267 
ON THE QUALITY CHECKING OF THE AIRBORNE LASER SCANNING-BASED 
NATION WIDE ELEVATION MODEL IN FINLAND 
E. Ahokas*, H. Kaartinen, J. Hyyppa 
Finnish Geodetic Institute, Geodeetinrinne 2, 02430 Masala, Finland, eero.ahokas@fgi.fi 
Commission I, WG 1/2 
KEY WORDS: Airborne laser scanning, Optimization, Aerial survey, Accuracy, Quality control 
ABSTRACT: 
Nationwide airborne laser scanning (ALS) has been carried out in European countries such as the Netherlands and Switzerland; in 
Germany large parts of the country has been scanned and the work is going on as federal basis. The decision makers of national 
mapping agencies are realizing the benefits of ALS, e.g. significantly improved accuracy, lower processing costs and higher 
automation, and, thus, plans of performing national laser scanning are planned in many other countries. The elevation model working 
group of the Ministry of agriculture and forestry of Finland recommended the usage of airborne laser scanning for the creation of a 
new multi-purpose nationwide elevation model in 2006, and thus the National Land Survey of Finland (NLS), in co-operation with 
the Finnish Geodetic Institute (FGI) had a more than 1000 km 2 practical test on using ALS for the new nationwide elevation model 
and its quality checking in late 2006 and early 2007. The objectives of the test were: a) to define the elevation quality of ALS flights 
from two point densities 0.5-1 and 0.1 points per square metre resulting in two flight altitudes 2000 m and 5000 m using the Optech 
ALTM 3100 and scanning angle of ±20 degrees and Leica ALS50-II and scanning angle of ±20 degrees, b) to compare the quality 
derived with two different sensors (Optech ALTM 3100 and Leica ALS50-II), c) to define the quality of ALS derived DEM in 
various surface types, d) to analyse the planimetric errors, and e) to develop methods to derive ground reference data effectively. 
1. INTRODUCTION 
The digital elevation model (DEM) in 25 by 25 m 2 grid covers 
the whole Finland and it has been updated gradually to 10 by 10 
m 2 model. The height accuracy (St.Dev.) of the first model is 
1.4 m (NLS 2008) and the latter 1.1m (Oksanen and Sarjakoski, 
2006). Using exactly the same study area as in Oksanen and 
Sarjakoski, 2006, the height precision (St.Dev.) of the 25 by 25 
m 2 grid model is 2.1 m. In May 2007 half of the area of Finland 
was covered by this new 10 by 10 m 2 updated model. However, 
the height accuracy does not fulfil the modem demands. For 
example, European Union floods Directive 2007/60/EC on the 
assessment and management of flood risks entered into force on 
November 26, 2007. It requires the mapping of flood risk areas 
also in Finland. Additionally, International Civil Aviation 
Organization (ICAO) requires accurate elevation models of 
airport areas and their neighbourhoods. Some practical tests 
have shown that errors up to several tens of meters exist in the 
present model and the quality is heterogeneous (Oksanen and 
Sarjakoski, 2006). 
New techniques enable producing more precise DEMs 
nowadays. Nationwide airborne laser scanning (ALS) has been 
carried out in European countries such as the Netherlands and 
Switzerland (e.g. Artuso et al. 2003); in Germany large parts of 
the country has been scanned and the work is going on as 
federal basis. The decision makers of national mapping agencies 
are realizing the benefits of ALS, e.g. significantly improved 
accuracy, lower processing costs and higher automation, and, 
thus, plans of performing national laser scanning are planned in 
many other countries. 
The elevation model working group of the Ministry of 
agriculture and forestry of Finland recommended the usage of 
airborne laser scanning for the creation of a new multi-purpose 
nationwide elevation model in 2006, and thus the National Land 
Survey of Finland (NLS), in co-operation with the Finnish 
Geodetic Institute had a more than 1000 km 2 practical test on 
using ALS for the new nationwide elevation model and its 
quality checking in late 2006 and early 2007. 
The objectives of the test and study were: 
a) to define the elevation quality of ALS flights from two point 
densities 0.5-1 and 0.1 points per square metre resulting in two 
flight altitudes 2000 m and 5000 m using the Leica ALS50-II 
and scanning angle of ±20 degrees, 
b) to compare the quality derived with two different sensors 
(Optech ALTM 3100 and Leica ALS50-II), 
c) to define the quality of ALS derived DEM in various surface 
types, 
d) to analyse the planimetric errors of ALS surveys, and 
e) to develop and test practical methods to derive ground 
reference data effectively for large-area ALS collection. 
The test site Salo was selected by the NLS. In earlier study by 
Ahokas et al. (2005), it was preliminary shown that scanning 
angles of up to 15 degrees can be used without scarifying the 
derived elevation accuracy. Based on specifications by the 
manufacturers the following accuracies are given. Leica's 
ALS50-II accuracy assuming 0 cm GPS error is 5 cm in xy and 
4 cm in z with slant range 500 m and pulse rate 150 kHz. 
Planimetric accuracy is 21 cm and vertical 8 cm with slant 
range 2000 m and PRF 58 kHz. Flying height 5 km is supposed 
to give accuracies of 53 cm in xy and 19 cm in z when PRF is 
26 kHz. Accuracies of 64 cm and 23 cm, respectively, are given 
for 6000 m altitude. Optech ALTM Gemini gives 9 cm 
horizontal and <10 cm vertical accuracy from 500 m flying 
altitude at 100 kHz pulse rate. Planimetric accuracy is 36 cm 
and vertical <15 cm with slant range 2000 m and PRF 100 kHz. 
Quoted accuracies do not include GPS errors. 
The evaluation of the quality of ALS is problematic, since the 
object, such as forest and slope may affect more on the accuracy