1. AHN: The need for a
nation wide digital
elevation model
Niels van der Zon
November 14, 2012
SPAR Europe 2012, The Hague

3. 2011
1926
Need for spatial
data is evident
2003

5. 1. The need for a national elevation model in the
Netherlands
2. National Dutch elevation model AHN2
3. Future: AHN3, development and challenges
Subjects

6. • The Netherlands is an example
of a country highly susceptible
to both sea-level rise and river
flooding.
• 55% of its territory is below sea
level where 60% of its
population lives and 65% of its
Gross National Product (GNP) is
produced.
The Netherlands
“Dry feet, sufficient clean water and reliable and useful
information.”

7. Facts and figures:
• Flood control, drainage and irrigation, water quality,
wastewater treatment
• Developing and managing national infrastructure networks
• 10 million people and € 2,000.- billion value behind dunes
and dikes
• 17,500 km dikes / 225,000 km canals and rivers
• 3,000 diked marshes
• Water system of 35,000 km2; 65,000 km2 incl. North Sea
Watermanagement ..

8. To know how the water flows:
• through the rivers and canals
• after a dike break
• after a heavy downpour of rain
• below ground
Why is elevation data so important?!

9. .. a water model consists of 80% from an elevation model ..
Watermodel
Source: Nelen & Schuurmans

10. .. a proper watermodel is a detailed model ..
Watermodel
the outside world
100m x 100m
0.5m x 0.5m
Source: Nelen & Schuurmans

11. Benefits from detailed watermodel
Source: Nelen & Schuurmans

12. • Rijkswaterstaat and waterauthorities jointly responsible
for safety of the Netherlands by proper water- and
floodmanagement;
• Up-to-date elevation data is indispensable
• Guarantee: quality, availability and accessibility
• Government: is in charge (directing role)
• Business: companies perform work out
• Further objectives: cost saving by cooperating and
promote use of data by third parties
Why government into the lead?!

13. Product:
• detailed and accurate elevation model
• Not only used for information on elevation,
but also used for 3D-mapping
AHN: solid basis for the use of height in other national
geometric base registrations
AHN: national elevation model

14. Major drivers for AHN2:
• Legislation (december 2009: Water Act): every 6 years
quality and safety check
• Information needs: detailed data of barriers beside
floodmanagement
• Technology push: lidar technique becomes more
mature
• Climate change adaption: sea-level raising and severe
rainfalls
• Outdated data: between 2003-2005 no data collection
(not by market and not by government)
• government takes control: quarantee quality,
availability, continuity
AHN2 (2008-2012)

15. In theory:
• Height accuracy: 5cm stochastic en 5
systematic
• Mapping accuracy: objects 2x2m max.
deviation of 50cm
• Filtering: terrain dataset and another
dataset with all filtered topography
• Grid file: 0,5m grid is base product
In practice:
• Some degrees of freedom for data
acquisition companies:
• Planimetric accuracy (sigma_XY)
• Point distribution (regularity)
• Point density (# point/m2)
Quality of AHN: geometrically verifiable
(location and height)!
AHN2: the quality

16. • About 400.000.000.000 LIDAR points
• Accuracy 5cm
• Also available in grids (0.5x0.5m)
AHN2: the details

17. • Contractor is responsible to ‘supply evidence’ of the quality
demanded
• QA: done by independent contractor
• Small market, limited number of parties (3 different parties)
• No direct relation between contractor DA and QA
• Methodology developed in collaboration with University of
Twente
• More efficient and cheaper than done by government itself
AHN2: Quality Assurance
Executive
organization
Contractor
Data Acquisition
Contractor
Quality Assurance

18. Water authorities and Rijkswaterstaat
• Quality and safety check barriers
• Drainage maps
• Hydrological models
• Monitoring of coastal erosion / subsidence
• Scheduling of major infrastructural work
• Ordinances relating to water levels
• Planning buffers between ground water
levels and location of pumping stations
AHN2: applications

19. • WHY: water authorities need update for
watermanagement (i.e. safety check dikes; Water Act)
• WHAT: AHN3 is the new updated and improved
version of AHN2
with additional targets:
• Hybrid model: special products (if needed) only by
dedicated funding
• Business opportunities between companies and
government
• Funding by participating government
parties for nationwide elevation model;
aim is funding by central government
Next-generation: AHN3 (2014-2019)
basic-product AHN3
specials
business

20. 1. Further classification: terrain, building, infastructures
+ the rest
2. RGB coding and laser intensity value per point
3. Echo count for automatic classification (by business)
(e.g. vegetation); full wave optional
4. Pointclouds in LAZ format.
5. Griddata: INSPIRE-proof (RD/NAP and ETRS89/EVRS)
6. Shorter leap time (max 6 months) between data
acquisition and supply
AHN3 vs AHN2: the major changes

21. LIDAR
INSAR
DENSE matching
Mobile Mapping Systems
UAV’s
AHN3: based on LIDAR ..

22. Data integration:
Keep the AHN up-to-date without the need for
LIDAR acquisition on a yearly basis, at a
national scale
Satellite images Aerial photogrammetry 
airborne LIDAR
Major challenge AHN3

23. Production of pointclouds based on stereo
photography (corresponding points; in theory every pixel
1 height value)
Accuracy / density depends on quality of images:
• orientation parameters, overlap, resolution, radiometry
• Central governmental program for aerial photography of
the Netherlands: 10cm on a yearly basis (since 2012)
Dense matching
Resolution
Overlap
10cm
60%/30%
10cm
80%/30%
10cm
80%/60%
vertical accuracy 1σ = 9.7cm 1σ = 2.9cm 1σ = 2.8cm
% mismatch 9% > 29.1cm
(3σ)
11% > 8.7cm
(3σ)
8.5% > 8.4cm
(3σ)
AHN2
1σ < 5.0cm
0,3 % > 15 cm (3σ)

24. • Comparatively cheap .. if images are available
• Requires much computation time (hours per model)
• Suitable for pointcloud production of smaller areas
(neighborhoods, buildings, roads and bridges)
Disadvantages:
• Height accuracy less than LIDAR
• Poor in area with weak colour contrast (terrain,
streets)
• Occlusion in urban area’s; mismatch (%) is large
• (Almost) no points under vegetation/trees
• Large overlap (€) needed for accuarcy at LIDAR level
Dense matching

25. Niels van der Zon
n.vanderzon@hetwaterschapshuis.nl
AHN: The need for a
nation wide digital
elevation model