Decision makers need data from forest ecosystems

In order to make wise decisions and policies for climate protection, we need more understanding on the interactions between the carbon cycles and the changes in environment. A new research project “Understand forest ecosystem cycles using high density terrestrial laser scanning (TLS) time series data” will demonstrate, how understanding ecosystem cycles in forest can be enhanced through automatically collected and analysed TLS time series data.

Puita metsässä, jossa aurinko pilkistää puiden raosta
Photo:
Julia Hautojärvi

Laser scanning data reveals how trees are doing

‘We will model how tree growth is affected by water, temperature, sunlight and other environmental factors through constant observation of trees in forest. We will use high temporal and spatial resolution TLS time series. With an automated observing and analysing solution, it is possible to observe and to measure changes on stems, branches and leaves through a span of time. We can also link those changes with changes in local water, temperature, and sunlight conditions,’ says Senior Research Scientist Yunsheng Wang, who leads the project. 

Since tree growth stores carbon, this will help the researchers to understand how carbon is exchanged between trees and their outside environment. 

‘It will also help us to predict and to manage the carbon dynamics in a given forest environment,’ Yunsheng Wang adds.

Laser scanning produces an enormous amount of data

Laser scanning is a method to digitize the real world by collecting the 3D locations of laser return points that are backscattered from object surfaces. The Laser scanning produces a 3D point cloud describing the 3D structures of the objects in real world, so that the 3D shapes of the objects can be precisely reconstructed in a virtual world using computers. In the project, researchers use a unique long-term point cloud time series, which are produced by a TLS system installed permanently on a Hyytiälä forest station tower. Such system is capable of capturing the 3D shape of a baseball that is at 100 m distance. 

The system will be able to monitor leaf-level changes in a forest of 5 hectare size, with a repetition rate of twice per hour. The system will produce immense amount of data, about 10 GB / hour data for several years. The data analysis chain must be a very intelligent design in order to be able to process all data.  

Edge computing and data pipeline enable big data analysis

The main aim of the project is to develop an edge computing and artificial intelligence powered data pipeline to solve the problem of big data analysis. Edge computing improves response times and thus makes computing faster. In edge computing, computing and data storage are brought closer to where they are needed. Data pipeline is a series of data processing steps which makes a fluent and automated work flow from the input data to the expected output results.  

The researchers are using novel computing technologies, which offer many benefits. Data collection is enhanced with learning an optimal data collection scheme. Efficiency is achieved by close-to-real time processing locally at sensors. The researchers will create novel computer self-learning process for ecosystem cycle modelling, and a complete pipeline including dispatched modules. This all enables faster results to scientific community and to society.

‘This also facilitates an artificial intelligent process that enables the computers to automatically and continuously optimize the ecosystem cycle models with growing data. Such models can support decision making in climate protection by providing prediction of tree growth dynamics caused by changes in forest stands and weather conditions.’ explains Yunsheng Wang.

Project “Monitoring and understanding forest ecosystem cycles using high temporal and spatial resolution terrestrial laser scanning time series” is funded by the Academy of Finland. The project started in September 2020 and it will end in August 2024. 

More information

Senior Research Scientist Yunsheng Wang, +358 50 347 8903, firstname.lastname@nls.fi

Uusimmat uutiset ja blogit