HomeRail NewsSeeing the wood in the trees - Network Rail tackles troublesome trees

Seeing the wood in the trees – Network Rail tackles troublesome trees

Listen to this article

‘Leaves on the line’ – an announcement that every commuter dreads and a term often repeated by the British media to the point of ridicule. Low adhesion caused by poor rail head conditions can affect train performance and punctuality in a number of ways. It’s the scourge of the autumnal and winter months, with passengers facing disruption due to train services being delayed, subjected to last-minute alterations or even cancelled.

Britain’s 52,000 hectares of railway land are home to millions of trees, bushes and other vegetation. A mature tree has between 10,000 and 50,000 leaves and, each autumn, thousands of tonnes of leaves fall onto railway lines across the country.

It’s a problem that is taken very seriously by Network Rail. One of its divisions, the National Delivery Service, is tasked with the contracting out of Railhead Treatment Trains. Operating overnight on key sections of the network, these trains undertake water-jetting and the spreading ‘Sandite’ on the line to improve the grip between wheel and rail.

Vegetation management has a role to play too, involving the removal of trees and ensuring other vegetation is cut back and managed. In addition, each autumn, track teams work around the clock at key locations using de-scaling machines to clean the railhead. All in all, leaf-strewn railway lines cost the UK economy millions of pounds each year.


So the annual leaf fall is a problem, but then there’s the trees themselves. Trees, or parts of trees, falling onto the railway don’t just create a safety risk. They can be the cause of thousands of pounds of damage and thousands of minutes of passenger and freight service delays. In 2016, there were more than 470 recorded incidents precipitated by vegetation affecting the railway. These ranged from fallen leaves affecting train braking to fallen branches disrupting overhead line equipment and fallen trees blocking routes entirely.

The cost of all this disruption is hard to calculate and Network Rail is unable to place a precise figure on it. However, according to the Linear Infrastructure Network, a group made up of infrastructure asset owners, managers, consultancies, Government and its agencies, the annual cost of vegetation impacts on train performance to the UK economy is estimated to be in the region of £100 million.

National survey

Network Rail decided that in order to alleviate this problem it would need to conduct a census of its estimated 10 million trees. With 20,000 miles of lineside to cover, work started in 2014 on what was essentially an aerial survey that made use of LiDAR (Light Detection and Ranging) mapping. This remote sensing method uses 3D laser scanning technology to measure the position of objects relative to the laser source.

In conducting the aerial survey, use was made of fixed-wing aircraft, helicopters and drones. These three methods complement each other. Although manned aircraft get the job done quickly, they can’t fly everywhere across the network. The proximity of airports, conurbations and other such no-fly zones enforced temporary gaps in the coverage. But two years after the start of the survey, the network had been covered, taking in an area that extends 60 metres each side of the track centreline, including third party land.

LiDAR is a common enough land surveying technique, but in interpreting the results Network Rail has made clever use of an additional technique known as ‘feature extraction.’ A point cloud image is created of the lineside, within which fixed structures such as location cases, OLE stanchions and bridge parapets can be pinpointed. Relative to these fixed points, the 3D envelope of the lineside foliage is mapped. The really clever bit is the way in which the resultant foliage data is interpreted.

Expert help

It’s all done by the magic of software analysis, of course. In simple terms, Network Rail has worked with arboriculturalist experts to create a tool that can identify individual trees that have the potential to affect the safe and smooth running of the railway. So what, exactly, is an arboriculturalist? The definition is a person who deals with all aspects of the growth, maintenance and removal of woody plants for any reason other than as a timber crop. In other words, an amenity tree expert.

The extraction, collation and reprocessing of data captured as part of the nationwide LiDAR survey was undertaken jointly by Network Rail, Airbus Defence & Space and Forbes-Laird Arboricultural Consultancy (FLAC). Use has been made of FLAC’s POLESTORM environmental risk control system, which models tree failure modes during weather events. This is supported by FAILSAFE, a risk model created by FLAC to predict derailments due to fallen trees.

