Presentation by Carlo Lavalle, Joint Research Centre, European Commission at the OECD Workshop on Spatial Dimensions of Productivity, 28-29 March 2019, Bolzano. More info: https://oe.cd/GFPBolzano2019

1. The European Commission’s science
and knowledge service
Joint Research Centre

2. From place of residence to place of activity
Emerging data and tools for territorial analyses
Carlo LAVALLE, Filipe BATISTA, Ricardo BARRANCO,
Claudia BARANZELLI, Mert KOMPIL , Chris JACOBS-CRISIONI, Andrius KUCAS
JRC.B.3 – Territorial Development unit
Workshop on Spatial Dimensions of Productivity, 28-29 March 2019 Bolzano, Italy

3. Outline
 Emerging geospatial data
 Applications
 Localisation of productive activities
 Cities’ attractiveness
 Day/Night Time population mapping
 Functional Border Areas
 New insights in Territorial Modelling

4. Emerging geospatial data sources
Data generated as a by-product
Unintentional crowd-sourced data.
 Mobile network operator (MNO)
data
 Web activity (content, traffic,
searches…)
 Social media (tweets, check-ins,
photos…)
 Transactions (costumer,
financial…)
Data generated on purpose
Intentionally produced as a core
component of ICT-based service.
Aspects in common with of Big Data.
 Navigation/mapping data (e.g.
POI, road networks), but
volunteered/collaborative or
private initiatives
 Sensors (count of vehicles,
pedestrians, air-borne, satellite)

5. Points of Interest (POI)
 Physical structures on the Earth’s surface with a functionality relevant to human
or societal activities.
 Mapped as a precise points on a (digital) map.
 Many sources:
 OpenStreetMap (VGI, free and open source)
 Navigation /mapping / sector data (proprietary) (e.g. TomTom)
 Derived from mining web services (e.g. Booking.com, TripAdvisor)
 Different levels of quality, completeness, overlap
 Different classification systems
 Different quality (completeness, accuracy…)

6. Density of Points of Interest in Paris per 500 m cells
Commerce
Food
Commerce
Other
Education
School
Education
University
Health
General
Health
Hospital
Parks Recreation Restaurant
Source: TomTom Points of Interest
Elaboration: European Commission JRC B.3
LUISA Territorial Modelling Platform, 2018

7. POI data – Application: Land Use
Land use characterization
using POI data
Part of a wider project to refine
the thematic and spatial detail of
CORINE Land Cover and map
spatiotemporal population
densities (ENACT).
Main objective: To break down
CLC class 121 (“Industrial and
Commercial Sites”) into 3 more
detailed land uses.

8. CORINE
Land Cover
LUISA Base
Map
Localisation of industry-commerce-service clusters using POI data
CORINE Land Cover LUISA Base Map

9. 1211 Production facilities (ABCDE)
1212 Commercial/service facilities (GHIJKLMN)
1213 Public facilities (OPQ)
121 Industrial and
commercial units
Localisation of industry-commerce-service clusters using POI data
CORINE
Land Cover
LUISA Base
Map

10. Original Land
Use Map (CLC
2012)
Bologna
Parma

11. LUISA
base map
Bologna
Parma

12. Torino
LUISA base map -
Population
Bologna
Parma

13. Identification of clusters of mixed services as element for walkable neighbourhoods
 Input data:
 Open Street Map points of interest (POI)
 Filtered tags include: amenity, leisure, shop and tourism (values excluded if related to cars and,
in general, vehicular mobility)
 Methodology:
 Optimised hotspot analysis (index of agglomeration)
POI – Application: Walkability

14. High
Low
Agglomeration
index
Service centres - Milan

15. High
Low
Agglomeration
index
Service centres - Milan

16. Brussels Wien
Vilnius
Athens
Copenhagen Madrid
Milan
Warsaw
Service centres – various capitals

17. Service centres – various capitals

18. New set of POI data from Google Maps
enriched with Popular Hours data.
 Fine spatial resolution and 24/7
temporal detail.
 Multiple activity types.
POI: Spatiotemporal population

19. Web mining
Applied to extract useful information from websites.
Many applications:
Non directly geospatial-oriented
 Media monitoring (e.g. EMM).
 Mining of prices for price indexes,
inflation rates.
 Citizen and costumer sentiment
(widely used by private sector to
optimize business).
Geospatial-oriented
When information can be linked to a
geographical location by means of
coordinates or place names.

20. FDI data from www.fdimarkets.com (FT)
EU28 investors: most common value (frequency)
Period
Source
Country
Sub Sector Market Motive
2003-2018
United
States
(14576)
Software
publishers,
except video
games (5279)
Regional
(8028)
Proximity to
markets or
customers
(1357)
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
$Millions
EU28 FDI: Top 10 investors (2003-2017)
USA Germany France UK Netherlands
Switzerland Spain Japan Austria Sweden
Multidimension dataset on FDI:
Spatial: Source – Destination (country/region/city).
Temporal: Monthly for 2003-2018.
Thematic: Sector, activity, type, market, motive...
Capital expenditure and Jobs created.

21. Mapping hot spots of FDI origin and destination
Source (2003-2018) Destinations

22. African FDI flows (2003-2018)Egypt investitors analysis: most common value per period (frequency)
Period Source Cntr Sub Sector Market Motive
2003 - 2007
United States
(40)
Oil & gas extraction
(63)
Regional
(6)
Domestic Market
Growth Potential
(15)
2008 - 2012
United States
(60)
Retail banking
(91)
Regional
(6)
Domestic Market
Growth Potential
(16)
2013 - 2018
UAE
(39)
Retail banking
(88)
Domestic
(5)
Domestic Market
Growth Potential
(11)
All (2003-2018)
United States
(57)
Retail banking
(137)
Regional
(7)
Domestic Market
Growth Potential
(33)
FDI Analysis: Flows in Africa

23. Malta: 1230 houses
Density: 3.89 point/km2
Housing (ask) prices from Remax (Italy) Features (Total 34):
lat - Latitude
lon - Longitude
keys - ID
price - Selling Price (€)
totalRooms - Total Rooms
bedrooms - Bed Rooms
bathrooms - Bathr
totalsqm - Total Square Meters (m2)
lotsize - Lot Size (m2)
year - Construction Year
builtArea - Building Area (m2)
parkingspaces - Number of parking spaces
floors - Number of floors
floorlevel - Floor of the house
toiletRooms - Number of toilets
energyClass - Energy class ***
energyEff - Energy Efficiency (kWh/m2 per year) ***
Features (Total 17 - yes or no):
garage - Garage (yes/no)
pool - Swimming Pool (yes/no)
renovated - Renovated (yes/no)
fireplace - Fireplace (yes/no)
terrace - Terrace (yes/no)
balcony - Balcony (yes/no)
garden - Garden (yes/no)
liftelev - Lift or Elevator (yes/no)
parking - Parking places (yes/no)
heating - Heating System (yes/no)
solar - Solar panels (yes/no)
oil - Oil heating (yes/no)
ac - Air Conditioner (yes/no)
sewer - Connected to sewer (yes/no)
pool - Swimming Pool(yes/no)
security - Alarm or security system (yes/no)
kitchen - Kitchen (yes/no)

24. Convergenc
e
Housing Cost: Milan time-series (Sep 2016 – Aug 2018)

25. Conventional + Big Data: Mapping tourism
Goal
 Tourism density maps at the highest possible temporal and spatial resolutions,
for the whole of the EU28. Tourism density defined as the number of tourists
present in a given location at a given time.
Methodology
 Downscale monthly nights-spent to local level.
Input data
 Conventional statistical data: Nights-spent (ESTAT) & Seasonal curves (NSIs)
 Big data: location of accommodation establishments (Booking.com and
TripAdvisor)

26. Mapping tourism

27. Summer
Summer
Easter
Skiing
Summer
Summer
Santa
Claus
Aurora borealis
Seasonality: Very low
Seasonality: Very high
Seasonality: medium Seasonality: high
Tiroler Unterland, Austria

28. Application: Spatiotemporal population
Day & night-time 24 grids
Integrating conventional and non-conventional data in
a coherent framework

29. Difference in population between day-
and night-time
+ in night-time
+ in daytime

30. Spatiotemporal population
Ljubljana
Night-time
population
Day-time
population

31. Functional Border Areas
Modelling travel time to land border
crossings
‘Origin’ points
 Regularly spaced points at 10 Km intervals, +
 Centroids of "urban clusters" (continuously
populated areas of 300 inhab./Km2 and at least
5k inhabitants).
‘Destination’ points
 Land border crossings, including: Paved
roads/bridges (this includes the Copenhagen-
Malmo bridge) and river ferries.

32. Functional Border Areas
Dashed polygon = Polish NUTS3 of
Szczecinecko-pyrzycki (PL427).
Officially considered a cross-border
region, but…
…only a very small portion of its
territory is within 30 or 60 minutes
from border crossings.

33. Functional Border Areas

34. Functional Border Areas

35. Functional Border Areas

36. Functional Border Areas
Positive day- /night-time ratio =
more day-time population (workers,
students…) than nighttime
(residents).

37. Functional Border Areas

38. Functional Border Areas
Positive day- /night-time ratio =
more day-time population (workers,
students…) than nighttime
(residents).

39. Functional Border Areas
4.8
3.8
2.3
4.1
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
AT AT_border IT_border IT
%
Population change (2001-2011)
0
50
100
150
200
250
300
350
400
1961 1971 1981 1991 2001 2011
Inhabitants(thousands)
Austria - Italy
AT
IT

40. Functional Border Areas
7.6
24.4
0.0
5.0
10.0
15.0
20.0
25.0
30.0
AT IT
%
Day- / night-time population ratio
Positive day- /night-time ratio =
more day-time population (workers,
students…) than nighttime
(residents).

41. New geospatial data & Territorial Modelling
Role of new data in territorial modelling:
Complement spatial and thematic accuracy of statistical data
Estimate supply and cost of services (e.g. childcare, education etc.) at sub-national
level
Evaluate attractiveness and suitabilities for investments
Perfom ad-hoc spatial analysis (spill-over effects, super-linearity growth of cities)

42. Territorial Analysis and Modelling
Territorial Impact Assessment
RHOMOLO: Investments, Infrastructures, Human Capital
Territorial Trends
LUISA: GDP and Demographic trends (Reference)

43. New analysis: demographic projection by age
class. Aging patterns in urban areas
Population over 65 years old in 2030.
Population in the core city (including
those that over 65 years old) will
decrease.
Population in the Functional Urban
Ares (including those that over 65
years old) in most cases will
increase.

44. Stay in touch
KC TP Knowledge Centre for Territorial Policies:
https://ec.europa.eu/knowledge4policy/territorial_en
Community of Practice on Cities (CoP-CITIES):
https://ec.europa.eu/knowledge4policy/territorial/topic/urban_e
n#CoP_CITIES
UDP Urban Data Platform:
https://urban.jrc.ec.Europa.eu
STRAT-Board Territorial and Urban Development Strategies:
https://urban.jrc.ec.europa.eu/strat-board
GHSL - Global Human Settlement Layer:
http://ghsl.jrc.ec.europa.eu

45. Thank you
Any questions?
Carlo.Lavalle@ec.europa.eu