International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
.A.- The need for up-to-date topographical information 
showing the situation just before the emergency situation; 
B.- The need to acquire as quickly as possible heterogeneous 
geospatial data from the disaster area after the emergency 
situation in order to understand and monitor the situation, to 
evaluate the damages and the risk for additional damages and 
injuries, to plan and monitor the rescue/recovery operations. 
Same data (as high resolution imagery) should be recorded 
repeatedly over the same area for monitoring needs. In 
addition, the team has to address the problem of integrating 
these layers of data from divers origins, not necessary 
homogenous, having different resolutions and precision. 
In a real life situation, the separation between these two main 
cartographic problems is not so clear. In many areas the existing 
topographic information is not updated on a regular basis and 
some important changes for examples for the local road network 
(such as access roads to industrial facilities) are not included in 
the last version of the topographic database. The Emergency 
Mapping Group will face the challenge for example to use the 
new imagery layers recorded after the disaster for 'situation 
understanding’ purposes and in the same time for the updating 
of the old topographic information. Same confusions and errors 
could happen. 
Another possible issue is the delivery of geomatics outputs to 
the main group of users, the emergency services, in order to 
help the understanding of the ongoing changes. This issue is not 
discussed in the present paper. 
4. OUR TESTS IN OTTAWA SOUTH 
4.1 Description of the test 
The Centre of Topographic Information (CTI) began to 
investigate in 2001 the development of rapid mapping and rapid 
change detection systems for current mapping operations and 
for emergency situations. A contract was granted in 2002 to 
Mosaic Mapping Systems Inc. of Ottawa for an airborne survey 
of a test area of about 5 square kilometres (1.5 km x 3.5 km). 
The contractor had to provide a turnkey airborne survey, 
complete with data acquisition of LIDAR and digital camera 
data, and GPS field control. 
The overall objective of this project was to test and evaluate the 
acquisition, processing and handling of LIDAR DEM data 
collected simultaneously with optical data and their contribution 
to a rapid change detection system. 
Flight and LIDAR Parameters: 
Flight Altitude: apprx. 300 m AGL (Above Ground Level) 
Swath Width: apprx. 300 m 
Laser Wavelength: 0.9 um 
Scan Angle: up to +/- 30 degrees 
Flight Overlap: 40 percent 
Flight line Spacing:apprx. 140 m 
Point Density: 0.4 m x 1.0 m (approximate and adjustable) 
Scan Rate: 34 hz (approximate and adjustable) 
Data rate: 10 MHz 
Beam Divergence: 3.0 mrads (90 cm spot size at 300 m) 
Collection Mode: Last Pulse and Intensity 
Digital Camera Data: 
[mage size: 2300 x 3500 pixels, 
Resolution: about 8 cm, at 300 m 
Image Swath Size: apprx 400 m x 260 m, at 300 m 
A second flight (recording optical images only) was flown at 
900m above ground. 
The test area was chosen in the South of the City of Ottawa 
along the Rideau River, close to the Ottawa Airport. One of the 
selection criteria was to have a variety of topographic features 
and land use categories: roads, railways, two important bridges 
over the Rideau river, residential and industrial areas with 
buildings of different sizes, lakes, some vegetation spread 
across the residential area, a more dense forest area along the 
river and some terrain height variation. 
4.2 Reference data 
Landscape change detection analysis involves the comparison 
of two or more spatio-temporal datasets and the identification 
and location of differences in the patterns of two spatio- 
temporal datasets. For change detection operations it is 
important the have a good spatial registration between the two 
data sets. In order to test the change detection operations and 
other mapping operations needed for emergency mapping as 
merging/fusion of heterogeneous geospatial data sets, more 
imagery and digital topographic data were acquired for the tes 
area: 
- National Topographic Data Base (NTDB) data set 31G05; 
- Colour aerial photography from 1999 at 1:15 000 scale. 
In addition, for the evaluation of the LIDAR data and the 
associated digital camera, a data set of over 230 check points 
have been measured by the Mapping Services Branch personnel 
using the GPS kinematic technique. This data set was used only 
as independent checkpoints and was not communicated to the 
contracting company. A number of 6 points were signalized. 
Part of the GPS check points were recorded as profiles over 
different slope shapes: across and along a railway track, alonga 
country road with moderate, constant slope, across a 5 levels 
building and its attached flat parking lot, etc. 
4.3 Results of Ottawa tests 
4.3.1 A priori accuracy estimation: 
The expected accuracy was estimated a priori using the 
manufacturer specifications of all the components of the Mosaic 
Mapping LIDAR system and taking in account the company’s 
experience in operating the system in standard terrain 
conditions and in processing the data. 
Considering a confidence level of 95% (2 sigma), the expected 
accuracy of the LIDAR data was as following: 
a.- For the absolute Vertical Accuracy: +/- 0.15 to 0.25 
meters on Hard Surfaces; +/- 0.25 to 0.40 meters on 
Soft/Vegetated Surfaces (flat to rolling terrain); +/- 0.40 to 0.75 
metres on Soft/Vegetated Surfaces (hilly terrain). 
b.- For the absolute Horizontal Accuracy: +/- 0.75 to 1.0 
meter on all but extremely hilly terrain. 
4.3.2 Evaluation of the LIDAR DSM and DEM: 
The LIDAR data was delivered in two versions: 
- The ‘original’ data set containing all the laser signal returns 
called ‘all-returns’ was used to produce the digital surface 
model (DSM), 
- The filtered data set produced using 2 automatic filtering 
processes and an interactive editing, intended to eliminate all 
the noise, the buildings, the trees an all objects above the 
Inter 
—— 
grou 
(DE 
nam 
Visu 
A fi 
LID. 
softy 
(hig 
DEN 
mor] 
DSN 
num 
data 
emel 
the { 
brid; 
for t 
Ac 
phot 
phot 
phot 
spik 
terra 
reso 
iden 
furtl 
bee 
are 
havi 
of tl 
allo 
less 
map 
the « 
Som 
elev: 
the 
exis 
elev: 
appl 
‘Rar 
usin, 
Six 
redu 
were 
elev: 
the 
com 
anal; 
Eval 
a.D 
A St 
prec 
the 
desc 
3me 
With 
of-ra 
expe 
data 
and 
970