View Full Version : decision altitude (height)

23rd Jan 2005, 00:43
Hello all,
my question refers mainly to a CAT I ILS with a DH of 200'.

my question is the following:

is DA the altitude at which you make the decision to either continue visually or execute the missed approach (thus implying that if you make the decision to go missed, you WILL momentarily descend somewhat below DA if you account the reaction time, engine acceleration time and aircraft vertical acceleration time) or is it a HARD STOP altitude which you should never go below unless you have the required visual reference?

I have always been of the opinion that you cannot go not even 1 foot below DA, and therefore you should take whatever action necessary to initiate the go-around in order to maintain the aircraft at or above DA (in other words start adding power and rotating to GA pitch some X ft above DA).

However after reading several discussions on other forums (here's one for example (http://www.propilot.com/doc/bbs/messages//8323.html) ) and talking to other instructors I am noticing that many pilots have a different view on this.
They believe you can reach DA, make your decision, and that it is "understood" that you will descend somewhat below DA.

One of the arguments I heard was that if its called "decision altitude" it means thats when you make the decision. However, I don't think this argument holds, because if we compare this to a V1 scenario, "decision speed", thats not really the time you start making decisions. Thats the point where the decision has already been made for you.

I was wondering what was your input on this, especially when it comes to airline SOPs. If the power advanced prior to, or at DA?
What about pitch rotation?
Whould the answer be different if the autopilot was shooting the approach? (I've never flown with an advanced AP/FD/AT so I don't know if the autopilot would automatically execute a GA at DA or if the pilot needs to "manually" tell it to execute TOGA thrust and pitch attitude).

In most airline SOPs, is it "understood" that you will descend somewhat below DH?

If not, how much above DA do you generally start advancing the throttles?

I understand the amount of "driftdown" from the time the throttles are advanced to the time the aircarft starts climbing will be different from say, an older 747 with older engines and slower spool-up times having a momentum of over half-million pounds descending at 700/800fpm flying at Vref, to a C-172 at 450fpm shooting the approach at over twice Vs1 that can afford to stop the descent rather abruptly even without immediately adding power due to the higher stall margin. However, its also true that the airlines have the advantage of a 2 person crew, where the PNF is looking outside when approaching DA. In a single-pilot operation, if the pilot waits until he hits DA before looking out, then has to re-transition to his intruments and execute the go-around, there will also be some degree of altitude loss.

All inputs are appreciated.


PS. By the way, when flying in the US with a sensitive altimeter, an error of +/-75ft is allowed, therefore when your altimeter reads 200ft yo may well be only 125ft above TDZE...which is not necessarily AGL. Not the time I'd like to risk getting any lower.

23rd Jan 2005, 00:49
MDA is dead stop altitude with no margin.

DA can be breached provided you have initiated the go-around before you reach it (ie. you can sink below DA as long as the nose is up and power is applied)

HG :ok:

As for the altimeter error. As far as I am aware, that is accounted for when your DA is calculated.
It is basically aerodrome altitude + allowance (this is category dependent)

I think Cat C + D is about +150\' (but don\'t quote me)

23rd Jan 2005, 04:08
In Cat 1 approach, the DA is based on a barometric altitude. Look at the approach plates, the DA is runway elevation plus 200'. this equates to approx 200' from the ground, taking into account the terrain in the approach it may be higher or lower than 200' on rad alt but never lower than 200' from the runway.

It is an alt that at which you decide to continue for a landing or go around to execute a missed approach. There fore you can desend below, as you say for reaction time etc,,, for example heavies in a cat II approach allow for the wheels to touch the runway, but by this stage you have set G/A power, just the mommentum taking you below.

ILS approaches are precison approaches so the DA reflects this and allows you to desend below, not to sneak below for a better look, just the time it takes to pitch up and set power etc.

VOR or NDB approaches are non precision, therefore have the MDA and it is a hard deck, do not go below. Many operators will set the descision call 50' above the mda, this means you can perform a constant rate of desent method, ie pretend its an ils, and work out your heights (in advance) from the threshold, this works better than dragging the A/C in along the MDA then deciding at the MAP cutting the juice and dropping on the runway. You decide 50 above and can execute a GA without fear of breaking minimums...

the two crew works with PNF calling the heights, IE one hundred to go , or SOP equivialant, then decide. the idea being that he/she is looking in at speed, alt, engine parameters etc. But the roles can be reversed in Low Vis Proceedures, whereby the PF's head is looking in, PNF decides at minimums, calling either contact or G/A.... dependig on SOP he may continue for landing by taking control or calling g/a in which case the PF will execute,

23rd Jan 2005, 14:35
A precision approach (Cat 1) has a predetermined safety allowance below DH for a GA commenced at or above DH. For Cat 1 this includes altimetry and aircraft height loss. Some crew procedures use a call of “15 (ft) above” so that the PF can respond immediately, with no response or a “GA” call, a GA is commenced at or above DH.

The MOC for a non-precision approach should not be less than 75 m 246 ft or 295 ft no FAF (PANS-OPS), or 250 ft (TERPS) and the obstacle clearance assume no flight below MDA. However, due to the method of procedure construction the obstacle clearance for TERPS is slightly larger than PANS-OPS thus some TERPS based operators have been granted dispensation to use MDA as DH.
In Europe studies have shown that the slightly increased risk of obstacle collision by using PANS-OPS MDA as DH would be offset by the overall increase in safety afforded by flying a continuous descent with GA at MDA. Plans for MDA=DH are being discussed by JAA and ICAO.

See the draft JAA NPA-OPS 20 (http://www.jaa.nl/catalogue/npas.html) relating to the introduction of continuous descent approaches.
“The use of the MDH as a DH is a consequence of the method of not allowing level flight at MDH (or multiple step-downs during non-precision approaches). This method is widely used among major European, commercial operators and more than half of those operators do not apply an add-on to the DH. It was however felt to be prudent to remind the recipients of this NPA that the application of MDH as a DH is a matter, which needs to be evaluated and which may need to be discussed between the operator and his Authority. This matter is currently reviewed by ICAO OCP and OPSP.”

23rd Jan 2005, 15:27
PS. By the way, when flying in the US with a sensitive altimeter, an error of +/-75ft is allowed, therefore when your altimeter reads 200ft yo may well be only 125ft above TDZE...which is not necessarily AGL. Not the time I'd like to risk getting any lower.

Yes but don't forget that the glideslope is "safe" on a precision approach. If your altimeter overreads by 75 ft and you're still on the glideslope, your go around may be late but you won't hit an obstacle. By contrast, if you're 75 ft low on a non-precision approach, that comes straight out of the 250 ft obstacle clearance.

23rd Jan 2005, 17:21
Good point!

There is always discussion going on. See also a thread from last year:


25th Jan 2005, 05:54
I guess then i must have it wrong.:)

Thanks for the correction:ok:

So I understand DH DOES provide provisions for a temporary descent below DH.

I'm just curious how much altitude below DH would an airliner go if the GA was initiated right at DH with stardard pitch rate and engine acceleration?

Thanks again,


25th Jan 2005, 07:45

25th Jan 2005, 08:46
palgia I think that there is a simpler version of the equation shown below, but I cannot locate it at this time, so the JAR-AWO reference will have to do. The equation relates to the minimum approach break off height (MABH) in Cat 3 operations, which I think approximates to the height loss during a GA or a value slightly greater than the actual height loss … you did ask!!!

Ref JAW-AWO 313
Ah = 0.155 (VS squared) + (0.0017 (square root W) + 3.2) VS
Ah = MABH (ft)
Vs = rate of descent (ft/sec) at MABH in still air using the approach speed scheduled for 90% of W
W = maximum landing weight (lb)

Sprite is far more practical.

25th Jan 2005, 15:49
In Canada, when executing a missed approach from a non-precision CDA approach you are permitted to descend up to 50 feet below the MDA.

I don't know if this is a general rule yet but it has been approved for air carriers who use constant descent angle to be at MDA at the visual reference point required to complete a landing in a normal configuration.

Just make sure that you don't start any lateral maneuvers until after the MAP!

There are still a lot of instances where it is far more reliable to level and stabalize at MDA well befor the MAP. Mountainous terrain for example where local conditions can vary from one end of the runway to the other and a circling approach with lots of options. (Local knowlege may be required.)


28th Jan 2005, 01:09
I believe a rejection at V1 allows for a reaction time of 2 seconds.

28th Jan 2005, 06:57
OK, its been a while and I do not have ICAO PANS OPS 8168 to hand....

However, for an ILS approach, terrain/obstacle separation is provided by a complex series of sloping surfaces (Obstacle Assessment Surfaces OAS) extending out from the surface (touch down area of the runway). Quite simply, if nothing penetrates these surfaces (or the ICAO Annex 14 surfaces) then operations to Cat III are possible so long as the ILS installation, lighting, surface signs, LVP protection etc are in place and to appropriate criteria to support such ops.

In the case of an ILS to Cat 1, with no terrain or other obstacle to affect, then the minimum (pilot/operator may require a heigher DH) decision height would be 200ft above the threshold. Normally, for a Cat 1 ILS, the controlling obstacle would be a B747 SP (tail height 20.06 metres) holding at the normal Cat 1 holding point at the RWY threshold at 45 degrees to the RWY centre-line (to get the tail as near to the RWY as possible). If a particular airport has taxiways other than just at the threshold, the procedure designer will have to check that a 747SP holding at these does not penetrate the OAS surfaces. The A380 may change this!

Remember, every approach may result in a missed approach, so obstacles/terrain have also to be considered for this event. If there are obstacles affecting the missed approach, then the decision height would have to be increased so that start of climb commences earlier on the approach to clear this (these) obstacle(s).

In terms of practical clearance of obstacles, it is assummed that the aircraft will remain within full scale deflection of the LLZ and GP. This will keep the acft safely within the OAS "funnel".

At decision height, well, it is quite simple. The assumed height loss margin, different for each acft category A thru' E (using a pressure altimeter for CAT 1 ILS so extra height factored in to cater for altimeter errors) is simply added to the obstacle to give an Obstacle Clearance Height (OCH) or altitude (OCA) and converted to DH/DA (by the operator)with the proviso that the DH will not, in the case of CAT 1 ILS (pressure altimeter) be less than 200ft.

For acft category A (light acft) versus category D (heavy acft) the height loss margin due to inertia is less. On an approach chart produced by the procedure designer you might see the following

Acft CAT A OCH 190
Acft CAT B OCH 199
Acft CAT C OCH 209
Acft CAT D OCH 227

The DH itself calculated by the operator. DH could equal OCH, so long as it is not less than 200ft of course, but in practice many operators/administrations may add a buffer to this to cater for acft type/performance/crew experience etc.

At DH then, decide! All you have between you and the obstacle is the assumed height loss margin! And don't forget, although buffers to allow for pilot reaction time and acft reaction time to commence climb are calculated and factored into the OCH, if the controlling obstacle for the approach is in the missed approach area you need to start climbing! Obstacle clearance for the mapt case (climb gradient 2.5%) is only 98ft for the initial mapt phase! But that is another story.....