How England’s complicated political geography is confusing coronavirus rules

James Cheshire, UCL and Alex Singleton, University of Liverpool

As England emerged from its second national lockdown in early December, Boris Johnson, the UK prime minister, faced an onslaught of questions from MPs on both sides of the House of Commons. Each demanded clarity on what the arrangements would be for their particular constituency under the multi-layered tiers that would impose different COVID-19 restrictions on different areas.

They saw an ad-hoc logic behind the system outlined in the bill they were being asked to vote into law. In some cases – such as in Kent – restrictions were too general. In others – such as Slough – they were too specific.

Johnson responded by saying future restrictions would be “as granular as possible … to reflect … the human geography of the epidemic”. In theory, a more localised tiered approach is exactly what is needed once national infection rates come under control. It rekindles the “whack-a-mole” strategy for the flare-ups Johnson referred to earlier in the year. In reality, however, the government – like the rest of us – is looking increasingly confused by the complicated geographical units used to govern and map the country.

It could opt for obscure statistical units that best capture local outbreaks but that few people understand, or choose from a long menu of options used by local or national government. There’s something of a pick ‘n’ mix strategy at present that betrays how the UK’s geographic units were designed by different bodies, with little coordination, for a whole range of conflicting purposes – none of which were managing a pandemic. The result is a confusion of seemingly conflicting messages across government communications.

This is not helped by the fact that maps based on the same data produce very different pictures of the crisis if you split up the country differently. Depending on the size of the population of the area, you can come out with an infection rate as low as 295 per 100,000 people or as high as 736 per 100,000.

Maps showing how COVID infection rates appear different depending on how you define a geographical area.
How is the infection spreading? Depends on how you count.

For this reason, scientists tracking the spread of the virus prefer to use units that encompass roughly the same number of people, which are geographies developed for the census (so called “output areas”). This approach has several advantages. COVID-19 hotspots can be linked to other contextual data, for example, such as on the ethnic makeup or the deprivation of an area.

But these units are not how the country is governed. For that, England is divided into constituencies and counties and “combined authorities” – to name just a few of the different units of governance. Map COVID-19 rates across these boundaries and you will get even more different infection rates, since a constituency can include a densely packed town and a sparsely populated rural area, for example. It’s an impossible problem to solve, but it can be managed through consistent policies and geography.

This is important because, as we’ve seen, local councils, MPs and metro mayors want to negotiate their own lockdown terms. Many combined authorities (city regions) are bristling at being treated as similar even if they are experiencing significantly varying disease patterns at local levels.

In London, many are questioning the rationale for treating the entire capital the same and cracks are appearing in the one-size-fits-all approach. Greenwich council, for example, entered into a heated argument with central government over its unilateral decision to close schools.

These disagreements show what happens when there is confusion about how data on infections should be interpreted. And when local, regional and national governments can’t agree, the public becomes confused too. That reduces compliance with the rules and ultimately allows the virus to spread more rapidly.

The law that England’s tiered restrictions are based upon has done little to simplify things. It previously listed the geography of counties and unitary authorities, but the public communication included the larger and more regional geography of combined authorities. The most recent legal amendments that have placed Greater London, and parts of Essex and Hertfordshire into Tier 3 are, in some cases, being set at a different geography again. The likes of Rochford District Council now make the list, for example, rather than being included in the broader Essex County Council as it was previously.

If more localised restrictions are to have a fighting chance of success, they need to do a better job of reflecting this complex and conflicting geography, even if only to give a clearer picture of how COVID-19 is spreading. The government would then be able to better communicate why particular restrictions are necessary to help control the pandemic. If people are told clearly why, and where, restrictions are being applied, they are much more likely to comply – potentially saving their own lives and the lives of others.

James Cheshire, Professor of Geographic Information and Cartography, UCL and Alex Singleton, Professor of Geographic Information Science, University of Liverpool

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Atlas of the Invisible

Really excited to announce that Atlas of the Invisible, the third book I have co-authored with Oliver Uberti will be in bookshops soon! You can pre-order it here!

