I am working on a trajectory optimisation program for commercial, large aircraft, in which I want to incorporate realistic air traffic control regulations. One of the constraints I want to incorporate is that of ATC requiring step-climb cruise at specific flight levels.

Specifically, the hemispheric rule, though not universal, is a good example of how air traffic is managed by enforcing that westbound flights take even flight levels (FL020, FL040, etc.), whereas eastbound flights take odd flight levels (FL010, FL030, etc.).

So, aircraft should fly at one of these flight levels during cruise, when in level flight, thus resulting in a step-climb cruise, as opposed to a continuous-climb cruise profile. E.g., if an aircraft flying east starts its cruise at FL330, it would want to increase its altitude at some point for increased fuel efficiency. If cleared by ATC, it should climb to FL350 (maybe even FL370, if feasible), and so on.

My question is; are there any restrictions on how long the climb part between each of the level portions can take, at maximum (in terms of distance, time, or some other measure), and if so, what is the specific FAR/JAR or CS25 chapter that mentions it?

One could imagine that if such regulations aren't imposed, the climb segment in between level steps could take up so much time, theoretically, that effectively the cruise would still be continuous-climb. Although this is probably not too much of a problem in real life, it is a problem in my trajectory optimisation program. I.e., without proper bounds on the duration/length of this in-between climb leg, the solution always tends towards a continuous-climb solution by making the level flight portion short and the climb portion large.


3 Answers 3


The FAA's IFR handbook sheds some light on what they expect (page 2-48)

If the pilot receives the term “climb at pilot’s discretion” in the altitude information of an ATC clearance, it means that the pilot has the option to start a climb when they desire and are authorized to climb at any rate, and to temporarily level off at any intermediate altitude as desired, although once you vacate an altitude, you may not return to that altitude. When ATC has not used the term nor imposed any climb restrictions, pilots should climb promptly on acknowledgment of the clearance. Climb at an optimum rate consistent with the operating characteristics of the aircraft to 1,000 feet below the assigned altitude, and then attempt to climb at a rate of between 500 and 1,500 fpm until the assigned altitude is reached.

Its somewhat contradictory, first they say you can climb at any rate you want but then they say in the absence of any specific instructions you should climb at an "optimum" rate to within 1000 feet of your assigned altitude and then climb between 500fpm and 1500fpm.

This is specifically spelled out in the AIM with more or less the same language

When ATC has not used the term “AT PILOT’S DISCRETION” nor imposed any climb or descent restrictions, pilots should initiate climb or descent promptly on acknowledgement of the clearance.Descend or climb at an optimum rate consistent with the operating characteristics of the aircraft to1,000 feet above or below the assigned altitude, and then attempt to descend or climb at a rate of between 500 and 1,500 fpm until the assigned altitude is reached. If at anytime the pilot is unable to climb or descend at a rate of at least 500 feet a minute, advise ATC.

Interestingly this adds the note that if you can not climb at 500 f/min then you should alert ATC. I would take this to be a minimum climb rate.

There is at least one permitted case of the drift climb that I know of. The Concorde operated above all other commercial traffic and aside from the SR-71 there really was no one else up where they were. As such once they passed a specific floor they were allowed to drift up freely as they burned fuel. This was, as far as I know unique to their operation and specifically permitted. They had a lot of other specific addendums due to their unique design.

  • $\begingroup$ That is a very clear comment, thanks. One thing that seems especially contradictory to me in the IFR handbook is that you are allowed to temporarily level off, thereby lowering your RoC to 0 fpm. This doesn't agree with the implicit minimum climb rate of 500 fpm. Then again, I reckon real-life applications wouldn't benefit from such strict regulations. $\endgroup$
    – Sam
    Dec 17, 2019 at 1:21
  • $\begingroup$ One of the mental skills a controller has to have is the ability to anticipate whether conflicts could arise as a result of the instructions he/she gives to traffic, and this would include the parameters around a climb instruction. So the type of clearance or instruction given would take into account the potential for conflicts using what you might call rules of thumb that include ample safety fudge factors. $\endgroup$
    – John K
    Dec 17, 2019 at 1:33
  • $\begingroup$ @Sam, would you please quote the full part about the pilot being allowed to temporarily level off during an IFR climb? This needs context, because generally no, if you are climbing to an assigned altitude you cannot arbitrarily just level off. $\endgroup$ Dec 17, 2019 at 16:13
  • $\begingroup$ @MichaelHall I think he is referencing the line "and to temporarily level off at any intermediate altitude as desired" from the IFR manual (in my above quoted passage) $\endgroup$
    – Dave
    Dec 17, 2019 at 16:54
  • $\begingroup$ @Dave - is that a direct quote then? (I don't see it on page 2-28 of the linked document.) $\endgroup$ Dec 17, 2019 at 17:15

To address what's discussed in comments...

The East/West rule governs cruise altitude. When you're climbing to or descending from your cruise altitude, those rules don't apply. ATC obviously has to consider opposite direction traffic when they contemplate where/when to issue those climb & descent clearances, but for the pilot, that's just a matter of a few minutes before the clearance gets issued.

