We can attribute it to the slow growth rate of RT instability during the season (Jun Sols). In fact, the few resultant bubbles from such effect are also slow in rate of decay, a process that may extend to early hours of the day. They are usually within the scale sizes that can cause modulations to VHF signals. Best of luck!
I suggest you read the paper by Sharma and Muldrew, Seasonal and longitudinal variations in the occurrence frequency of magnetospheric ionization ducts, which can be found in ReserchGate. The asymmetry in the angle between the geomagnetic pole and the the sun-earth line at night affects the occurrence frequency of ducts and spread F and I suspect it would also affect plasma drifts. If possible you could verify this by observing a longitudinal affect.