Tree risk assessments| 15methods compared

Eric Frei · 20th July 2010, 09:39 PM

Interesting read from Martin Norris of The University of Melbourne.

Also there's been a lot of QTRA stuff going on lately, a licenced system for providing a hazard rating.

Seems the individual users own psyche has a lot to do with the output score.

Quote:

The 12 experienced arborists who assessed eight trees using eight different methods, produced the most
interesting results. The hypothesis was that experienced arborists would apply similar values in similar
circumstances and hence the differences produced by each method would be evident. However, it appears that
the differences in arborist applied assessment values to these eight trees across all methods was so diverse that
the influence of each method is not obvious and that the variation produced by the arborists is the greatest
influence on the risk output values created by each method. This observation was also confirmed by the
comparison of ‘intuitive risk’ values that were chosen by the arborists.
The application of a simple ten-point personality test (TIPI) identified a medium correlation between a TIPI
derived attribute and the arborists’ intuitive risk values, suggesting that to some degree an individual’s inherent
attitude in relation to risk possibly influences their risk assessment.
Uncertainty is a major factor in risk assessments, however, it is largely ignored by the assessment methods
reviewed. Uncertainty was obliquely addressed by the ‘Confidence Level’ assessment required by one method,
this found that the arborists assessing trees applied a wide range of confidence levels to their assessments,
indicative of uncertainty caused by lack of knowledge or data.
Of the eight methods assessed by the 12 arborists, the method that scored the highest value (and by a large
margin) in the ‘how well it works’ question was the simple two category Bartlett method, which given that none
of the arborists had previously used the method is interesting.
Analysis of the data currently suggests that the assumption of validity, completeness, robustness, and
repeatability should be challenged, including the base assumptions, and underlying modelling (particularly
weighting and mathematics). Equally, the wide range of assessment input variables chosen by arborists resulted
in a wide range of risk output values, which would suggest that individual differences amongst arborists question
the current value of risk assessments.
What Works – What Does Not – Can We Tell? Whilst analysis of the data gathered for this research is still to be
completed and reviewed, based on the methodology and data analysis to date, sensitivity analysis will identify
the weighting of the various components and likely distribution curve of output, however, it would appear that in
field tests the inter-arborist variability is such that ‘we cannot tell’ which models or methodology provides the
most complete measurement of tree risk.

Garry Brockley · 20th July 2010, 09:44 PM

This doesn't surprise me really as we have differing types of people and the human factor will always cause a differing result. Good find though Ekka

Eric Frei · 20th July 2010, 09:47 PM

It's been out a while but lately I keep arguing with people who fall for the bells and whistles bull of QTRA numbers.

I was told recently that a tree definitely has a 1:200 chance of failure because an "expert" says so via their QTRA result.

Licence to rate risk of trees, licence to infrared image them, soon be (if some have their way) a licence to work via some bull organisation where near on crims are housed.

treeshaveneeds · 21st July 2010, 09:54 PM

It would have been more interesting to have had the results of the 12 arborists, for each type of assessment, to better compare results between the arborists and between methods. The assumption that we all work with the same assumptions, based on the same experience was faulty to start.

Eric Frei · 21st July 2010, 10:38 PM

I see the whole process of trying to numerically rate the risk as flawed, it is a typical type of human conditional behaviour and not a science.

CaityR · 29th July 2010, 02:06 AM

Treelogic had a tree risk assessment weekend seminar - I think the US speakers were Matheny and Clark - and at the end of it they did an exercise where the 150 delegates assessed 4 trees and they came up with 12 different assessments for each tree. QTRA especially seems less transparent that, say, visual tree assessment. The 1 in 10000 'acceptable' risk is the most ridiculous aspect of QTRA because it started out as a health writer's opinion about where acceptable risk was. It was not based on research and was arbitrary.

Alon · 4th October 2010, 07:33 PM

I am not familiar with QTRA myself, but using statistics on trees? Really!? Where are the numbers coming from backing up these statistics? And this is for any species as well? Give me a break. Statistics in general are problematic, even when there are not many variables, and we can all agree that there are too many in tree risk assessment. Plus what are people supposed to do with it anyway? A tree has a 1:1000 chance of failing - will I cut it down based on this information? It's useless.

However, this study comes as no surprise to me. It is impossible to put anything tree-related on a flow chart and have a computer do the work. Trees are too complicated and need the expert's perspective and knowledge as well. Just like regular doctors, when in doubt get a second opinion.

Eric Frei · 14th November 2010, 08:21 AM

Here's a story where the passenger was killed by a failed tree that impaled her, condolences to her family.

I wonder what the QTRA numbers would have been, and if the weather variable factor was used? Use of the weather variable reduces the risk further.

Sure some will say 100mph winds is crazy and it could happen anywhere anytime. But the tree had no leaves as it was a deciduous tree, means a lot less sail. The breakage point from watching the video seems "suspicious". Also in the video they spoke of banked up cars and redirecting traffic so regardless of such storm conditions there was still people driving past.

Yorkshire woman killed as falling tree smashes into car (Video) - Yorkshire Post

JBardey · 13th January 2011, 11:55 PM

My condolences to the family of the woman involved.

The point of failure of that lime looks to me as though it was through lack of maintenance. The church was disused, suggesting that the systematic pollarding of the limes in the chruch yard would not have taken place, leading to large and heavy regrowth on potentially decayed anchor points.

A tragic reminder of the importance to keep up systematic maintenance once it has been started.