So let's say I'm on a westbound flight. I know that I'll be cruising at FL300, 320, 340, 360, 380, or 400 (since my aircraft can't go above 410 and any flight that's long enough to think about a step climb won't be cruising at an altitude as low as 280 or below). Typically, there will be some optimum altitude that I'll climb to, given the winds and my initial weight. Let's say that the FMC says that my ideal altitude for "right now" is 34,300'... I can't cruise there, so I'll climb to FL340. Time passes, and that ideal cruise altitude increases. After a while, I'll be better off at FL360. That point doesn't occur when the ideal passes FL350; you're less efficient 1000' above the optimum altitude than you are 1000' below it. Let's say, for simple math, that the break even point is 500' above optimum is the same burn (per mile) as 1500' below. So when my FMC optimum altitude is 35,500, I'll request FL 360. And when ATC has a clear path -- no oncoming traffic at FL 350 -- they'll clear me up to FL360. And a couple of hours later, that process will repeat when I'm ready to step up to FL380.

As far as a model, if you assume 500 FPM average climb from 340 to 360 (or any other step), that's four minutes of climbing for each step, and that's probably a pretty reasonable estimate. In most cases, the aircraft will climb faster than that, but when those steps happen a couple of hours apart, is the difference between 2.0 or 3.1 minutes vs 4 minutes really significant for the model? If it is, then you'll need far more detailed performance data for specific aircraft you're modeling than any rule of thumb can give you. But, I'd suspect that for most purposes, assuming 4 minutes (or somewhat less) for the step climb to the next even (or odd, in the case of eastbound flights) is probably a workable assumption that will give you a reasonable approximation of what aircraft typically burn in such a scenario.

  • $\begingroup$ Thanks, indeed this summarises and clarifies some of the points discussed in the comments. Especially your last paragraph is what I intended to get out of this question; the RoC requirement poses an upper bound on the time to climb to a new FL. Because I got this answer from Dave's response first, I left his as the accepted answer. You're right, 2 or 4 minutes makes no difference to the model, but without the upper bound of 4 minutes, the optimisation algorithm pushes the climb towards solutions with very small rates of climb, thereby not modelling realistic ATC-compliant trajectories. $\endgroup$
    – Sam
    Dec 18, 2019 at 7:38

ATC does not specifically require a step climb. The only reason they would hold you at an altitude lower that your intended final cruise altitude would be conflicting traffic, or airspace constraints. For example, a departure controller will clear you up to the top of their sector, but you will need to contact a center controller to be cleared higher.

The only real requirement related to duration of climb is that ATC requires you to report if unable to maintain your climb at a minimum of 500ft/minute. If you report unable to meet this they may have you level off, but there is no minimum time required at any intermediate level offs before reaching final cruise altitude.

My understanding is that some airliners will step climb for reasons of fuel efficiency at certain weights. (and/or because of climb rate) However, step climbing for this reason wouldn't be regulatory or imposed by ATC, it would be an airline policy, or requested by the pilot.

  • $\begingroup$ Okay, then my understanding of the hemispheric rule is probably somehow off. I found that under visual (VFR) or instrument flight rules (IFR), one has to stick to certain flight levels according to the aircraft's heading, whenever in controlled airspace. Isn't step climbing a logical consequence of this rule, then? And if not, could you explain how these rules (hemispheric rule and allowing continuous climb) aren't contradictory? $\endgroup$
    – Sam
    Dec 16, 2019 at 23:32
  • $\begingroup$ @Sam - FYI, it is actually the aircraft's magnetic course that determines the hemispheric rules. (in case heading is adjusted for wind and you are crabbing 355 but drifting 005...) Your understanding of the hemispheric rules seems correct, however, I'm not sure how this relates to step climbing. If I file IFR westbound at FL 240, why would ATC need me to level off on the way up if I can maintain 500FPM all the way up to 240? (presuming traffic is no factor) $\endgroup$ Dec 16, 2019 at 23:39
  • $\begingroup$ So basically the climb duration is related only to the informal requirement of maintaining a minimum RoC at 500 fpm. Isn't it strange that those climb requirements aren't concretised in a specific regulation? As to your last point, it seems to me that the hemispheric rule is related to step climbing, because the level "steps" at which the aircraft flies are defined by those FLs defined by that rule. Indeed, if you can maintain 500 fpm all the way up, there's no need to level off in between; the only thing limiting a continuous climb would be not meeting a minimum RoC, it seems. $\endgroup$
    – Sam
    Dec 16, 2019 at 23:56
  • $\begingroup$ FYI, the "requirement" to report if unable to meet 500FPM is actually in the AIM, so it isn't regulatory in the sense that the CFRs are. But I'm still not quite understanding how you could think that step climbing is a logical consequence of the hemispherical cruising rules. What made you first think that you would need to level off for some random period of time at 2000' intervals? I assume you have flown in an airplane at some point in your life? Did it do this? $\endgroup$ Dec 17, 2019 at 0:05
  • $\begingroup$ Ah I now see where I didn't make myself clear. I didn't think you'd have to stop at 2000' intervals when climbing. Rather, I assumed that due to the hemispheric rule, you wouldn't be allowed to climb gradually (and slowly) during cruise. From a trajectory optimisation point of view, this gradual climbing as you burn fuel would be most fuel-efficient. However, it seems that this is inhibited by the requirement to fly at a specific flight level. As for my background; I haven't piloted, but have flown in an airplane as a passenger and I study aeronautical engineering. $\endgroup$
    – Sam
    Dec 17, 2019 at 0:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .