Difference between revisions of "Above-ground biomass estimation in Finland"

From tessera
Jump to navigation Jump to search
Line 7: Line 7:
 
==Dataset==
 
==Dataset==
  
The dataset is a georeferenced subset of the [https://nascetti-a.github.io/BioMasster/ BioMassters dataset], containing 11,043 patches representing AGBM in Finnish forests from 2017 to 2021. Ground-truth measurements are obtained with airborne LiDAR.  
+
The dataset is a georeferenced subset of the [https://nascetti-a.github.io/BioMasster/ BioMassters dataset], containing 11,462 patches representing AGBM in Finnish forests from 2017 to 2021. Ground-truth measurements are obtained with airborne LiDAR.  
  
Each patch consists of 256×256 pixels, with each pixel covering a 10×10 meter area. The dataset is split into 8,689 training patches and 2,354 testing patches.
+
Each patch consists of 256×256 pixels, with each pixel covering a 10×10 meter area. The dataset is split into 8,689 training patches and 2,773 testing patches.
  
  
Line 15: Line 15:
 
<gallery widths="250" heights="250">
 
<gallery widths="250" heights="250">
 
File:Biomassters_locations.png|Patch locations by year of data collection
 
File:Biomassters_locations.png|Patch locations by year of data collection
File:Agbm_distributions.png|AGBM distributions in train and test data. The area of each histogram is normalized to sum up to one.
 
 
</gallery>
 
</gallery>
 
</div>
 
</div>

Revision as of 12:52, 28 October 2025

Objective

This subproject evaluates the performance of Tessera representation maps in above-ground biomass (AGBM) estimation in Finnish forests. Accurate AGBM estimation is essential for assessing the carbon sequestration capacity of forests and tracking changes caused by deforestation and recovery.


Dataset

The dataset is a georeferenced subset of the BioMassters dataset, containing 11,462 patches representing AGBM in Finnish forests from 2017 to 2021. Ground-truth measurements are obtained with airborne LiDAR.

Each patch consists of 256×256 pixels, with each pixel covering a 10×10 meter area. The dataset is split into 8,689 training patches and 2,773 testing patches.