Visualising the energy costs of commuting
From static graphs to online, maps via infographics
Robin Lovelace, University of Leeds (GeoTalisman)
@robinlovelace, github
Motivation• Peak oil, obesity, climate change, recession• Energy: 'master resource', affects all
See Berners-Lee and Clarke (2013)
Thinking about energy costs of transport
See Lovelace et al. (2011)
Here!Sense
of future
Where we're at: regional variabilityEnergy use per average one way trip to work: Mega Joules per trip (Etrp in MJ)
Data source: 2001 Census
Train network
• Cut continous variable in bath-sized chunks:
la$Baths <- cut(la$ET/ (la$all.all * 3.6 * 5), breaks=c(0,1,2,3)
)
One bath = 5 kWh = 3.6 * 5 MJ(MacKay, 2009)
Add baths and textin image editor
Individual-level variability
Data source: National Travel Survey (2002 - 2008)
Inequalities within areas
Lovelace et al. (2013)
Going Dutch
• Scenario of high cycling uptake
• Realistic based on Dutch data
• 'What if' not 'it will' approach
source: London Cycling Campaign
Going Dutch: Energy savings from high cycling uptake scenario
National-level comparisons
Average energy costs per one way trip to work in English regions (2001) and Dutch provinces (2010)
Going Finnish
Going Finnish: assumptions
Based on work by Finlanders Helminen, Ristimäki, M. (2007)
Energy saving from telecommuting results
Compare with cycling uptake (below)
Making analysis reproducible• Link to data: Dutch data taken from
Statistics Netherlands and English data from Casweb
• Most analysis + visualisation in R
• Result reproducible: RPubs documents + uploaded .zip folder
• RMarkdown runs code 'live'
Key functions for mapping in Rx = c("ggplot2", "sp", "rgeos", "mapproj", "rgdal",
"maptools")lapply(x, require, character.only = T)
gors <- readOGR(".", layer = "GOR_st121")
fgor <- fortify(gors, region = "ZONE_LABEL")fgor <- merge(fgor, gors@data[, c(1, 2, 3, 8, ncol(gors@data))],
by.x = "id", by.y = ZONE_LABEL")
p <- ggplot(data = fgor, aes(x = long/1000, y = lat/1000))
p + geom_polygon(data = fgor, aes(x = long/1000, y = lat/1000, fill = ET/all.all,
group = group)) + ...
Making that dynamic
• Gas guzzler map - video
• Work needed here
• Ideal would be interactive
Google's Fusion Tables
• Shpescape = for loading shp files
• As described by Dr Rae• Pros
– Fast, user friendly– Sensible presets– no need for servers
• Cons– Not flexible– Data ownership (NSA?)
Geoserver on Amazon Web Server
• Experimented with Geoserver • Running on Amazon's Web Services
(AWS), with 1 year free• Upload shapefiles, server side (Geoserver
interface) + client side (OpenLayers) edits• Not currently set-up• Pros: Flexibility, control of information,
massively scalable (geodb)• Cons: Tricky, time consuming and some
cost
Impact
• People seem to relate to research more when it's in visual form
• Very good response from people in range of other disciplines
• Still struggling to engage 'policy makers'
• Increased accessibility and potential 'impact' (in context of REF)
Taking it further
• Geo-visualisations with 'processing'
• Flow mapping in R
• Energy use at the road level
• Comparisons with other energy users
Conclusions
• Range of visualisation options available now is wider than ever - take advantage!
• Each option has pros and cons - decision should be context-specific
• Advantages of moving beyond static graphs and maps, esp. in age of 'big data'
• Don't get caught up in the details, focus on message
Go references + questionsBerners-Lee, M., & Clark, D. (2013). The Burning Question: We can’t burn half
the world's oil, coal and gas. So how do we quit? Profile BooksHelminen, V., & Ristimäki, M. (2007). Relationships between commuting
distance, frequency and telework in Finland. Journal of Transport Geography
Lovelace, R., Ballas, D., & Watson, M. (2013). A spatial microsimulation approach for the analysis of commuter patterns: from individual to regional levels. Journal of Transport Geography
Lovelace, R., Beck, S. B. M. B. M., Watson, M., & Wild, A. (2011). Assessing the energy implications of replacing car trips with bicycle trips in Sheffield, UK. Energy Policy
Email: R . Lovelace @ Leeds . ac . uk
'Eco-localisation'
• It's the localisation of economic activity (North 2010; Greer 2009)
• Extent of process depends on your perspective
• Tried to model it...
• But some things are best not quantified (and so says Vaclav Smil)