I made sure that if a time span contains observation with different mode but with modes regarding the same 
component the results are still returned for both. Additionally, a certain mode may not be for a long period.
So givin a mode a, it can happen that a certain component in the RTSM is not tested and result as green. This
a long standing issue. To solve this issue the RTSM should provide the mode in which the observation is perfomed. However, to mitigate the issue I am now excluding all the observations that do not have errors in a specific mode but have errors in another mode. This can still include spurious errors if such observation does not presents any RTSM error! ('luckly' it is very rare).