Can anyone explain the curve at the apparent baseline of my spectra? These are blank spectra (Water)



Hi, 

These are LC-MS spectra from a Synapt G2-si linked to Waters' Premier Acquity LC system. Running Water (blank) injections (1ul) over a 24min gradient of A: Water, B:MeOH both with 0.1% formic acid.

I am a little intrigued by the lifting spectral baseline, almost forming a boat like curve on either ends of the spectra. What could this be?

The system recently had contamination issue so we cleaned it, changed the sampling parts, some tubings and the injector valve. This improve the massive curve shaped spectra we were getting before, but we are still left with this mild curve in our mass spectrum.

Is this normal, can this be ignored, how to remedy this?