Comparing terrestrial laser scanning (TLS) and satellite imagery for estimating carbon stocks in trees of a semi-arid savannah
(2024) In Student thesis series INES NGEM01 20241Dept of Physical Geography and Ecosystem Science
- Abstract
- Due to the intensifying impacts of climate change, accurate estimation of global carbon stocks is crucial, with aboveground vegetation biomass representing a significant carbon reservoir. Savannahs, which cover approximately 50% of the African continent, play a vital role in carbon storage, as trees within these ecosystems contribute substantially to the continent’s total carbon stocks. Estimating aboveground biomass carbon stocks can be achieved through various methods, including terrestrial laser scanning (TLS), which provides data from a local scale, and satellite imagery, which enables broad-scale assessments. Both techniques capture tree metrics that, when combined with allometric equations, allow for the estimation of carbon stocks.... (More)
- Due to the intensifying impacts of climate change, accurate estimation of global carbon stocks is crucial, with aboveground vegetation biomass representing a significant carbon reservoir. Savannahs, which cover approximately 50% of the African continent, play a vital role in carbon storage, as trees within these ecosystems contribute substantially to the continent’s total carbon stocks. Estimating aboveground biomass carbon stocks can be achieved through various methods, including terrestrial laser scanning (TLS), which provides data from a local scale, and satellite imagery, which enables broad-scale assessments. Both techniques capture tree metrics that, when combined with allometric equations, allow for the estimation of carbon stocks. For TLS applications, segmentation algorithms are necessary to isolate individual trees and derive these metrics. Commonly applied algorithms in tree segmentation include Dalponte2016, Watershed, Silva2016, and Li2012.
This study aims to estimate the carbon stock of trees within a savannah ecosystem located in Dahra, Senegal, using TLS. The results from the TLS data were compared to satellite-derived estimates from the study by Tucker et al. (2023). A secondary objective was to identify the most effective segmentation algorithm for processing TLS data. Among the algorithms tested, Dalponte2016 performed best, as it achieved a higher match rate with inventory data, exhibited lower absolute errors, and showed consistent estimations.
The TLS-derived estimate for carbon stock was 2.61 tonC ha⁻¹, whereas the satellite-derived estimate for the Dahra savannah was 2.10 tonC ha⁻¹. Relative to inventory data, TLS provided more accurate results, with lower root mean square error (RMSE) and absolute bias, and identified a greater number of trees. These findings suggest that large-scale remote sensing methods may underestimate a significant amount of stored carbon, which, if accounted for, could bring estimates closer to actual carbon storage levels. TLS data thus has the potential to serve as a benchmark for calibrating or establishing error margins in large-scale satellite-based assessments, enhancing the precision of global carbon stock estimates in savannah ecosystems. (Less) - Popular Abstract
- With climate change accelerating, it’s essential to accurately measure how much carbon is stored globally, and one major carbon reserve is the biomass of trees. Savannahs, which cover about half of Africa, play a big role here, as their trees contribute significantly to the continent's carbon storage. There are a few ways to measure the carbon in these trees. One method, called terrestrial laser scanning (TLS), uses ground-based lasers to obtain tree structure over smaller areas. Another approach uses satellite imagery to cover larger areas. Both methods capture information from trees that helps us estimate the carbon they hold. For TLS to gather these details, we use segmentation algorithms that isolate individual trees; some commonly... (More)
- With climate change accelerating, it’s essential to accurately measure how much carbon is stored globally, and one major carbon reserve is the biomass of trees. Savannahs, which cover about half of Africa, play a big role here, as their trees contribute significantly to the continent's carbon storage. There are a few ways to measure the carbon in these trees. One method, called terrestrial laser scanning (TLS), uses ground-based lasers to obtain tree structure over smaller areas. Another approach uses satellite imagery to cover larger areas. Both methods capture information from trees that helps us estimate the carbon they hold. For TLS to gather these details, we use segmentation algorithms that isolate individual trees; some commonly used algorithms include Dalponte2016, Watershed, Silva2016, and Li2012.
In this study, TLS was used to estimate the carbon stored in a savannah ecosystem in Dahra, Senegal, with results compared to satellite-based estimates by Tucker et al. (2023). The study also aimed to identify the best segmentation algorithm for analyzing TLS data. Among those tested, Dalponte2016 performed best, with more accurate matches to inventory data, lower errors, and consistent results.
The TLS method estimated the area’s carbon storage at 2.61 tons of carbon per hectare, while satellite imagery gave a lower estimate of 2.10 tons per hectare. Compared to inventory-based estimates, TLS provided more accurate results and identified more trees. This suggests that satellite methods may miss some carbon storage, and using TLS data to refine satellite estimates could bring these measurements closer to reality. TLS can thus help improve the accuracy of carbon estimates from satellite imagery, providing a more complete picture of carbon storage in ecosystems like savannahs. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9177956
- author
- Buschinelli Carreño, Debora LU
- supervisor
- organization
- course
- NGEM01 20241
- year
- 2024
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Physical Geography and Ecosystem analysis, TLS, Satellite imagery, Carbon stock, Savannah
- publication/series
- Student thesis series INES
- report number
- 684
- language
- English
- id
- 9177956
- date added to LUP
- 2024-11-18 10:33:57
- date last changed
- 2024-11-18 10:33:57
@misc{9177956,
abstract = {{Due to the intensifying impacts of climate change, accurate estimation of global carbon stocks is crucial, with aboveground vegetation biomass representing a significant carbon reservoir. Savannahs, which cover approximately 50% of the African continent, play a vital role in carbon storage, as trees within these ecosystems contribute substantially to the continent’s total carbon stocks. Estimating aboveground biomass carbon stocks can be achieved through various methods, including terrestrial laser scanning (TLS), which provides data from a local scale, and satellite imagery, which enables broad-scale assessments. Both techniques capture tree metrics that, when combined with allometric equations, allow for the estimation of carbon stocks. For TLS applications, segmentation algorithms are necessary to isolate individual trees and derive these metrics. Commonly applied algorithms in tree segmentation include Dalponte2016, Watershed, Silva2016, and Li2012.
This study aims to estimate the carbon stock of trees within a savannah ecosystem located in Dahra, Senegal, using TLS. The results from the TLS data were compared to satellite-derived estimates from the study by Tucker et al. (2023). A secondary objective was to identify the most effective segmentation algorithm for processing TLS data. Among the algorithms tested, Dalponte2016 performed best, as it achieved a higher match rate with inventory data, exhibited lower absolute errors, and showed consistent estimations.
The TLS-derived estimate for carbon stock was 2.61 tonC ha⁻¹, whereas the satellite-derived estimate for the Dahra savannah was 2.10 tonC ha⁻¹. Relative to inventory data, TLS provided more accurate results, with lower root mean square error (RMSE) and absolute bias, and identified a greater number of trees. These findings suggest that large-scale remote sensing methods may underestimate a significant amount of stored carbon, which, if accounted for, could bring estimates closer to actual carbon storage levels. TLS data thus has the potential to serve as a benchmark for calibrating or establishing error margins in large-scale satellite-based assessments, enhancing the precision of global carbon stock estimates in savannah ecosystems.}},
author = {{Buschinelli Carreño, Debora}},
language = {{eng}},
note = {{Student Paper}},
series = {{Student thesis series INES}},
title = {{Comparing terrestrial laser scanning (TLS) and satellite imagery for estimating carbon stocks in trees of a semi-arid savannah}},
year = {{2024}},
}