Next slide please: what’s wrong with the UK government’s coronavirus charts

“If you can’t convince them, confuse them.” If you watched the UK government’s COVID-19 briefing to announce and England-wide lockdown, you might have been reminded of this quote by Harry S Truman. Following slide after slide of maps and charts, there was growing frustration about the way nationally important statistics were being presented to the public.

Getting these things right is important. We’ve seen previously and in this pandemic that trust in government influences whether people follow public health guidelines. And in a UK survey earlier this year, those who had low levels of trust in the government’s ability to handle the outbreak were twice as likely to think its response had been confused and inconsistent. While a set of confusing slides won’t alone dictate how people behave, these things add up.

We don’t need high production values, or even much polish – it’s nice to feel like we’re seeing the latest data rather than something endlessly adjusted – but being comprehensible and looking professional will help support the message. At the moment, these slide decks are reminiscent of rushed conference presentations pieced together while the previous presenter was speaking. Here’s how to fix that.

Explain your working

Perhaps the biggest betrayal to an audience eager to understand is the phrase “as you can see”. It’s repeated many times at these briefings, and it’s too quickly followed by “next slide please”. The information shown is complex and takes a moment to digest. The presenters – the UK government’s chief medical adviser Chris Whitty and its chief scientific adviser Sir Patrick Vallance – need to slow down.

On the map below from briefing in question, how many of us noticed that the weekly case rates per 100,000 people didn’t increase by the same amount each time in the key? We had intervals of 25 for the first two categories, but then jumps of 50 until 200+. The map’s design also failed to show that the rate far exceeded 200 cases per 100,000 people in some areas. Wigan, for example, had 622 cases per 100,000 people.

Map showing infection rates per 100,000 people across England, with different shades of purple marking severity
www.gov.uk

One goal of a map maker is to reveal patterns that may exist in the data, and colouring is key to this – they have to decide when to move from one colour to another. In some cases it’s preferable to split up a narrow part of the distribution into lots of colours and then assign the rest to a few. Or you might assign each part of the distribution equally. Either is fine, but it needs to be explained, or else it’s a nuance that will get missed or misinterpreted.

The choice made for this map overemphasises small leaps in small numbers at the expense of big leaps in large numbers. Unless the values up to 25 and those between 25 and 50 had significance in policy, they could have been lumped into 0-50. Likewise, the map suggests anything greater than 200 doesn’t really matter – that a rate of 201 deserves the same colour as a rate of 601. This doesn’t seem right to me. But the point is, this system needs to be explained, because choosing different intervals can create a very different impression.

Consider the following graphs. The top left is the same as the government’s purple one above, whereas the others present exactly the same data, just with different sized intervals.

Four maps of England based on the same data as the graph above, but using different keys and showing different patterns
Author provided

On this point as well, the presenters made their lives hard by using national maps when most of the action is in cities. These are hard to see at this scale. The maps pulled out London, but should have done the same for other urban areas.

Presentation matters too

On this map, we were supposed to focus on the dark brown areas – these are bad news. But instead our eyes can’t resist the greens. Whitty had to tell us that the brown areas were what we should be looking at.

Map showing changes in infection rate across England, with brown showing increases and green decreases
www.gov.uk

These greens and browns are an industry standard colour palette, but more intuitive alternatives exist, such as light blues for slowing rates and darker reds for the increasing ones.

Many people online also complained that the slides didn’t fit the screen. This was an error seen on the BBC only, which had set them up wrong, and wasn’t the government’s fault. However, it does suggest the government isn’t considering what devices people will use to view the press conferences. They appear to be designing for the 50-inch television they are viewing and not for the many people streaming or catching up on their phones.

It’s always a risky strategy to push content right to the edge of slides, as things can get cut off. The layout also failed to account for the chyrons that appear at the bottom of news broadcasts, which could easily have been anticipated and designed for.

Try and keep it simple

“This is a complicated slide,” said Sir Patrick Vallance as he drew things to a close, forgiving us for not fully understanding it. But this slide was crucial. It was the climax to the case for lockdown. The 16 maps and graphics that came before were just preamble. The two graphs on this slide told us that the NHS would likely run out of capacity to treat the sickest patients in only a few weeks if we didn’t act. It was all he needed to show.

Graph projecting when NHS bed capacity would run out during the second wave
www.gov.uk

Unfortunately, the dates were misaligned on both graphs (the one on the right takes us to the end of the year, the left mid-December). It’s splitting hairs perhaps, but it demonstrates again that no one took a breather to dot the Is and cross the Ts.

The abundance of acronyms and specialist language is also symptomatic of trying to throw too much at a general audience to build credibility through complexity. This approach risks alienating the audience – when actually there was one key message on Saturday: without lockdown we’ll run out of hospital beds within a few weeks and people we could otherwise save will die.

I want to be clear that I have tremendous respect for the teams of people involved in creating these maps and graphics. I also have sympathy with the scientific advisers themselves, who are treading the increasingly strained tightrope between science and politics. The fact that they are showing such a rich array of data in some quite interesting ways is a really good thing, and we need more of it.

But data visualisation and communication is different to epidemiological modelling. It’s hard to do well, even harder under pressure, though it is possible. Unfortunately, if the government briefings are anything to go by, it remains an overlooked and undervalued skill.

James Cheshire, Professor of Geographic Information and Cartography, UCL

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Why eye-catching graphics are vital for getting to grips with climate change

Like many people, the first graph I ever saw explaining climate change was in a school geography textbook. It showed the “hockey stick” curve of the Earth’s surface temperature over time, which has become one of the world’s most recognisable line graphs.

Despite relatively minor fluctuations, the line on the graph depicting global surface temperature remains almost horizontal across centuries, before suddenly inclining to an almost vertical trajectory over the past 50 years. Since 1970 the rate of global temperature increase has hit an unprecedented 1.7°C per century.

One challenge of understanding the information contained in this hockey stick graph – and this is a gift to climate-change deniers – is the inclusion of the grey fuzz of “uncertainty data”: outlying data points that can be cherry-picked to raise doubts about the mass of evidence supporting a general warming trend.

Global surface warming: the hockey stick

A graph displaying changes in global surface temperature
Changes in global surface temperature as reconstructed (between 1-2000) and observed (between 1850-2020). IPCC 2021

Uncertainty is a complex thing to communicate in a single chart. In 2018 the UK-based climate scientist Ed Hawkins chose to omit it altogether when he presented his “warming stripes” graphic to help clearly visualise key trends in climate data. Hawkins explained that the warming stripes were designed to remove all superfluous information, leaving behind only the undeniable scientific evidence of a steadily warming world.

Climate change in warming stripes

A series of stripes, shifting from blue on the left to red on the right, represents warming global temperatures over time.
Wikimedia

If getting to grips with all the data and complexity in the hockey stick required a long read, Hawkins’ climate stripes give us the headline. The stripes are now a global phenomenon, having appeared on the lapels of US senators, the ties of TV weather presenters and on the front cover of The Economist.

As calls for change grow louder in light of the latest IPCC (Intergovernmental Panel on Climate Change) report and in the run up to COP26 conference in Glasgow this November, it’s time to focus on how data visualisation can help people grasp the challenges that lie ahead.

The power of maps

One misconception about the climate crisis is that warming will be uniform across the world. Deniers cite cold fronts or blizzards as evidence that warming is exaggerated, or hark back to past heatwaves – such as that experienced by the UK in 1976 when temperatures exceeded 35°C – as proof that the scientists have got it wrong.

Apart from this misleading conflation of weather (daily conditions) and climate (long-term conditions), this kind of argument misses the complex patchwork of effects that interact to create what gets reported in the headline figures.

Maps can be an invaluable weapon against this misunderstanding. For the first time, the IPCC has released an “interactive atlas” with its latest report, allowing audiences to pan and zoom through the data themselves. But if you give the IPCC’s atlas a try, you can see how it’s hard to capture complexity for a specialist audience while retaining simplicity for a global audience.

Most users are unlikely to closely engage with towering datasets named ‘CMIP5’ or ‘APHRODITE’, or with the mass of code that constitutes the IPCC-WG1 repository on Github. Although it’s a step in the right direction, what is needed are more universally accessible visualisations that are able to show where we’re heading in no uncertain terms.

With that in mind, when I set out to map global warming for a new book entitled Atlas of the Invisible, my co-author Oliver Uberti and I chose to combine the most important lessons from the warming stripes with the intricacies of geographical context.

This intriguingly named “Peirce quincuncial” projection, which you can see below, is a type of 2D map that flattens the Earth into a grid of 130 mini maps called tiles. Like all projections, it’s not a perfect representation of the 3D Earth, since some areas are stretched more than others. But it lets us create a series of tiles representing the planet in each year from 1890 to 2019, coloured by how and where temperatures deviated from a reliable baseline measured between 1961 and 1990. Blue areas represent temperature anomalies between -2°C and 0°C, while red areas represent anomalies between 0°C and 3°C and grey represents insufficient data.

Heat gradient map

A series of coloured maps represents the global warming trend
This data visualisation depicts the trend towards warming across the world, from the upcoming book Atlas of the Invisible. Author provided

Reading the images from left to right reveals that while heatwaves and cold spells speckle the grid, tiles representing the current century are increasingly filled with warm tones. For example, compare the few pink splotches in 1976 when the UK experienced its famous heatwave to years later in 2006 and 2016 when ruddy hues spanned the globe. In fact, the ten hottest years on record have occurred since 2005.

Heat gradient map: specific years

A visualisation of increasing global temperatures from 1976-2016, from Atlas of the Invisible.
A visualisation of increasing global temperatures from 1976-2016. MET Office Hadley Centre Hadcrut 4.6

Time to think local

When mitigation targets aim to keep the overall global temperature increase at an average of below 1.5 or 2°C, we need data visualisations to remind us that there can still be large local variations even when such targets are achieved, with the warming creating drastic and often devastating conditions for those living in affected areas.

Generalised warming will inevitably affect some places far worse than others, causing knock-on effects like sea-level rises and storms in different areas. For proof, look to the 2021 summer heatwave experienced by many parts of Europe yet escaped by the UK, the “heat dome” that scorched British Columbia in June, or the Arctic, where temperatures are rising at twice the global rate.

Even within cities, conditions can vary from neighbourhood to neighbourhood. Across the US, global warming is compounding the legacy of racist housing policies enacted through a process known as redlining. This rated the “investment risk” of urban areas, condemning many black neighbourhoods to a “hazardous” rating and thus to reduced infrastructure and increased poverty.

As the New York Times has expertly mapped, such areas saw a lack of investment in – amongst other things – green spaces and street trees. This has resulted in some historically redlined neighbourhoods suffering summers that are up to 7°C warmer compared to their non-redlined counterparts.

Maps reveal these social injustices in the UK, too. Local authorities are under pressure to turn a blind eye to flood-risk maps in order to permit thousands of “affordable” homes to be built for those priced out of higher ground.

The power of maps lies in their ability to show us simultaneously that as global average temperatures rise, local conditions threaten to become ever more extreme. We now need to better harness that power to inspire action.

James Cheshire, Professor of Geographic Information and Cartography, UCL

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Different Maps, Same Data

The maps above were created for an article in The Conversation entitled Next slide please: data visualisation expert on what’s wrong with the UK government’s coronavirus charts. In it I argue that there needs to be better data visualisations in government briefings. I give a particular example about how maps can appear differently depending on the choice of colours – in this case case rates of COVID-19 in the week up to the 30th October. The maps were designed to be created in R and I have posted the code below.

The case data comes from here, and the boundary data here. It’s a bit messy to join them so I have done this already, data url is in the code.