20th July 2010, 09:44 PM	#2
Garry Brockley Bayside Tree Care Brisbane Join Date: Jun 2008 Location: Brisbane Aus Posts: 1,649	Re: Tree risk assessments\| 15methods compared This doesn't surprise me really as we have differing types of people and the human factor will always cause a differing result. Good find though Ekka __________________ My business:- Brisbane Bayside Tree Care

20th July 2010, 09:47 PM	#3
Eric Frei Admin - Razor sharp and independent 2 X Diploma Level 5 qualified arborist Join Date: Jan 2007 Location: Brisbane Posts: 12,820	Re: Tree risk assessments\| 15methods compared It's been out a while but lately I keep arguing with people who fall for the bells and whistles bull of QTRA numbers. I was told recently that a tree definitely has a 1:200 chance of failure because an "expert" says so via their QTRA result. Licence to rate risk of trees, licence to infrared image them, soon be (if some have their way) a licence to work via some bull organisation where near on crims are housed. __________________ Protect your rights to work - Join Australian Tree Association My businesses:- Qualified Brisbane Tree Lopping \| Stump Grinding Brisbane \| Brisbane - Gold Coast Consulting Arborist

21st July 2010, 09:54 PM	#4
treeshaveneeds Veteran Heritage Status Join Date: Jun 2010 Location: Toronto, Ontario, Canada Posts: 1,399	Re: Tree risk assessments\| 15methods compared It would have been more interesting to have had the results of the 12 arborists, for each type of assessment, to better compare results between the arborists and between methods. The assumption that we all work with the same assumptions, based on the same experience was faulty to start.

21st July 2010, 10:38 PM	#5
Eric Frei Admin - Razor sharp and independent 2 X Diploma Level 5 qualified arborist Join Date: Jan 2007 Location: Brisbane Posts: 12,820	Re: Tree risk assessments\| 15methods compared I see the whole process of trying to numerically rate the risk as flawed, it is a typical type of human conditional behaviour and not a science. __________________ Protect your rights to work - Join Australian Tree Association My businesses:- Qualified Brisbane Tree Lopping \| Stump Grinding Brisbane \| Brisbane - Gold Coast Consulting Arborist

29th July 2010, 02:06 AM	#6
CaityR Former Member Join Date: Dec 2009 Location: Newcastle Posts: 17	Re: Tree risk assessments\| 15methods compared Treelogic had a tree risk assessment weekend seminar - I think the US speakers were Matheny and Clark - and at the end of it they did an exercise where the 150 delegates assessed 4 trees and they came up with 12 different assessments for each tree. QTRA especially seems less transparent that, say, visual tree assessment. The 1 in 10000 'acceptable' risk is the most ridiculous aspect of QTRA because it started out as a health writer's opinion about where acceptable risk was. It was not based on research and was arbitrary.

4th October 2010, 07:33 PM	#7
Alon Sappling Join Date: Aug 2010 Location: Israel Posts: 20	Re: Tree risk assessments\| 15methods compared I am not familiar with QTRA myself, but using statistics on trees? Really!? Where are the numbers coming from backing up these statistics? And this is for any species as well? Give me a break. Statistics in general are problematic, even when there are not many variables, and we can all agree that there are too many in tree risk assessment. Plus what are people supposed to do with it anyway? A tree has a 1:1000 chance of failing - will I cut it down based on this information? It's useless. However, this study comes as no surprise to me. It is impossible to put anything tree-related on a flow chart and have a computer do the work. Trees are too complicated and need the expert's perspective and knowledge as well. Just like regular doctors, when in doubt get a second opinion.

14th November 2010, 08:21 AM	#8
Eric Frei Admin - Razor sharp and independent 2 X Diploma Level 5 qualified arborist Join Date: Jan 2007 Location: Brisbane Posts: 12,820	Qtra Here's a story where the passenger was killed by a failed tree that impaled her, condolences to her family. I wonder what the QTRA numbers would have been, and if the weather variable factor was used? Use of the weather variable reduces the risk further. Sure some will say 100mph winds is crazy and it could happen anywhere anytime. But the tree had no leaves as it was a deciduous tree, means a lot less sail. The breakage point from watching the video seems "suspicious". Also in the video they spoke of banked up cars and redirecting traffic so regardless of such storm conditions there was still people driving past. Yorkshire woman killed as falling tree smashes into car (Video) - Yorkshire Post __________________ Protect your rights to work - Join Australian Tree Association My businesses:- Qualified Brisbane Tree Lopping \| Stump Grinding Brisbane \| Brisbane - Gold Coast Consulting Arborist

13th January 2011, 11:55 PM	#9
JBardey I'm new here so be nice Join Date: Jan 2011 Location: Warwickshire, England Posts: 3	Re: Tree risk assessments\| 15methods compared My condolences to the family of the woman involved. The point of failure of that lime looks to me as though it was through lack of maintenance. The church was disused, suggesting that the systematic pollarding of the limes in the chruch yard would not have taken place, leading to large and heavy regrowth on potentially decayed anchor points. A tragic reminder of the importance to keep up systematic maintenance once it has been started.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Danger tree Assessments / Fire line	kiwi_tree_steve	Ask an Arborist here	2	13th August 2009 04:52 AM
Palm Cleaning, what we see compared to the customer	Eric Frei	General Tree Chat	24	14th March 2009 09:38 PM
Urban Tree Risk Management USDA Forest Services	Eric Frei	Tree Information and Facts	0	27th January 2007 06:19 PM