Hero Image
ESA

Deriving improved nation-wide raster databases using satellite images and airborne laser data

In Sweden the NFI provide nation-wide wall-to-wall raster maps with estimates of forest variables representing the state of the forest in year 2000, 2005, and 2010 (kNN Sweden 2000, 2005, and 2010). Presently, the Swedish Forest Agency (Skogsstyrelsen) and SLU are producing a new nation-wide forest raster dataset based on the ALS dataset (0.5-1 pulses/m2) that the Swedish Land Survey is acquiring for the production of a new national DEM.

The main objective of the proposed project is to develop and implement methods for updating the ALS based estimates derived by the Forest Agency and SLU using a combination of optical satellite data with terrain features and vegetation metrics derived from the national ALS dataset. The updated raster dataset should include estimates of of total stem volume, biomass, tree species composition (which is not included in the present dataset), tree height, diameter, and basal area representing the state of the forest in 2015.

Two different approaches for deriving the new dataset will be tested and compared with each other. The first approach is to estimate growth and tree species composition by combining optical satellite data, terrain features, and laser metrics using the kMSN method. The second approach is to use the ALS based estimates derived by the Forest Agency and SLU in combination with optical satellite data and terrain features for the kMSN estimation. Landsat or SPOT data from the national satellite database SACCESS will be used in the project.

The final estimates representing the state of the forest in 2015 will be calculated by adding the plot-wise growth estimates derived using kMSN to the ALS based estimates derived by the Forest Agency and SLU. A “link” from each pixel to the NFI plots that according to the kMSN algorithm is closest in feature space distance will also be established and stored in the raster database.
The new nation-wide raster dataset will provide the NFI, the Forest Agency and other users with an improved product that can be used in various applications.

Diarienummer
228/14
Lärosäte
Sveriges lantbruksuniversitet
2015
600 000 kr
Total
600 000 kr