 75% of residents live within 1km of public transport.

2024 UK Election Maps

The Labour Party has swept to power in the UK and the ruling Conservative Party has lost over half of their seats. At the time of writing the new Prime Minister Keir Starmer has 410 seats in the new government (with 11 seats to declare). This means that he has won well over the 326 seats needed to form a majority government.

The BBC election map includes a cartogram view which does a very good job at visualizing the scale of the Conservative losses in this election.

The map shows that the Conservative Party failed to win any seats in Wales, has won only three seats in Scotland. This cartogram view of the election reveals that the Tories have now become a regional party with a core base remaining mainly only in Southern and Eastern England. However even in the South of England the Conservative Party seems to have been all but wiped out in some previous Tory strongholds, such as Berkshire and Oxfordshire.

The BBC map does a very bad job at visualizing the surge in support for the extreme right in this election. The Reform Party won over 4 million votes. However one of the quirks of the UK's first past the post system is that even though Reform received more than a half a million more votes than the Liberal Party they won only 4 seats, compared to the Liberals 70 seats. Consequently while the Reform Party won 14.3% of the UK vote it has little representation in Parliament and little visual impact on the BBC map.

This means that the BBC map does not provide a great picture of where support for the individual parties may truly lie. The Guardian election map does include a Labour Party and Conservative Party 'vote share gains' view. This view (shown above) reveals what a good job the Labour Party did in improving their vote across the whole country, but particularly in Scotland. Conversely the Conservative Party failed to improve their vote share in any seat in the whole of the UK.

The Financial Times map includes a 'Change' view which shows which seats have changed hands in this election. This map reveals how Labour managed to pick up seats from the SNP in Scotland and from the Conservative Party in the North. In the South both the Liberal Party and the Labour Party were able to win a number of seats from the Conservatives. The Conservative Party were able to win just one new seat - in Leicester East, where a couple of independent candidates split the previous Labour vote.

Again, however, the Guardian and FT maps do not include a simple vote share view for each political party. This means that it is therefore impossible to use any of the maps to see where in the country the individual parties won their most and least support. I've read that the Reform Party was the second placed party in 92 constituencies and so far none of the maps I've seen have attempted to visualize this surge in support for the extreme right in this election.

Update: Open Innovation's General Election 2024 map does include an option to view constituencies colored by the second placed party in Thursday's election. This reveals that Reform were in second place in 98 constituencies. 

In his typically ungracious acceptance speech Reform owner Nigel Farage warned that the party was 'coming for Labour'. The Open Innovation map shows that Reform is in second place in many constituencies in the Labour strongholds of the North and Wales. This should be a stark warning to the triumphant Labour Party that a failure to deliver in these regions could prove disastrous for the party and for the UK.

I was actually in the process of creating my own election map (because I wanted to view the second placed party in each constituency) when I stumbled on the Open Innovation map. This means that I now don't have to worry about that (if you want to view the second placed party you can now view the Open Innovation map).

However I will still post a link to my UK Election 2024 Map, as it includes a couple of features I haven't yet seen elsewhere. My map includes political party filters which allow you to select to only view the seats of individual parties (or any combination of parties). 

If you click on a constituency on my map you can also view data on the majority and the vote share of the winning candidate. This allows you to explore the margin by which a candidate won a seat. For example if you click on South Basildon and East Thurrock you can see that Reform won the seat by 98 votes and with a vote share of only 30.79%. This reveals that the seat is extremely marginal and Reform may well struggle to retain this seat at the next election.

The Riskiest Places to Live in America

Los Angeles County is a dangerous place to live. According to the National Risk Index the residents of Los Angeles County are more likely to suffer from a natural hazard than anywhere else in the United States. 

The residents of Los Angeles County are more likely to experience an earthquake than any other US location. They also have one of the highest chances of experiencing a landslide and have a high chance of risk from heatwaves, wildfire, tornadoes and lightning. On the plus side if you live in Los Angeles County you have a very low risk of experiencing extreme winter weather.

You can discover your levels of risk from natural hazards on FEMA's National Risk Index map. The map colors US counties and census tracts to show which are most at risk for 18 natural hazards. If you select a natural hazard risk from the map's drop-down menu you can view the counties which are most and least at risk across the whole country.

For example if you select to view the map of earthquake risk, as you might expect, counties along the San Andreas Fault are shown to be most at risk. The hurricane risk map reveals that counties in Florida have the most risk of experiencing a tropical storm. The risk from avalanche is highest in the mountain range counties of Wyoming.

If you live outside the US then you can discover your degree of earthquake risk on the Global Seismic Risk Map. The Global Earthquake Model Foundation is a non-profit organization working to assess and help manage the risk from earthquakes and seismic activity around the globe. Part of its mission is to assess and share open data on earthquake risks and hazards.

The Global Earthquake Model Foundation has released two interactive maps, the Global Seismic Risk Map and the Global Seismic Hazard Map, which can be used to explore the risk from earthquakes at locations around the world. The estimated hazards are based on the foundation's own OpenQuake engine, an open-source seismic hazard and risk model.

