Tree Risk Assessment for Arborists

A homeowner asks if a big oak is safe. There is no honest yes or no. Every tree can fail, so the real questions are what could fail, what it could hit, how likely that is, and how bad it would be. A tree risk assessment turns a gut feeling into a reasoned rating you can write down and defend. It also sets the limits of what you are claiming. This guide teaches the logic. It does not qualify you to sell assessments. The International Society of Arboriculture's Tree Risk Assessment Qualification (TRAQ) is a widely recognized credential for that work, and some places have their own rules. You learn to read trees from experienced assessors, on real trees, over years.

What You Will Be Able to Do

• Explain what each level of assessment covers
• Define a time frame, target zone, and occupancy
• Rate the likelihood of failure and of impact
• Use the two matrices to reach a risk rating
• Recommend mitigation and state the residual risk
• Separate risk to the public from risk to your crew

Risk Is More Than Defects

A dead tree in a forest with no trail has a high chance of falling and almost no risk. A sound tree beside a school has a low chance of failing and a much worse outcome if it does. So a defect alone is not a risk. Risk needs three things:

1. Failure: A part of the tree breaks or the whole tree falls.
2. Impact: That part hits something that matters.
3. Consequences: The hit causes harm.

You assess each tree part and each target separately. A tree can have one dead limb over a patio and a root problem facing a street. Those are different risks, and each needs its own rating.

Levels of Assessment

ISA's system has three levels. The client and the job decide which one fits.

Level 2 tools include a mallet, a probe, binoculars, and a trowel. Level 3 might mean aerial inspection, decay testing with a resistance drill or sonic tomograph, root collar excavation, or load testing. The drill wounds the tree, so it is used only when the answer is worth it.

ISA's Basic Tree Risk Assessment Form is built for Level 2. It is a data form, not a report. A report adds your scope, limits, and recommendations in words a client can use.

Set the Time Frame

Likelihood means nothing without a time period. Everything fails eventually. The ISA form uses a stated time frame, usually between one and five years. Write it down. A rating of "possible" over one year is not the same as over five. Decay and new defects also change a tree, so you should recommend when to reassess.

Define the Targets

A target is a person, property, or activity that could be hurt, damaged, or disrupted. Be specific. "Sidewalk" is weak. "People at the bus stop on the north sidewalk" is useful.

Target zone is how far the tree or part can reach.

Worked example. A tree stands 70 feet tall. A house sits 60 feet away, so it is inside 1 x height. A playground sits 95 feet away. 1.5 x 70 is 105 feet, so it is within reach if brittle branches shatter and fly.

Occupancy is how much of the time a target is in the zone.

A house is a constant target. A walking trail may be occasional. Note anything that protects the target, such as a roof, a shelter, or lower branches.

Likelihood of Failure

This is the rating that takes the most experience. It has four levels.

Imminent stands apart, because it overrides the time frame. A tree with a split trunk and a root plate lifting after a storm does not wait for a year to pass. Clear the area and act now.

What You Weigh

• Load: Wind exposure, crown size, and any recent change. Trees that lost neighbors to clearing suddenly take wind they never carried.
• Defects: Decay, cracks, weak attachments, included bark, poor taper, and dead wood.
• Response growth: Wood a tree adds in reaction to stress. It can be a sign of a problem, and it can also mean the tree has corrected for it.
• Site: Grade changes, trenching, cut roots, saturated soil, or compaction.
• Species: Some species shed limbs in summer or grow weak wood. Local experience matters here.

Lean counts only with context. A lean a tree has corrected for over decades is very different from a new lean with soil heaving on the far side.

Hollow Trunks and the t/R Ratio

If a trunk has decay, you may hear about t/R. It compares the thickness of sound wood in the shell (t) with the radius of the trunk (R).

Worked example. A trunk is 30 inches across, so the radius is 15 inches. A sound shell measures 3.5 inches. 3.5 divided by 15 is about 0.23.

Researcher Claus Mattheck and his colleagues reported that trees rarely broke when t/R was above about 0.3 to 0.35 and at least half the trunk's girth was intact. That ratio is a widely used screening rule. It is not a pass or fail test.

• Many trees stand for years below 0.3.
• Large openings change the math.
• Later research has shown the limits of the rule, so treat it as a rough guide.
• A strong trunk does not protect against root failure.

Use t/R as one piece of evidence. Do not rate a tree on one number.

