Like other countries that are signatories to the Chicago Convention, Australia uses different classes of airspace (classes A, tightly controlled airspace, through class G, uncontrolled airspace) for air traffic controllers to separate aircraft depending on where that airspace is located. Capital city airports use an airspace classification that is quite different to that located at small country airports. This seems obvious. But what is not so obvious to the casual observer or even large sections of the Australian aviation industry is how combining various airspace classes into a symbiotic system is supposed to work.

Whilst different classes of airspace are not created equal the outcome of applying the various classes is supposed to be exactly the same. Put simply, applying airspace classifications correctly would mean that aircraft operating at small country airports would have roughly the same risk exposure as those operating at our biggest and busiest like Melbourne and Sydney.

Clearly to achieve this outcome the airspace class used at Sydney, Australia’s busiest airport would need to include significant risk mitigating features such as radar and full separation services by air traffic controllers, features that are not necessarily employed for the class of airspace used at smaller lower risk country airports.

For high-level en-route airspace where two airliners approaching head on advance on each other at almost twice the speed of sound, Australia like other countries uses class A airspace, the most restrictive and tightly controlled class. Generally lying below class A in the mid level altitudes we use class C (Class B is not used in Australia). Class C is similar to Class A in that all airline traffic is separated and it sits below class A and extends all the way down to the ground at places like Sydney and Melbourne.

Australian airliners spend a lot of time in Classes A and C as long as they fly between the capital cities. Even trans continental flights to Perth via the remote southern route over the Great Australian Bite, are well protected by high level class A and again by class C on descent and landing.

The big downfall of Australian airspace design is at regional and smaller airports. If you were to board a flight in Sydney for Byron Bay on the north coast of NSW the flight would be subject to a justifiably high level of class C surveillance and control during the climb out of the Sydney basin. When the aircraft descends into Ballina however it is subject to less and less assistance from air traffic control the closer the aircraft gets to the airport. The aircraft will have descended from Class A (the highest level of risk alleviation) into class C, then into class E (less risk alleviation features) and finally into class G (uncontrolled airspace with very little risk mitigation).

Science and common sense tells us that the closer an aircraft is to the point of convergence and divergence of traffic (an airport) the higher the risk of collision. An airport is a lot like the intersection of two major roads, which is why we use traffic lights to separate the converging and diverging traffic. Traffic lights used appropriately are dollars well spent to save lives.

Similarly the closer to the ground an aircraft flies the risk of unintentionally impacting it becomes higher too. In fact statistically validated risk modelling by the FAA in America shows that an aircraft which is 8 nautical miles away from an airport is 4 times less likely to be involved in a mid air collision than if it were half as close. That’s to say if you half the distance an aircraft is from an airport, you quadruple the risk of collision.

So why does Australia use highly protective Class C airspace at the mid level altitudes and employ less protective Class D and G at Australia’s regional airports? In essence our airspace architecture in the regional areas has been designed upside down. It is this logically flawed design of Australia’s airspace system that I believe will be the culprit of our first major jet airliner accident with the loss of hundreds of lives.

Its flaws have been highlighted by a number of close encounters whilst on approach to regional airports over the last few years, including an alarming proportion involving airliners. The Australian Transport Safety Bureau responsible for investigating these mid air incidents has remained noticeably silent and its reports have never included any recommendations to upgrade the airspace.

What is most disquieting about the ATSB’s lack of interest in airspace design as a contributing factor to 2 fatal crashes and 3 near misses involving airliners in recent years is that the Civil Aviation Safety Authority introduced airspace changes as far back as 1995. These airspace changes would have prevented all five of these incidents and accidents but the changes have never been implemented. Why has the ATSB failed to mention this in all five incident reports and why haven’t the airspace changes been implemented?

Each of the incidents happened in airspace that was observable on radar but was designated as uncontrolled class G.  That means that Air Traffic Control was not separating the planes from each other or the ground despite being detectable by their radar system. Four of the incidents occurred in IMC (Instrument Meteorological Conditions, meaning the aircraft was in cloud and unable to navigate by visual reference to the ground) and the fifth incident involving a Qantas Boeing 737 on approach to Canberra occurred at night near mountainous terrain. None of the major carriers in Australia have been spared from these recent incidents. Virgin, Jetstar and Rex have also been involved.

In the 1995 airspace re-design, all low and mid level altitude uncontrolled class G airspace that was visible by radar, was to be upgraded to class E controlled airspace. Class E is a flexible use airspace classification that allows controllers to keep Instrument Flight Rules (IFR) planes including airliners separated from the ground and other IFR planes at all times. The introduction of class E airspace in Australia has been controversial to say the least.

