Controlling the RoR with air on a drum roaster is very difficult. Drum roasters pull air from under and around the drum into the bean mass. The air temp around the drum can vary greatly depending on flame and ambient conditions. sometimes turning up the air will pull in cooler air, sometimes hotter for a period. At a minimum you would need a probe where the air enters at the back of the drum to measure inlet temperature to make an attempt. I'm working on setting one up this week.
Better to control the RoR with heat settings. How far in advance these changes need to be made depends on the thermodynamic properties of the roaster. The rumor is gas makes fast changes compared to electric, but my roaster is gas with thick cast iron front and back plates, and all I can say is adjusting the heat is like steering the Titanic. It requires a lot of experience with roaster and bean, and a meticulously repeating process to get it right. Sometimes it works perfectly, sometimes not.
Here is a roast I just did tonight. The blueprint roast is in the background and you can see where the heat changes occur. When I made that blueprint I was using a constant air rate of 50%. Tonights roasts had the air at 25% until 3:35. You can see the MET bump where I increased air to 50%. Usually I reduce the flame about 30s before opening the damper to avoid too much of a BT spike. So yes, I need to make an updated blueprint.
The tricky part is the last heat adjustment before 1C. Rao has posted that changes to heat should not be made within 45s on either side of 1C. So that last adjustment at 1C - 45s determines the RoR trajectory for the next 1:30 during the most dynamic part of the roast! Different coffees react very differently at 1C and to maintain a steadily declining RoR requires more art than science. There are just too many variables for my limited cognitive powers. But if my roast hygiene was disciplined throughout the pre-roast and early roast process, the variation in outcomes shrinks dramatically.
You can see on the blueprint where the last heat change was made. For some reason the RoR on this roast plateaued just a bit at 1C, then dipped too much before recovering. I wouldn't call it a crash, but it's not ideal and that dip will show up in the cup. I like to take this coffee to 400*F at <18%, but the dip sent the RoR on a path that required carrying it longer to even get to 398*.
This is why roasting darker and into 2C is so much harder than light roasts. You need to carry more heat throughout the roast to get to higher drop temperatures and still not allow a flick. If you can do it, you get unbearably sweet, chocolatey coffee with no raostiness* whatsoever.
* Freudian typo.