Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Estimating above ground biomass in a Salix plantation using high resolution UAV images

Nåfält, Simon LU (2018) In Student thesis series INES NGEM01 20181
Dept of Physical Geography and Ecosystem Science
Abstract
In forest biomass estimations, three dimensional (3D) canopy structure data derived from images captured by unmanned aerial vehicles (UAVs), also known as drones, has shown potential due to the flexibility and cost efficiency of the method. However, little research has been done for its applicability in bioenergy forest. In this study, a method was developed with the purpose of remotely estimating the above ground biomass of a bioenergy Salix stand in southern Sweden. The main aim was to design the method so that all the required input could be collected from a consumer grade UAV, making the application simple and at a relatively low cost. A 3D structure, or point cloud, of the Salix canopy was developed through structure from motion... (More)
In forest biomass estimations, three dimensional (3D) canopy structure data derived from images captured by unmanned aerial vehicles (UAVs), also known as drones, has shown potential due to the flexibility and cost efficiency of the method. However, little research has been done for its applicability in bioenergy forest. In this study, a method was developed with the purpose of remotely estimating the above ground biomass of a bioenergy Salix stand in southern Sweden. The main aim was to design the method so that all the required input could be collected from a consumer grade UAV, making the application simple and at a relatively low cost. A 3D structure, or point cloud, of the Salix canopy was developed through structure from motion analysis of multiple overlapping aerial images. Images collected from both a RGB sensor and a multispectral sensor were tested when developing the point cloud. The biomass was estimated in each 3.55 meter cell of a grid by an allometric equation based on the structural characteristics of the canopy. The results were compared to the harvest yield. The obtained standing biomass showed an underestimation by 8%, of the 105 ton harvest yield, which indicates that the method performed well for the study site. However, it was concluded that more testing is needed to fully evaluate the method. This study provides a suggestion on how 3D data can be used for remotely estimating biomass in a bioenergy Salix forest which can be valuable information for land owners in management decisions to improve the economic viability (Less)
Popular Abstract
From drone images to biomass

Bioenergy forests are fast growing plants, such as willow, that are harvested after 3-5 years and used for heat energy. After harvest the total wood volume is weighted and you will know how much money you made. But what if the money you made is less than you spent on e.g. irrigation and fertilizers to make the forest grow as fast and big as possible? It makes more sense to investigate how large the total wood volume of your forest is before you cut it down, and see if you actually will make a profit. Methods for pre-harvest estimations of bioenergy willow plantations usually include cutting down part of the forest and from this get an idea of the volume in whole plantation. In this study, an attempt was made... (More)
From drone images to biomass

Bioenergy forests are fast growing plants, such as willow, that are harvested after 3-5 years and used for heat energy. After harvest the total wood volume is weighted and you will know how much money you made. But what if the money you made is less than you spent on e.g. irrigation and fertilizers to make the forest grow as fast and big as possible? It makes more sense to investigate how large the total wood volume of your forest is before you cut it down, and see if you actually will make a profit. Methods for pre-harvest estimations of bioenergy willow plantations usually include cutting down part of the forest and from this get an idea of the volume in whole plantation. In this study, an attempt was made to use drone images.

Humans can see three dimensional shapes due to the fact that we have two eyes which provide two slightly different images of the object we observe, we have stereo vision. The same principle was used in this study to get a three-dimensional model of a willow plantation by capturing multiple overlapping drone images that provides a view of the forest from different angles. It is not possible to measure the biomass directly from the drone data. Instead something that is more easily measured can to be used to calculate the biomass. The developed 3D model provides information on the height and position of the willow stems in the plantation. The height information was the main data of the study and used to develop input parameters to an equation used to calculate the total wood volume, or the biomass, of the plantation.

The estimated biomass for the study area was 96.8 ton which only differed about 10 ton from the 105 ton harvest weight that was done a few days after the data for this study had been collected. It can therefore be concluded that the developed method performed well for this unique study area. The question is how well the method would perform in other areas.

This study provides a suggestion of how drones can be utilized for effective and low-cost biomass estimation in willow plantations. The information provides the land owners with more control over their plantation which can help maximizing the economic profit. (Less)
Please use this url to cite or link to this publication:
author
Nåfält, Simon LU
supervisor
organization
course
NGEM01 20181
year
type
H2 - Master's Degree (Two Years)
subject
keywords
biomass, Salix, structure from motion, unmanned aerial vehicle, allometry, point cloud
publication/series
Student thesis series INES
report number
465
language
English
id
8963727
date added to LUP
2018-12-02 14:08:24
date last changed
2018-12-02 14:08:24
@misc{8963727,
  abstract     = {{In forest biomass estimations, three dimensional (3D) canopy structure data derived from images captured by unmanned aerial vehicles (UAVs), also known as drones, has shown potential due to the flexibility and cost efficiency of the method. However, little research has been done for its applicability in bioenergy forest. In this study, a method was developed with the purpose of remotely estimating the above ground biomass of a bioenergy Salix stand in southern Sweden. The main aim was to design the method so that all the required input could be collected from a consumer grade UAV, making the application simple and at a relatively low cost. A 3D structure, or point cloud, of the Salix canopy was developed through structure from motion analysis of multiple overlapping aerial images. Images collected from both a RGB sensor and a multispectral sensor were tested when developing the point cloud. The biomass was estimated in each 3.55 meter cell of a grid by an allometric equation based on the structural characteristics of the canopy. The results were compared to the harvest yield. The obtained standing biomass showed an underestimation by 8%, of the 105 ton harvest yield, which indicates that the method performed well for the study site. However, it was concluded that more testing is needed to fully evaluate the method. This study provides a suggestion on how 3D data can be used for remotely estimating biomass in a bioenergy Salix forest which can be valuable information for land owners in management decisions to improve the economic viability}},
  author       = {{Nåfält, Simon}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Student thesis series INES}},
  title        = {{Estimating above ground biomass in a Salix plantation using high resolution UAV images}},
  year         = {{2018}},
}