Due to lack of support by superiors, I started HPLC in the neurosurgery lab in Giessen with N2 over mobile phase pressure containers (made by myself from appropriate pipes and needle valves). Any backpressure changes changed the flows, I had to keep everything super clean to get repeatable results. Later I built in a flow regulator, that took care of this problem. As I understand this, a piston pump regulates flow, so that pressure drops out as a factor, that is apparently what Dancho is saying. Now if a faulty pump is not compensating backpressure fluctuations, such that the flow changes, the baseline does not change as sharply as in the present chromatogram, I have seen many flow change-baseline changes at the time when the pressure vessels were used.
The above looks to me as if somene/something just changed the baseline setting as someone mentioned already in the form of an attenuation change.