Most tree assessments produce an opinion. QTRA produces a number. That sounds like a small difference in reporting style. It is not — it changes what the assessment can be used for, who can act on it, and whether it survives being challenged.
How the number is built
The Risk of Harm is the product of three assessed ranges.
- Target. What is beneath the tree and how often it is occupied, across six ranges. A road with continuous traffic, a playground used daily and a rear fence line used monthly are different targets under an identical tree.
- Size. The diameter of the part that could fail, across four ranges from 25 mm to over 450 mm. This is a proxy for the energy delivered on impact.
- Probability of failure. The likelihood of that part failing within twelve months, across seven ranges, from visual assessment of structure, decay, defects and loading.
The calculator combines the three using Monte Carlo simulation to derive a mean, and the result is reported to one significant figure — 1 in 8,000, 1 in 250,000. Reporting to one significant figure is deliberate: it signals that this is a calibrated estimate, not a false precision.
The thresholds it is read against
| Risk of Harm | Region | Indicated response |
|---|---|---|
| Greater than 1/1,000 | Unacceptable | Action required |
| 1/1,000 to 1/10,000 | Tolerable — upper | Closer consideration where the risk is imposed on others |
| 1/10,000 to 1/1,000,000 | Tolerable — lower | Manage and monitor |
| Less than 1/1,000,000 | Broadly acceptable | No action indicated; re-inspect on cycle |
The 1/10,000 line matters more than it looks. It separates risk a person has chosen to accept from risk imposed on them — a guest, a resident, a worker, a passer-by. Duty holders are held to a different standard for the second category, and the threshold is where that shows up in the arithmetic.
It is not a prediction that the tree will fail. A risk of harm of 1 in 12,000 does not mean the tree fails once every 12,000 years. It means that, integrating failure likelihood, target occupancy and impact energy, the annual probability of someone being harmed is about one in twelve thousand — a figure you can compare against other risks you already tolerate or control.
The four things a number does that a rating cannot
1. It lets you defend retention
This is the underrated one. Most people commission a tree assessment expecting it to justify removal. Just as often the useful outcome is the opposite: a documented, quantified basis for keeping a tree that somebody wants gone. “Broadly acceptable at 1 in 400,000, re-inspect in two years” is a position you can hold in front of a committee, a neighbour or a council.
2. It sorts a budget
Thirty trees rated “moderate” give you no ordering and no way to spend a limited budget rationally. Thirty numeric risks sort themselves, and the cut line falls where the money runs out rather than where the assessor's adjectives change.
3. It maps onto the legal test
Section 19 of the Work Health and Safety Act 2011 (Qld) requires control of risk so far as is reasonably practicable — a weighing of likelihood, degree of harm, knowledge, availability of controls and cost. You cannot weigh an adjective against a dollar figure. You can weigh a probability against one.
4. It survives a change of assessor
Because the inputs are recorded as ranges, a later assessor can see exactly which input drove the result and re-run it when something changes. If a courtyard is built under a tree that was previously over a service lane, only the target range changes — and the new number falls out without a fresh argument about the tree's condition.
An honest caveat
Quantification is not the same as objectivity. The probability-of-failure input is still a judgement made by a person looking at a tree. Research published in Arboriculture & Urban Forestry compared QTRA, TRAQ and untrained assessors and found no statistically significant difference in variability between the methods. Two competent QTRA assessors can land on different ranges for the same tree.
What the number gives you is not certainty. It is transparency — the reasoning is visible, the inputs are recorded, and a disagreement becomes a specific argument about one input rather than a clash of opinions.