International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
resolution satellites only 29% of the European gas pipeline
network can be monitored!
In figure-4 an overview is given of the monitoring period of
each observation (the number of days passed after the last
observation). The scheduling algorithm targets at a monitoring
period between 7 and 14 days. The dip at 9 days is caused by
the orbit pattern. By investigating more advanced scheduling
algorithms possibly a some higher effectiveness can be
obtained.
Monitoring days per observation
250.000 E.
tT |
I |
| |
200.000 Peu ies |
E |
2 Ei |
T 150.000 1-——— — : La ner + N EN ere ims aS
e i
o
o
9 100.000 BEE
c
50.000 +++... ers Je ee reir
ANI 2 5 i
0 - NRA 7 ARR N A
e e e 2 8 & 8 a S2 3 3
MESA EEE dns omefdawm a : ; =
—— no clouds no clouds, sun>15 »— clouds, sun» 95. |
Figure-4: Distribution of monitoring period per observation
In order to get an impression of the spatial distribution of this
monitoring capacity in figure-5 an overview is given of the
number of days that each element is not monitored during the
year, this means all days extending the 14 day monitoring
periods. It can be seen that dense network areas and network
trajectories in the north-south track direction are monitored
best.
áIDHAédOg*iu--t^i KAA A
Figure-5: Number of days a network element is not monitored,
green X2 days, yellow x21 days, red > 21 days.
A
[n a next step several system parameters have been varied in
order to get a feeling of the influence of the sensitiveness of the
system to these parameters. Here attention will be paid to the
availability of cloud information for the scheduling, the number
of satellites, the swath width and the pointing range.
Use of cloud information
The simulation results of the situation with clouds as shown in
table-2 and figure-4 have been generated for the case that ideal
information on the cloud situation is available for the satellite
scheduling. In case no information on the cloud situation at
time of observation is available for satellite scheduling, the
results are much weaker, as can be seen in table-3.
Table-3: Simulation results related to use of cloud information
observations total |-observ. 7-14 days | monitoring days
number . times| number times number. % full
case elements network} elements network monit.
Full cloud info 582.860 22,7| 274.909 10,7] 2.717.495 29,0%
1 hour old cloud info) 492.099 19,1| 208.716 8,1| 2.088.109 22,3%
No cloud info 344.288 13,4| 133.981 5.2| 1.334.648 14,2%
The effectiveness of the system drops from 29% to 14.2%. In
case cloud information of 1 hour old can be used the effect of
this information still is significant: 22.3%. The effect of the use
of cloud information also is shown in figure-6, where the
distribution of the number of yearly cloud free observations is
shown.
Different options and strategies for the use of cloud information
are thinkable. They are dealt with in more detail by (Algra,
2003).
Effect of cloud information
1.200
1.000 4 mes
800
600
nr of elements
400
| 200 | + eta | |
=
$e a e a +
|
0 ;
- u» e wn e AO e wn © wn ©
- + ~N N e e <r "f wn
nr of visits
| | no cloud info cloud info + cloud info 1hr old
Figure-6: Distribution of the number of cloud free observations
for different situations of cloud information.
Number of satellites
As described, with a number of 4 satellites only 29% of the
fully required monitoring capacity are obtained. Additional
simulations with constellations of 1, 2, 6 and 8 satellites have
been made. The result is shown in figure-7. With 8 satellites the
capacity increases to 49%. When also observations after 15 to
21 days are accepted the capacity would reach to 68%.
% oftotal monitoring days for all elements
70% ee A mem
|
|
|
50% Frm tes rics BB
40% up || || uperod 1521 days |
30% |
EI period 8-14 days
20%
| 10% | § 1 |
| {
| 0% - ; i
|
% total monitoring days
T
e N <r aq eo
nr of satellites
Figure-7: Monitoring capacity in % of required monitoring days
for different number of satellites
1064
Inte
Sw:
Wit
incr
the |
swa
of a
Whe
com
A ci
ofn
10kı
is a
traje
% total monitoring days
Poin
Final
track
effec
range
pipel
avoic
à COI
From
than |
% of total monitoring days
Figur