#Packages
library(tmap)
library(rgdal)
library(sp)
library(spatialEco)

# load in file

input<- readOGR(dsn="http://www2.geog.ucl.ac.uk/~ucfa012/COVID_Oct30_Rate.GeoJSON", layer="COVID_Oct30_Rate")

# Remove the Nas (this is Scotland and Wales and a function from the SpatialEco package)
input<- sp.na.omit(input, col.name="LA_Cln")
          
#Government choice - extra colours for low values           
opt1<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, breaks=c(0,25,50,100,150,200,736.2), title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)
  

#Suggestion 2 - same colour breaks up until 200+ plus cases          
opt2<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, breaks=c(0,50,100,150,200,736.2),title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)

#Suggestion 3 - consistent breaks up to 500+ cases          
opt3<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, breaks=c(0,100,200,300, 400, 500,736.2),title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)


#Suggestion 4 - natural breaks in the distribution using the Jenks algorithm         
opt4<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, style="jenks",title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)


#Suggestion 5 - equal numbers of local authorities in each colour band         
opt5<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, style="quantile",title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)

#Suggestion 6 - extra colours for higher values          
opt6<-tm_shape(input) + 
  tm_fill("Rate_num", palette = "Reds", legend.hist = T, breaks=c(100,200,300,400,500,550,600,650,700,736.2),title="COVID-19 Cases\nPer 100k (Weekly)") +
  tm_borders(col="black", lwd=0.1)

#uncomment below if you want to save as PDF. I have only selected 4 of the above options.
#pdf("map.pdf",width=20, height=5)
tmap_arrange(opt1,opt2,opt6,opt4, ncol=2)
#dev.off()

#Voila!

Mapping Nitrogen Dioxide Across Europe

Here is a short video I did for the Royal Geographical Society to show how new satellite data can reveal the daily changes to nitrogen dioxide (and other pollutants) in the atmosphere in unprecedented detail. For a longer version of the video and further explanation see here.

How the Victorians Mapped London’s Cholera

It is, of course, John Snow who is credited with using maps to demonstrate that the clusters of deaths from cholera in London’s Soho during London’s 1854 outbreak were caused by contaminated water. This marked a major shift in thinking away from the disease being transmitted through dirty air: the more widely accepted theory at the time.
However, it wasn’t just Snow producing innovative maps and charts to support his cause. Snow was part of an arms race to get the best data communicated by the most compelling maps/ charts, to evidence his side of the debate against his contemporaries – people like William Farr who was also a master data visualiser.
The Wellcome Collection’s image catalogue contains many great examples of maps and charts produced around the 1850s . These images are high resolution and free to use under a CC-BY 4.0 license. Most have very little information associated with them, but I think many are worth sharing because they such amazing examples of Victorian data visualisation. I have pasted them here with the catalogue details where they have them (in no particular order). I’ve included the original links back to the high resolution image in the collection. Enjoy!

Plan Showing the Ascertained Deaths from Cholera (John Snow)

Source
Details: Plan Showing the Ascertained Deaths from Cholera in Part of the Parishes of St. James, Westminster and St. Anne, Soho, during the summer and autumn of 1854. The plan is from a report on the cholera outbreak in the Parish of St. James, Westminster, during the autumn of 1854 / presented to the vestry by the Cholera Inquiry Committee, July 1855. Report is on the cholera outbreaks in central London in 1832, 1848-9,1851,1852, 1853 and (specifically) 1854. Much reference is made to a public water pump in Broad Street. Meteorological conditions are given. Information is reported about the population, housing, sanitation, sewerage, cess pools etc. and the water supply in the Soho area.

Report on the mortality of cholera in England. Charts showing the temperature and mortality of London for every week of 11 years (1840 – 1850).

Source

Diagram of cholera deaths in England during 1849


Source
Details: Diagram representing the deaths from cholera and diarrhoea in England on each day of the year 1849 with the meteorological phenomena registered at Greenwich on the corresponding days.

Map of England showing prevalence of cholera, 1849

Source
Details: Map of England shaded to show the prevalence of cholera in the several districts during the epidemic of 1849. The relative degree of mortality is expressed in the darkness of the shading. The dates indicate the time at which the epidemic broke out.

Map showing the distribution of cholera in London and its environs, from 27th June to 22st July, 1866.