The software can look at the complex shape of the tree canopy and can deduce from this the positions of the individual tree trunks. Coordinates are given relative to known fixed objects and, crucially, to the lie of the land. The proximity of a tree to the railway track is clearly important in assessing whether it’s a hazard, or potential hazard, but its position relative to other trees and the landscape topography is important too. In perfecting the software algorithms, previous risk models have been incorporated.

Assessing risk

Surprisingly, trees that have fallen onto railway lines have rarely been old and rotten. Usually, they are healthy trees that have become vulnerable to wind forces. Factors that affect the stability of trees include the ways in which they are aligned or grouped. Isolated trees or those growing on steep slopes would be deemed of higher risk than those forming a group on the flat. Worst of all would be an isolated tree at the top of a rock cutting. Other higher risk trees include those growing adjacent to stations, near over-bridges, tunnel portals or power lines.

Size is important too, of course, with trunks greater than 150 mm in diameter being regarded as a potential threat should they fall across the track. Less than this and the damage they could potentially cause is not deemed a significant safety risk. The survey has been able to deduce the size of potentially vulnerable trees from their height and thereby assess the threat level they present. In total, approximately 100 different tree attributes have been included within the LiDAR survey analysis software.

The survey algorithms can produce a ranking of risk, presented to highlight areas that require early intervention. In this way, the available manpower can be deployed to maximum efficiency. The presentation format has been tailored to provide information that is easy for route asset managers and maintenance teams to access and use. Remedial works can then be prioritised in a logical plan based around seven different levels of risk.

In its simplest form, the survey data, or more specifically the risk modelling, is imported into a Geographic Information System (GIS) that includes the railway infrastructure network model. The simplest output of this is a series of red dots where the highest risk trees are located along the railway. A 2016 trial revealed that a ‘heat map’ was one of the best methods of depicting geographical concentrations of the higher risk outputs and areas of overhanging tree canopies.


The tree census has arisen from the Offering Rail Better Information Services programme (ORBIS). This £330 million, seven-year project is designed to support the railway industry by collating, analysing and making use of data to create a detailed digital model of the UK’s rail network.

Launched in 2012, the programme aims to digitise maintenance of the UK’s railway infrastructure to help Network Rail manage assets more efficiently, cost-effectively and safely. It is predicted to save up to £1 billion over the next decade.

By helping mangers and engineers in decision making, Network Rail estimates that ORBIS has so far helped to save £207 million, by allowing engineers make smarter decisions and by ‘predicting and preventing’ incidents.

New additions to the ORBIS programme, such as the tree database, are forecast to save a further £281 million by 2019.


The previous requirement was that Network Rail would need to undertake line walking surveys at three year intervals, looking for any potentially problematic trees. This was supplemented by a more thorough tree assessment every five years. There has been an immediate cost saving by identifying areas to be managed and eliminating the need for a manual tree survey in those locations.

Furthermore, by using the tree census, Network Rail expects a dramatic reduction in the number of trees falling onto the tracks. As well as reducing delays to rail users this should bring significant safety benefits, not just by reducing the risk to trains, but also by reducing the time workers need to spend trackside. This, in turn, should also free up manpower for other activities.

Paul Meads, Network Rail’s head of lineside, said: “Our use of the tree census is emerging. Feedback will be useful for us to evaluate the benefits it can bring and we will be increasing our understanding of how it can best be used. Just now, the usage of the tree census is very much in its early days, but it promises to bring a real benefit to the rail industry as a whole.”

So far this year, Network Rail targeted and managed ‘hotspot’ areas for vegetation management covering 900 hectares, which accounts for roughly two per cent of the network. The work will continue and will include a reduction in the tree canopy in order to reduce leaf fall on the tracks. There will also be reduction in the lineside tree species that create the worst effect on rail adhesion.

This article was written by Stuart Marsh


  1. Millions of tons of biomass are imported from the US for use in Drax power station. What happens to the biomass produced during lineside vegetation clearance? If even a fraction of the 52,000 hectares of railway land are covered in trees and bushes the annual potential production of wood chips would run into six figures.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.