When we read a poll without allowing for the Media Stimulus at/of the moment, we misunderstand the poll's ultimate impact. All polls are but a measure of opinion at a particular moment in time. A failure to allow for the historical influences of "that moment in time," and we fail to understand the ultimate impact of said poll.

Most recent Gallup Poll fineds that 52% of those surveyed believe Trump should be removed from office while 46% say "no" to that issue.  Understand that Gallup gets to "52%"  because they asked 8% more Democrats than Republicans.  In other words,  minus the 8% weighted survey,  you have a one to one poll,  straight up,  and only 44% polling for impeachment/removal.  You should know that if you asked more Democrats than Conservative,  you may get a poll that recommends impeachment/removal, and that is exactly what has happened.

In June,  this same poll,  showed 45% for impeachment/removal and 53%  against impeachment/removal.  Again,  the 8%  marker still applies,  giving us only 37% of the population in favor of removal.

If you factor in Media Stimulus at the moment,  the June percentages were in response to the notion that the Mueller Investigation had failed to find an impeachable offence,  while the October stimulus was all about the whistleblower, his complaint coupled with little or no time for a qualified defense versus the Complaint.  Expect to see a reversal in these percentages if the whistleblower is abandoned for a new and different strategy.

Conclusion:  the October poll versus the June poll does not demonstrate a trend line, but rather a differing response to different historical issues presented by the media.  If Gallup was measuring the publics' view of impeachment as a result of the Mueller Report in BOTH polls,  then "yes,"  you have a trend line.  Otherwise,  all you have is a public response to new stimulus AT THE MOMENT excluding any sort of trend.