Source

Map of the Parish of Bethnal Green, Shewing the Cholera Mist in 1848-1849

Source

Cholera Map of the Metropolis 1849

Source

Cholera map showing intensity of cholera attack in sections

Source

Mortality of cholera graphed against elevation

Source

London mortality; plague years & cholera years

Source
Details: Diagrams showing mortality in London during plague years of 1593, 1603, 1625, 1636, 1665, and the cholera year of 1849.

Report on the mortality of cholera in England. Chart for the year 1849.


Source

Madras Cholera Clock (not London I know, but really nice chart!)Source

Point Pattern Analysis using Ecological Methods in R

Here is a quick example for how to get started with some of the more sophisticated point pattern analysis tools that have been developed for ecologists – principally the adehabitathr package – but that are very useful for human data. Ecologists deploy point pattern analysis to establish the “home range” of a particular animal based on the known locations it has been sighted (either directly or remotely via camera traps). Essentially it is where the animal spends most of its time. In the case of human datasets the analogy can be extended to identify areas where most crimes are committed – hotspots – or to identify the activity spaces of individuals or the catchment areas of services such as schools and hospitals.
This tutorial offers a rough analysis of crime data in London so the maps should not be taken as definitive – I’ve just used them as a starting point here.
Police crime data have been taken from here https://data.police.uk/data/. This tutorial London uses data from London’s Metropolitan Police (September 2017).
For the files used below click here.

#Load in the library and the csv file.
library("rgdal")
library(raster)
library(adehabitatHR)
input<- read.csv("2017-09-metropolitan-street.csv")
input<- input[,1:10] #We only need the first 10 columns
input<- input[complete.cases(input),] #This line of code removes rows with NA values in the data.

At the moment input is a basic data frame. We need to convert the data frame into a spatial object. Note we have first specified our epsg code as 4326 since the coordinates are in WGS84. We then use spTransform to reproject the data into British National Grid – so the coordinate values are in meters.

Crime.Spatial<- SpatialPointsDataFrame(input[,5:6], input, proj4string = CRS("+init=epsg:4326"))
Crime.Spatial<- spTransform(Crime.Spatial, CRS("+init=epsg:27700")) #We now project from WGS84 for to British National Grid
plot(Crime.Spatial) #Plot the data

The plot reveals that we have crimes across the UK, not just in London. So we need an outline of London to help limit the view. Here we load in a shapefile of the Greater London Authority Boundary.

London<- readOGR(".", layer="GLA_outline")

Thinking ahead we may wish to compare a number of density estimates, so they need to be performed across a consistently sized grid. Here we create an empty grid in advance to feed into the kernelUD function.

Extent<- extent(London) #this is the geographic extent of the grid. It is based on the London object.
#Here we specify the size of each grid cell in metres (since those are the units our data are projected in).
resolution<- 500
#This is some magic that creates the empty grid
x <- seq(Extent[1],Extent[2],by=resolution)  # where resolution is the pixel size you desire
y <- seq(Extent[3],Extent[4],by=resolution)
xy <- expand.grid(x=x,y=y)
coordinates(xy) <- ~x+y
gridded(xy) <- TRUE
#You can see the grid here (this may appear solid black if the cells are small)
plot(xy)
plot(London, border="red", add=T)

OK now run the density estimation note we use grid= xy utlise the grid we just created. This is for all crime in London.

all <- raster(kernelUD(Crime.Spatial, h="href", grid = xy)) #Note we are running two functions here - first KernelUD then converting the result to a raster object.
#First results
plot(all)
plot(London, border="red", add=T)


Unsurprisingly we have a hotpot over the centre of London. Are there differences for specific crime types? We may, for example, wish to look at the density of burglaries.

plot(Crime.Spatial[Crime.Spatial$Crime.type=="Burglary",]) # quick plot of burglary points
burglary<- raster(kernelUD(Crime.Spatial[Crime.Spatial$Crime.type=="Burglary",], h="href", grid = xy))
plot(burglary)
plot(London, border="red", add=T)


There’s a slight difference but still it’s tricky to see if there are areas where burglaries concentrate more compared to the distribution of all crimes. A very rough way to do this is to divide one density grid by the other.

