OB–stars have the highest luminosities and strongest stellar winds of all stars , which enables them to interact strongly with their surrounding ISM , thus creating bow shocks .
These offer us an ideal opportunity to learn more about the ISM .
They were first detected and analysed around runaway OB–stars using the IRAS allsky survey by van Buren et al .
( [ 1995 ] ) .
Using the geometry of such bow shocks information concerning the ISM density and its fluctuations can be gained from such infrared observations .
As to help to improve the bow shock models , additional observations at other wavelengths , e.g .
H \alpha , are most welcome .
However due to their low velocity these bow shocks have a size of \sim 1 ° , and could only be observed as a whole with great difficulties .
In the light of the new H \alpha allsky surveys ( SHASSA/VTSS ) this is no problem any more . We developed different methods to detect bow shocks , e.g .
the improved determination of their symmetry axis with radial distance profiles .
Using two H \alpha –allsky surveys ( SHASSA/VTSS ) , we searched for bow shocks and compared the different methods .
From our sample we conclude , that the correlation between the direction of both proper motion and the symmetry axis determined with radial distance profile is the most promising detection method . We found eight bow shocks around HD 17505 , HD 24430 , HD 48099 , HD 57061 , HD 92206 , HD 135240 , HD 149757 , and HD 158186 from 37 candidates taken from van Buren et al .
( [ 1995 ] ) .
Additionally to the traditional determination of ISM parameters using the standoff distance of the bow shock , another approach was chosen , using the thickness of the bow–shock layer .
Both methods lead to the same results , yielding densities ( \sim 1 cm ^ { -3 } ) and the maximal temperatures ( \sim 10 ^ { 4 } K ) , that fit well to the up–to–date picture of the Warm Ionised Medium .