ArbNation Join free

Pull tests and tree stability testing: what four published studies report

A static pull test loads a tree with a winch line and measures how far the stem base tilts. A doctoral thesis and three Forests papers record what the tests measure, how far they predict uprooting and where they fall short.

Checked against the sources at the bottom of this page on October 16, 2026. Rules, fees and pay change: the source has the last word.

The studies named below are described as published. It is not training and does not teach the work. They are a 2021 doctoral thesis from the University of Milano-Bicocca and three papers in the journal Forests, from 2023, 2024 and 2025.

What a static pull test measures

The Forests 2023 paper by Detter, Rust and Krišāns describes the idea. A winch applies a moderate load to the tree while a high-precision inclinometer at the stem base records the response. The data are then extrapolated to estimate the minimum anchorage strength of the root system. The paper adds that stability in storms is estimated by comparing that capacity with modelled wind loads for the tree's location.

Both papers on uprooting put the working range of a standard test at a small tilt. The 2023 paper uses 0.25 degrees of root plate inclination, and notes that a change in stem base inclination of no more than 0.5 degrees during pulling is fully reversible and counts as non-destructive. The 2025 paper from Milano-Bicocca and Dundee gives the same 0.25 degrees as the typical elastic limit. The thesis names the Static Integrated Method (SIM), with pulling to 0.2 degrees of root plate rotation.

Four studies compared

Published pull-test studies in this page's sources
StudyTreesWhat was compared
Milano-Bicocca doctoral thesis, 2021Almost 100 pulling tests from 2017 to 2020 at 11 places in Lombardy and Liguria, with 18 soil samplesTest records against the classic interpretation, with a more general equation proposed
Detter, Rust and Krišāns, Forests 2023More than 280 trees of mainly 15 species, uprooted in winching tests in Europe and North AmericaAnchorage strength predicted from non-destructive pulling against measured uprooting
Rust and Detter, Forests 202459 trees, mainly European beech and pedunculate oakStrength loss in the stem from static load tests against sonic tomography
Andreozzi and co-authors, Forests 202521 trees of eight species at the University of Dundee Botanic Gardens, 15 of them uprootedPredicted against measured overturning moment, pull direction and angle, soil properties

How well the predictions matched uprooting

The 2023 paper found that rotational stiffness at 0.25 degrees is a suitable proxy for anchorage strength. Both evaluation methods it tested systematically underestimated anchorage strength by about one third on average, and rarely overestimated it. For small trees the underestimate was sometimes up to a factor of two. The authors found no systematic difference between urban and woodland trees, nor between sites with very different soils, and they treat extrapolation beyond the tree sizes tested with caution.

The 2025 Dundee study reached a similar picture on a smaller sample. Of 15 uprooted trees, only 4 had their overturning moment overestimated by the prediction methods. For the other 11 both models underestimated it. The authors read this as support for the general applicability of the moment-based and stiffness-based approaches.

The same paper reports where the numbers vary. Peak moments in the uprooting tests occurred at inclinations of 1.59 to 2.93 degrees, and the authors take this as the reason to test to failure when the ultimate limit state is wanted. Soil shear strength, water content and grain size were at least as important as species. Trees of the same species and similar size showed no correlation in maximum overturning moment, and the paper concludes that stability depends much more on site-specific conditions of each tree. The overturning moment was largely unaffected by a change in pull angle.

Pull tests and sonic tomography

The 2024 paper tested the hypothesis that static load tests and sonic tomography give comparable estimates of lost strength in a stem. Earlier studies on six and seven trees had come to contradictory results. On 59 trees the correlation was weak, and for oak it was absent. The authors attribute part of that to how tomogram colours are sorted into carrying load and no longer carrying load, and to the simplified elliptical cross-section used in the load test analysis. They conclude the two methods cannot replace each other but can complement each other. Tomography has its own page, the guide to sonic and acoustic tomography.

Limits of the method in the thesis

The Milano-Bicocca thesis notes that trees fail biomechanically for meteorological, phytopathological and human causes, and that current stability evaluations have limitations. It calls SIM the main method for judging root stability because it is often the only one known, and says no alternative method matches its reliability and repeatability. The thesis also describes a deterministic approach to raising a tree's safety. It says that approach can only be repeated for trees of particular value, such as monumental trees, because of its complexity and cost.

Sources