both<-burglary/all
plot(both)
plot(London, border="red", add=T)


This hasn’t worked particularly well since there are edge effects on the density grid that are causing issues due to a few stray points at the edge of the grid. We can solve this by capping the values we map – in this we are only showing values of between 0 and 1. Some more interesting structures emerge with burglary occuring in more residential areas, as expected.

both2 <- both
both2[both <= 0] <- NA
both2[both >= 1] <- NA
#Now we can see the hotspots much more clearly.
plot(both2)
plot(London, add=T)


There’s many more sophisticated approaches to this kind of analysis – I’d encourgae you to look at the adehabitatHR vignette on the package page.

The Ultimate Gift List for Map Lovers

Here’s my 2017/2018 Ultimate Gift List for Map Lovers! All the recommendations are for products I own – or have seen – and can genuinely endorse. I’ve listed them under broad categories of people you might want to buy them for. Hopefully they cater for a range of map-related interests and budgets. Enjoy!

For those who like a good coffee table book


(Amazon UK; Amazon US)
Okay, I may be a little biased putting Where the Animals Go on here, but is has received significant critical acclaim and won a number of awards so I know others have enjoyed it! Not many people are creating original map books from scratch these days so we put a huge amount of effort into getting the cartography right for this – there’s lots of innovations in approach that we’re really proud of. Its not just for animal lovers – there’s data & technology to think about as well. We’ve been amazed at the age range of people coming to our talks from kids to grandparents and all in between – each enjoying the book for different reasons so it seems a safe bet if you want something for the whole family. It also comes as Die Wege Der Tiere and Atlas De La Vie Sauvage.

As I said above, there aren’t many people making books of original map content, instead there’s an increasing number of books containing compilations or edited selections of maps. In this crowded market I’m becoming increasingly discerning about what books are genuine contributions rather than a rehash of other people’s work. That said, there are a few lovely examples of map compendiums that contain great commentary and some nice surprises. The Phantom Atlas (US link) is one of those. It is also beautifully produced with some nice design touches and high print quality. I got it for Christmas last year and it remains a favourite.

You would have seen Maps (US link) in a bookshop before now. It’s designed for kids but the illustrations are brilliant and it’s really immersive…just make sure the person you’re giving it to has tall bookshelves since it’s a big book!

For those in search of a good poster

I have this Upside Down World Map (US Link) up on the wall of my office and almost anyone who comes in does a double-take when they see it. It’s a great conversation starter and nicely makes the point that the maps can be used to make us see the world differently! (Note that it comes folded rather than rolled).

The British library has an extensive map collection and it’s possible to purchase prints of some of them. There’s a variety of formats on offer and lots to choose from. Click here for more.

For the armchair geographer

Nowherelands (US Link) is a wonderfully produced book detailing countries that once existed but have since disappeared. It has nice touches like the countries’ stamps alongside a small map of their location. Maps are a relatively small part of the book, but it is packed with geopolitical history.

For the fashion-conscious


Threadless has loads of map themed t-shirts, most of which I think I own/ have owned in recent years.

I don’t own a t-shirt from Citee, but I’ve seen a sample and also they are cropping up more and more at conferences I attend. You can get a map for practically any city you can think of. They come in either white or black.

One Hundred Stars does a whole range of map-based clothing. I purchased one of they scarves and have been really impressed with it. More here.

For the map enthusiast


The Red Atlas (US link) has recently been published and it showcases the extraordinarily detailed maps the Soviet Union were able to create in a time before widespread satellite surveillance etc. It’s probably more suited for someone you know who really likes maps/ mapping since it can get a bit technical in places – I really enjoyed it.

Pictorial maps have brought a lot of fun to cartography. This book includes over 150 of them across six thematic sections such as  “Maps to Amuse” and “Maps for War”. All examples are taken from the US (as the title suggests), but it’s very nicely produced and full of interesting/ insightful commentary from Stephen Hornsby. UK link. US link.
If you’re still searching – click here for a gift list from a few years ago.