Despite its widespread and successful use in many countries, both the Australian industry and government bureaucracies have resisted the 1995 airspace change with almost pathological zeal for the last 16 years. Exactly why the bureaucracy has resisted it is unclear but industry players like pilots and controllers have resisted it due to fear and a real lack of understanding of how it is supposed to work.

The model put forward in 1995 was a copy of the FAA airspace architecture used in America, a proven airspace design with countless millions of flight hours to validate its risk modelling. To introduce this design model Australia was to reclassify mid level class C airspace in the en-route environment to class E. Generally the FAA risk modelling dictates class C as appropriate only for use within 10nm of an airport but not for use in the mid level en-route environment where less restrictive class E is used. Pilots and air traffic controllers resisted these changes as it meant opening up mid level en-route airspace to VFR (Visual Flight Rules) and light aircraft who would then not necessarily be actively separated from airliners.

Many argued that to change the mid level airspace from class C to class E was to reduce the level of safety for airliners and other IFR aircraft.  Class E airspace has been demonised ever since to the point where the authorities will not even upgrade uncontrolled class G airspace to controlled class E where common sense indicated we need it most. That is, near the ground and near towered and non-towered regional airports.

Opponents of class E often claim that unknown VFR or light aircraft who may not be known to air traffic control will create an unacceptable risk to airliners despite the millions of hours airliners fly through class E in the United States every year where they encounter orders of magnitude more traffic than in Australia. The perceived risks of applying class E here as per the U.S. model are simply not supported by all available scientifically validated data.

In November 2003 Airservices introduced the replacement of Australia’s mid level class C airspace with class E and subsequently reversed it citing reduced levels of safety in a class E risk study. The Airspace Risk Model (ARM) used by Airservices to justify the reversal has been resolutely discredited. Professor Terry O’Neil of the ANU, an expert in actuary and applied statistics who was commissioned by Airservices to review the Airservices study found the methodology they used was flawed and the data was not subjected to validation, a finding upheld by other statistics experts such as Dr Bob Hall, president of the Australian Sport Aircraft Confederation.

Pilots and controllers in Australia have been highly critical of class E largely due to the perceived decrease in safety of it replacing the higher more tightly controlled class C. And by extension resist replacing uncontrolled class G with controlled E. This misunderstanding of the risks associated with class E is the stimulus of resistance to the US airspace architecture in Australia. The reality is that class C is no safer than class E. Nor is class A more safe than class G. The class of airspace does not determine the level of safety, rather the density of traffic among other things determines the class of airspace required to mitigate the varying levels of risk from departure to arrival.

Captain Ian Woods, the president of the Qantas pilots union AIPA, in his June 2007 submission to the federal government’s Aviation Policy Statement, reveal the level of misunderstanding the Australian pilot fraternity has in regard to airspace classifications and how they are supposed to work.

“The Association’s position is that Class E airspace allows potentially unknown or unverified VFR traffic to be in conflict with IFR passenger transport operations…”

So does uncontrolled class G. But interestingly the 5 near misses and accidents that have occurred in class G since this statement was made involved airliners and other IFR aircraft only and no VFR aircraft.

“Class G airspace implies a lower risk which in turn, requires a lower level of ATS [Air Traffic Services]…”

This is a remarkably misguided understanding, as the class of airspace does not imply a lower risk. The level of risk implies which class of airspace is required to mitigate that risk. Unfortunately this lack of understanding permeates throughout the industry.

Today, eight years after the first meaningful attempt to implement the US airspace model in Australia was wound back, we still have aircraft descending into airports like Hobart and Alice Springs where airplanes receive greater risk mitigation services from air traffic control at 50 miles away than they do at 5 miles.

Arguably however the most astonishing aspect of our regional airport airspace design is that aircraft operating within 30 to 40 nautical miles of a regional airport with a control tower are forced to participate in radio reporting of position and altitude for separation in the same way it was done prior to the Second World War and the advent of radar. Once again, the adoption of the FAA model would have done away with this draconian procedure had it been implemented.

So how does Australia update its airspace architecture so aircraft get greater protection rather than less the closer they are to regional airports? Whatever the way to change it must include strong leadership within pilot and air traffic control unions that demand a risk based approach to airspace design tied to scientifically validated risk and cost benefit analysis. So far not a single peer-reviewed airspace risk study in Australia has been conducted using validated data. Additionally the Office of Airspace Regulation within CASA and Airservices Australia leadership must stop dithering on what has been government policy now for 20 years to introduce the proven US airspace system.

 About the author:
Matthew Bowden is a Boeing 737 airline pilot. Has has flown extensively in the United States, Asia and Australia.