Some observations on
LEM’s post.
From my experience there is very little in aviation that is legally sound; the best course of action is “to do, and be seen to do” the correct thing for the circumstance i.e. take the safest option. With this in mind, it would be preferable for an operator to have SOPs for those abnormal circumstances where the captain has to satisfy himself with the visual cues e.g RVR below minima after OM.
Specific policy/procedures should define less restrictive visual cues in the event of a system failure; e.g to aid the crew in aligning or correcting a deviation resulting from an engine failure. In the event of an incident, it is more difficult for an individual to show that the required level of expertise existed or that correct judgement was made in the circumstances, as opposed to using a careful pre considered SOP to cover such situations – a task for management.
The visual segment maths are fine, providing all runways have closely spaced lighting. In addition, I think that you have identified a weakness in the JAR-OPS wording which was drafted before the wider use of closely spaced ‘Cat 3b’ lighting. The initial draft of subpart E required use of minimum visual segments as determined by experimental flight trials (ECAC Doc 17). I recall that for Cat 2 manual landing, this was 250 m, but as the old 400m RVR limit has been reduced to 350/300m RVR, the vis seg may now be 225 – 180 m depending on aircraft type. The first point of contact and duration with the visual references are important aspects in assessing the aircraft flight path and alignment.
Cat 3 autoland required a smaller visual segment at 50 ft (90 m), approx 3 consecutive sets of ‘lights’. However, as the smaller ‘autoland’ visual segment would not give an equivalent level of safety for a manual landing (without guidance, HUD, etc), JAR-OPS Interpretive and Explanatory Material (IEM) suggests a GA for any failure above DH in RVR less that 300 m, but the wording is loose enough to allow manual landing below DH.
A good explanation of the relationship between RVR, visual segment, and slant visual range is given in the Airbus document
Getting to grips with Category II and III Operations.
Re RVR falling below the approach ban limitations. In stable fog conditions, typical of Cat 3, any change in RVR with time is normally very small. However, for blowing snow or dust then the change could be large and indicate weather conditions to avoid.
In Cat 2, the RVR is more variable due to fog forming or clearing. Even a ‘normal’ Cat 2 decision can turn out to be wrong if cloud or fog banks blow across the runway, thus ‘legality’ before or after DH, still requires good crew judgement. This means taking the safest option, which ideally has been well thought through on the ground and thus less likely to suffer human error in an unfamiliar airborne scenario.
Finely, the thread – uptrim. Most systems, such as uptrim are in an aircraft for a good reason; if they are important to safety there should be a warning or physical limit, although in older aircraft the manufacturers relied on a crew procedure. If the reason for the equipment is not known, then avoid making assumptions about any operation if it fails.
Thus, does the 737 have a light, limit, or procedure for no uptrim or system failure?