Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 Inte 
    
es of the Photogrammetry, Remote 
  
International Archi 
    
  
  
  
  
   
    
    
     
    
  
   
   
  
      
     
   
  
  
depicted in Figure 3. 
1008 
Table 3. Information reported for the slick from the oil rig From the 61 checked oil slicks 34 (56%) could be verified and mui 
27 (44%) were false positives. eta 
The 
Table 5. Results of the satellite/aircraft campaign 2003 con 
cam 
55928' 18" N sate 
5906'00" E stati 
The 
Sea 
Confidence fron 
Sen: 
The aircraft was at the reported position 2 hours and 5 minutes Rel 
after the satellite overpass. The reported length, width and area Scar 
of the slick are different. The reason for the area difference can the ) 
be the modification (spreading) of the oil slick extent (wind, The 
current) during the period between the satellite overpass and the F Leu 
verification. The oil volume can not be estimated from SAR avail 
data since the oil layer thickness can not be obtained Vers 
In total 40 images were acquired over the North Sea and the Nort 
Baltic Sea (Table 4). 
{ : Results of the satellite/aircraft 
D campaign 2003 7 
of Germany Van 
NM 6 
e s . 
V $ Es 
19:24 UTC Ss 8 
Figure 3. Results of the satellite/aircraft campaign 2003 i 
The verification results separated into the analysis of : 
RADARSAT-1 and ENVISAT images are shown in Table 6 and 1 
7. 
Table 6. Verification results RADARSAT-1 
Figure 2. oil releasing oil rig detected on 15 July 2003 by RADARSAT-1 
RADARSAT-1 at 17:19 UTC (large image) and by the German a-priori ich Medium 
aerial surveillance (SLAR) at 19:24 UTC (Radarsat-1 data  Confidence i een: Figu 
copyright Canadian Space Agency/agence spataile canadienne ee Ta 
2003, processed and distributed by KSAT) PA de ox 
Table 4. Acquired images of the North Sea and Baltic Sea 3 an : Le 
fable 7. Verification results ENVISAT 7 
a re PSE Ar « 
confidence i 
|g 30 
In total 97 possible oil slicks were reported from satellite : 3 
observation. The average time difference between the satellite Table 6 (RADARSAT-1) shows that of the 4 high confidence | 20 
overpass and the time when thé aircraft was at the reported oil slicks checked all Were verified as being oil. Only ] of 12 
position was 2 hours and 15 minutes. Of the 97 possible oil medium confidence slicks were false positives but 3 of 21 low 10 
slicks 58 were observed in the North Sea and 37 in the Baltic confidence slicks were found to be false positives. Of the 2 high : 
Sea. The verification is classified in spills which couldn't be confidence slicks detected in ENVISAT images and checked 
confirmed by the aircraft crew, spills which were tonfimncd, — Only | Was found to be oil. For the medium confidence slicks 
spills which couldn't be checked (e.g. out of range), reported checked 3 of 4 slicks were found to be false positives. For low 
spills when no aircraft was airborne (e.g. due to technical confident slicks 9 of 18 were found to be false poste Figur 
problems or bad weather) and some spills which were checked Comparing the analysis results from RADARSAT-I an | 
by a Dutch surveillance aircraft (NL) and the Danish ENVISAT show that analysis of RADARSAT-1 images seems 
authorities (DK) (Table 5). The location of the slicks is to have produced more reliable oil slick detection. One reason Only : 
could be that the operators analysing the satellite imagery have EEZ il