Preprint of a paper published in
Classical and Quantum Gravity,
vol. 20, No 17 (7 September 2003), S665–S676


 (1) 


 (4) 
 (5) 

 (7) 

 (9) 
 (10) 
 (11) 
 (12) 
 (13) 
 (14) 
 (15) 
 (16) 

 (18) 

 (21) 

 (24) 
 (25) 
 (26) 
 (27) 
 (28) 

 (30) 
 (31) 
 (32) 
 (33) 
 (34) 
 (35) 
 (36) 
Figure 1: Quality of the EXPLORER data. The xaxis gives the number of the 11hour block of data from the 13 day data run and the yaxis gives the corresponding probability values of the KS statistic. 
As we do not have a detection of a gravitationalwave signal we can make a statement about the upper bound for the gravitationalwave amplitude. To do this we take our strongest candidate of signaltonoise ratio d_{o} and we suppose that it resulted from a gravitationalwave signal. Then, using formula (25), we calculate the signaltonoise d_{ul} of the gravitationalwave signal so that there is 1% probability that it crosses the threshold F_{o} corresponding to d_{o}, where F_{o} = 2 + (1/2)d^{2}_{o}. The d_{ul} is the desired 99% confidence upper bound. For d_{o} = 8.2, which corresponds to the signaltonoise ratio of our strongest candidate, we find that d_{ul} = 5.9. For the EXPLORER detector this corresponds to the dimensionless amplitude of the gravitationalwave signal equal to 2×10^{23}. Thus we have the following conclusion:
In the frequency band from 921.00 Hz to 921.76 Hz and for signals coming from any sky direction the dimensionless amplitude of the gravitationalwave signal from a continuous source is less than 2×10^{23} with 99% confidence. Our analysis has been done using two days of data. We note that the upper bound will decrease as length of the data analyzed increases; d_{ul} is proportional to the inverse of the square root of the observation time T_{o}.