Climate Change by Electoral District

With the UK on the eve of a general election one political issue that has been conspicuous by its absence during the election campaign is climate change. I don't blame the political parties themselves too much for this. They spend a fortune on opinion polls and focus groups. Presumably they therefore know that the voting public are not concerned by climate change.

They should be!

A new map from University College London reveals that much of the UK has already warmed by 1°C since the 1970s. Using the Climate change projections by constituency over Great Britain map you can click on your electoral district and view the local temperature increase since the 70s. The map also allows you to view the 'past changes' in precipitation, and the increase in temperature on the hottest day of the year.

The UCL map can also be used to view a 'near-term future' projection of climate change by constituency. These changes are predicted using  a range of different global projections'.

The hexmap of UK constituencies used by the UCL map was created using Open Innovations' Hexify. This is a tool that can be used to turn defined areas into a hexagon map. You can use Hexify to create a HexJSON file which can then be used with Hexviz to create a hexmap. A hexmap of UK 2024 General Election constituencies is also available from Automatic Knowledge. Automatic Knowledge's UK Constituency Files contain a number of downloadable election resources, including a hexmap of UK electoral areas.

Tory Candidates Violently Attacked

Election riot on 26 June 1865 in Nottingham

Disturbing reports of violence, carried out by supporters of the Liberal Party, have been reported across the whole of the UK. In Shoreham a gang of 300 Liberal supporters attacked poll clerks. In Grantham Liberals set polling booths on fire. In Alfreton a crowd of thousands attacked properties 'associated with Conservatives'.

These acts of electoral violence took place during UK elections in the 19th Century and are all documented on the Victorian Election Violence Map. The Victorian Election Violence Map visualizes nearly 3,000 incidences of violence which occurred in England and Wales during national elections held between 1832 and 1914. The map shows where violent election events took place, from minor incidents (such as the breaking of windows) to major political riots involving the deaths of many people.

There were 20 general elections in Britain between the Great Reform Act of 1832 and the Great War starting in 1914. These elections were often accompanied by extreme violence. This violence often included major riots involving thousands of people, leading to the deaths of many people and large scale property damage. For example on the first day of polling in the 1868 General Election, there were at least 18 different riots across England & Wales. 

Supporters of the Conservative Party were not averse to using violence themselves. For example in Worthing the Liberal candidate for Mid-Sussex, Mr Hubbard, was "showered with stones upon his arrival to an election" and a "member of his party, Major Gaisford, was reportedly badly cut in the face". In "Andover, the Conservative party brought in railway navigators from Salisbury and Bishopstoke and plied them with beer, then set them loose in the crowd."

France Surrenders to Facism

The far-right National Rally has won 33% of the vote in the first round of France's parliamentary elections. You can view the results in each constituency in Le Monde's French elections: Map of the first round's results.

Le Monde's map is a sea of brown, revealing that the National Rally has managed to become the most popular party in electoral districts across the whole country. There is a degree of the non-equal representation of districts on the map over-egging the popularity of the National Rally. Generally the party performed better in the larger rural areas and not so well in the smaller inner-city districts.

Zoom in on many of France's largest cities and you can see that the National Rally is not quite so popular in the more multi-racial areas of the country. For example there is a distinct lack of support for the National Rally in most of the electoral districts in Paris, Lyon and Toulouse (although the National Rally were the most popular party in nearly every district in Marseille - the second most populous city in France).

However despite the fact that Le Monde's election map visually over emphasizes the scale of National Rally's success there is no denying that the far right party were the most popular party in the first round of the French parliamentary elections.

If you click on a district on the map you can view which candidates have qualified for the second round of the election. On Sunday France will hold the second-round of elections. In previous elections the center-right and center-left parties have often agreed to stand down candidates in order to avoid splitting the anti-National Rally vote. The left-wing alliance, the New Popular Front, has already announced that they will withdraw all their third-place candidates from the next round of elections. The French President Emmanuel Macron has also asked voters to support 'republican and democratic' candidates.

In The main lessons of the first round in six maps (paywalled) Le Parisien has mapped out the success of individual parties across the whole of France. For example the map above shows the rise and fall in support for Macron's 'presidential bloc' since 2022. Macron's bloc has seen a significant fall in support from 26% to 20% in just two years. Le Parisien has also mapped the high turnout in Sunday's election across the whole of France (almost up 20% since 2022). 

One result of the high turnout was that a number of candidates were able to be elected in the first round. The main winners were again the far-right National Rally, who already have 39 deputies elected to the French parliament from the 76 candidates directly elected in the first round of the French elections.

Bloomberg published an interesting map visualizing the surge of support for Le Pen's National Rally. In Le Pen’s Support Surges in Nearly Every City, Town and Village in France (paywalled) peak markers are used to show the rise in votes for National Rally at the municipality level. The map shows that the party increased its share of the vote across most of the country.