The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part BI. Beijing 2008 
4.4 Additional Equipment Used 
In addition to the aircraft and installed systems, a number of 
additional equipment items were needed. 
4.4.1 Transfer/backup dives used were 1 TB “Firewire” 
drives, and an minimum of 5 were required for each aircraft 
(one drive being filled with the current 4-6 mission, one backup 
of the previous 4-6 mission retained in the field pending data 
verification at the processing center, one in transit to the 
processing center, one at the processing center being verified 
and one in transit back to the field). 
4.4.2 GNSS base stations, of which at least five were 
required. One used at the airport field base, plus a minimum of 
two for each east-west cross-strip at the northern and southern 
ends of the block. 
4.4.3 Large-capacity batteries for GNSS base stations 
were used. Each battery could power a base station for up to 4 
days. 
4.4.4 Four-wheel-drive trucks, of which two were used. 
Many of the GNSS base station locations were located on 
unimproved rural by-ways, where use of a traditional rental car 
would be impractical. In addition, the trucks were equipped 
with “camper caps” that could be used as a sleeping area in the 
event the ground crew monitoring the GNSS base station 
needed to stay overnight. 
5. CHARACTERISTICS OF RESULTING DATA 
To date, approximately 3.29 x 10 11 points have been acquired to 
date, assuming only a single return from each outgoing laswer 
pulse. Of those, approximately 2.64 x 10 11 points have been 
retain for archiving. The following section summarizes various 
attributes of the acquired data. 
5.1 Comparison with lPiA Data 
As mentioned previously, some of the data within the project 
area was collected at an earlier date using a conventional (i.e., 
lPiA) system. Data from the MPiA system used in this project 
provide accuracies at least as good as those from the previous 
lPiA system. 
5.2 Comparison with Advertized System Specifications 
Data collections are consistently meeting the targeted RMSE 
accuracy specifications of 20 cm Z and 50 cm XY. Typical 
estimated accuracies for ALS50-II systems, expressed as 
standard deviation, are 13 cm Z and 20 cm XY, assuming 10 
cm GNSS error. 10-15 cm GNSS error was typically achieved 
by using a minimum of 5 base stations for each flight, one 
located at the airport and 2 on each operating cross strip at the 
job site. Factoring in the slightly higher GNSS error than used 
on the ALS50-II accuracy estimates, the remaining difference 
between ALS50-II estimated accuracy and the achieved 
accuracy is accounted for by the fact that final accuracy 
achieved is quoted in RMSE terms, as opposed to standard 
deviation. 
It should be noted that, though some of the data is collected at 
northern latitudes, there was not noticeable degradation in GPS 
accuracy due to this. No particular effort was made to optimize 
flying times to maximize GNSS satellite coverage. Instead, 
flights were performed any time there was good weather. 
Despite this, there was minimal, if any, problem with poor 
PDOP. 
5.3 Forest Floor Penetration 
The project areas consisted of both prairie and forested 
mountain areas. The nature of the northern Canadian forest is 
such that there was little, if any need to “over-collect” in order 
to meet the desired ground point density targets. Adequate 
space between tree stems and minimal undergrowth meant that 
nearly all laser shots resulted in a return from the forest floor. 
5.4 Relative Match Between Adjacent Flight Lines 
In areas of rugged terrain, the target point density was low 
enough compared to the terrain feature size such that any slight 
mismatch between data from adjacent flightlines was not 
discemable. However, the prairie areas where there is little 
terrain relief, small elevation difference where adjacent flight 
lines meet are more readily noticed. Even elevation differences 
that were well within both predicted and specification values 
were easily observed in prairie area. For these areas, additional 
effort was put into refining boresight calibration so that any 
difference could be kept small enough to be unnoticeable. In 
these areas, it was not uncommon to refine calibration values 
until any height were less than 3-4 cm. 
6. CONCLUSIONS 
From the extensive use of 3 MPiA-equipped ALS50-II systems 
on the project, it can be concluded that MPiA technology is 
well suited to improving the acquisition rate of LIDAR point 
cloud data. It is clearly demonstrated that significant reductions 
in data acquisition time can be realized. 
It can also be concluded that the use of MPiA technology does 
not sacrifice data quality. Data obtained using MPiA 
technology compared favourably with data acquired over 
similar terrain and operating conditions with conventional 
single-pulse-in-air systems. 
Finally, the use of MPiA technology provides additional 
benefits that are not directly quantifiable including reduced 
flight crew fatigue. This is mainly due to reduced turbulence at 
the higher flying heights allowed when using MPiA. 
From the foregoing it is readily seen that MPiA technology 
fully realizes predicted benefits and is capable of delivering and 
even exceeding the theoretical 2:1 benefit in terms of data 
acquisition cost over attempting to acquire similar density data 
using non-MPiA technology, especially were there is any 
significant terrain relief. It can therefore be recommended as 
the preferred alternative to conventional single-pulse-in-air 
systems. 
From the standpoint of acquiring massive terrain data sets at 
medium spatial resolution, it can be concluded that flight 
planning, mission execution and data processing logistics all 
benefit from some level of optimization. Given adequate 
weather conditions, very large projects of this type are feasible, 
even with a single aircraft, although there are some additional 
benefits of multiple aircraft operations. In particular, multiple 
aircraft could be flying in the same or adjacent project block, 
187