Likelihood of Impact

If this part fails, how likely is it to hit this target? Choose from very low, low, medium, or high. Think about occupancy, the direction of likely failure, and any protection. A limb hanging over a road used all day is more likely to hit a car than a limb over a seldom-used path.

Consequences

If it hits, how bad is it? Choose from negligible, minor, significant, or severe. Part size matters, and so does fall distance. A small dead twig on a lawn is negligible. A large limb onto people is severe.

From Ratings to Risk

ISA's form uses two steps. Read each line from left to right.

Matrix 1 combines likelihood of failure with likelihood of impact. Reading across very low, low, medium, and high impact:

• Imminent: Unlikely, Somewhat likely, Likely, Very likely
• Probable: Unlikely, Unlikely, Somewhat likely, Likely
• Possible: Unlikely, Unlikely, Unlikely, Somewhat likely
• Improbable: Unlikely in all four

Matrix 2 combines that result with consequences. Reading across negligible, minor, significant, and severe:

• Very likely: Low, Moderate, High, Extreme
• Likely: Low, Moderate, High, High
• Somewhat likely: Low, Low, Moderate, Moderate
• Unlikely: Low in all four

Worked example. A dead limb 6 inches thick hangs over a bus stop. In a one-year time frame, you rate failure as probable. People wait there for parts of the day, so you rate impact as medium. Matrix 1 gives somewhat likely. A limb that size could seriously hurt someone, so consequences are severe. Matrix 2 gives a rating of moderate.

Do this for each part and target. The tree's overall rating is the highest of the ratings you found.

Mitigation and Residual Risk

Mitigation means reducing risk. It works on any of the three parts.

Moving a target is often cheaper than touching the tree, and it is easy to forget.

Worked example. Using the bus stop above, moving the stop 40 feet down the road drops likelihood of impact to very low. Matrix 1 goes to unlikely, and Matrix 2 gives low. The limb stays, but the risk falls.

Residual risk is the risk left after mitigation, worked out the same way. A rating of none is only used when the tree is recommended for complete removal. Give the client options, not orders. Your job is to give the facts and a reasoned recommendation. The tree risk manager, usually the owner, decides how much risk to accept.

Record Your Limits

Your rating is only as good as what you could see. The form asks for inspection limitations, and you must fill them in honestly.

• Visibility: leaves, snow, fog, or vines
• Access: fences, slopes, or a closed part of the site
• Root collar buried under soil, mulch, or pavement

Mark your data as final or preliminary. If you could not get enough information, say so, and recommend an advanced assessment. Then give a reassessment time. Trees change, so an assessment describes a tree at a moment, for a set period. It is not a promise that the tree will stand.

Risk to the Public Is Not Risk to the Crew

A risk rating is for the people and property near the tree. Your crew faces a separate question. A dead tree in a remote corner may rate low for the public, and it may still be dangerous to climb, because the climber is the target and vibration from cutting is the load.

Before anyone climbs or cuts, the crew leader decides how the work can be done safely. Decay at a tie-in point, a dead top, or a lifting root plate can mean no climbing at all. That is when you use a lift, a crane, or a different plan. Say no when you need to.

Common Mistakes

Check Your Understanding

1. A tree 60 feet tall has a split trunk that widened in last night's storm. A daycare playground sits 50 feet away and is used all day. How do you rate failure, and what happens now?

Imminent, because failure has started. That overrides the time frame. Keep people out of the zone at once and get it dealt with right away.

2. You rate a limb as possible failure, and impact as low. What does Matrix 1 give you?

Unlikely. On Matrix 2, that is low whatever the consequences.

3. You review a colleague's report. It says a hollow beech must come down because t/R is 0.25, and it lists no other findings. What do you send back?

That ratio is a screening figure, not a pass or fail test, and many trees stand for years below 0.3. Ask for the rest of the evidence: openings, roots, load, and targets. The rating has to come from all of it.

4. Your report on a maple used a two-year time frame, and that time has passed. A new driveway has since cut through its roots. The owner asks if the old rating still holds. What do you say?

No. An assessment describes a tree at one moment, for a set period. That period is over, and cut roots change the site, so the tree needs a new assessment.

Key Takeaways

• Risk comes from failure, impact, and consequences together.
• Assess each part and target, within a stated time frame.
• Use the two matrices, and report the highest rating as overall.
• Mitigation can change the tree or the target.
• Record your limits, and recommend a reassessment date.
• A rating for the public does not make a tree safe to climb.