ILS installations have a critical area extending some distance back from the glideslope antenna (primarily due to most glideslope transmitters’ needing to bounce one of their two beams off the ground in this area); objects or vehicles within this critical area can cause large variations in glideslope transmissions, potentially resulting in approaching aircraft being guided dangerously low above terrain and/or other obstacles. A second critical area exists along the runway itself, where the presence of objects or vehicles can distort or scatter the the localiser beam, potentially causing dangerous lateral-guidance anomalies for approaching aircraft.1 This means that the ILS critical areas must be kept clear and unobstructed at all costs when an aircraft is performing an ILS approach, especially if visibility is low enough for aircraft to be reliant on the ILS all the way down to the runway (which simultaneously worsens the potential consequences of ILS-signal interference, as there is much less margin for error close to the ground, and makes guidance anomalies less likely to be detected immediately, as the pilots have little or no external visual reference with which to crosscheck the received ILS signals, and - unlike with higher-visibility approaches - will generally not be controlling the aircraft themselves, but, rather, monitoring the autopilot as it flies the aircraft down [human reaction times being far too long, and human control inputs far too rough and imprecise, to safely fly the aircraft all the way down to the runway without actual or simulated external reference]).
This raises the question of how big an object in the localiser or glideslope critical area has to be in order to cause significant distortion or interference to the ILS signals. Obviously, tiny objects such as bumblebees or snowflakes are not large enough to have a significant impact (otherwise, most ILSs would be near-permanently useless due to insect and precipitation interference); equally-obviously, large vehicles like snowplows and eighteen-wheelers and 777s are (otherwise, there would be no need to restrict operations of these vehicles in the ILS critical areas,2 or to restrict ILS approaches while these vehicles are present within one of the critical areas). But just how small does an object have to get before it no longer poses a threat - would a deer running across the departure end of the runway cause problems,3 or would it only be large enough to cause problems if it were in the glideslope critical area,4 or not even there? What about an eagle (no, not this one) perched on one of the glideslope (or localiser!) antennae, or a robin flying around them?6
1: The localiser critical area is separate from the glideslope critical area because the localiser antenna is (usually) located along the extended centreline beyond the departure (far) end of the runway (allowing it to provide lateral guidance all the way down through touchdown and rollout), while the glideslope antennae are located alongside the approach (near) end (as we are trying to guide the aircraft to a touchdown in the first portion of the runway, not the last).
2: Aside, of course, from the restrictions inherent to all vehicle operations near active runways, ILS or no.
3: Granted, a deer running across any part of an active runway would almost certainly cause problems (likely in multiple ways), but let’s restrict ourselves to discussing its impact (if any) on that runway’s ILS signals.
4: Localisers transmit in the 108-112-MHz band, near the middle of the VHF range and sandwiched between the FM broadcast band (88-108 MHz) and the VOR band (112-118 MHz), while glideslope transmitters operate in a much higher frequency band (328.6-335.4 MHz, in the lower portion of the UHF range and occupying a slot dividing the 225-328.6-MHz and 335.4-400-MHz bands used for military-aircraft radiocommunications); hence, the critical size for interference with glideslope signals should be about a third of the critical size for localiser signals.5
5: On the plus side (for the glideslope), this also means that the optimal antenna size for a glideslope transmitter is approximately one-third of that for a localiser transmitter, allowing the former to use smaller, cheaper, lighter, easier-to-handle-and-support-and-replace antennae than the latter.
6: I am well aware that birds and aircraft do not mix; however, as with the deer making an uncleared runway crossing, my concern here is limited to their potential effects on ILS transmissions.