airport wildlife risk modelling

Airport Wildlife Risk Modelling


7 Steps to Airport Wildlife Risk Management

This is part 4 of a 7 part series exploring the use of the ISO31000 risk management framework in Airport Wildlife Risk Management; part 1 was “Frameworks and Consultation”, part 2 was “Setting the Airport Wildlife Hazard Scene” and part 3 was “Lining Up the Usual (Wildlife) Suspects

So, now you are now equipped with a list of species you may want or need to control. To get that flow I mentioned my last article, you will need to set up a framework for analysis first. The goal of this framework is to, in the words of ISO31000, “comprehend the nature of risk and to determine the level of risk.”

In Need of a Model

Luckily, there already exists a number of approaches to this very problem. The one I’m most familiar with is the Paton Bird Risk Assessment Model developed by the University of Adelaide and Adelaide Airport.

Overall, the Paton model is based on a consequence-probability matrix but beneath that it uses relatively easy-to-quantify factors that go into the mix to produce a final score.

The consequence side of the equation is made up of 3 factors which are multiplied together to produce a consequence score. This score is equated to 1 of 6 consequence categories.

The 3 factors are:

  • Body Mass - Broken up into 6 categories ranging from less than 20 grams to greater than 5 kilograms, these categories are given a score ranging from 1 through to 32.
  • Flocking Behaviour - 3 levels of flocking behaviour with scores assigned from 1 to 4.
  • Flight Behaviour - 2 types, rapid direct and not, scored 1 and 2 respectively.

Likelihood is a little more complex multiple criteria used simultaneously to assign 1 of 4 categories and other factors used to modify this score. Generally, however, it is abundance and strike history which contribute to this score.

The Paton model provides quantitative and qualitative descriptors for this part of the process. A number of sub-criteria elaborate further and do help you get to an answer.

With the consequence and likelihood categories in hand, the overall risk matrix gives you a final risk category - one of six from negligible to extreme. And that is your final answer - lock it in.


The Paton model has a number of limitations. The first one Paton mentions in the paper - time. Timeframes aren’t considered as part of the above process.

Also, ground-based animals aren’t considered. I don’t think this is such a big problem as body mass is just as relevant and one only needs to change flight and flock to movement and herd and it works pretty well.

It is, of course, a model and this means that it won’t apply in all circumstances. As long as you go in to the process eyes open and give the results an arms length gross error check, the results should work and users can have some confidence in the process.

Other Models

There are other approaches and I’m not making a judgement call which is better. The Paton model just happens to be the one I have used the most. There are links to other models in the resources section below.

In part 5 we are going to take the time to set some criteria to which we will compare this analysis. It is a step often overlooked. But before we get there, wouldn’t you like to make a comment on your preferred model or technique in the area below?

Additional Resources More information on wildlife hazard management and bird strikes can be found on the web at: