Canadian Forest Service Publications

Estimates of forest growing stock volume of the northern hemisphere from Envisat ASAR. 2013. Santoro, M.; Schmullius, C.; Pathe, C.; Schwilk, J.; Beer, C.; Thurner, M.; Fransson, J.E.S.; Shvidenko, A.; Schepaschenko, D.; McCallum, I.; Hall, R. J.; Beaudoin, A. Proceedings of ESA Living Planet Symposium, Edinburgh, September 9-13, 2013. ESA Communication Office, Nordwijk, the Netherlands, pp. SP-722, CD-ROM.

Year: 2013

Issued by: Northern Forestry Centre

Catalog ID: 35591

Language: English

Availability: PDF (request by e-mail)

Mark record


This paper presents and assesses a dataset of forest growing stock volume (GSV) above 30°N estimated from multi-temporal backscatter measurements acquired by Envisat Advanced Synthetic Aperture Radar (ASAR) in ScanSAR mode. The retrieval was based on the BIOMASAR algorithm that combines standard SAR processing techniques for multi-temporal SAR data and an automated GSV estimation approach based on a Water Cloud-like model and a weighted combination of GSV estimates from individual measurements of the SAR backscatter. Approximately 70,000 ASAR images were used to estimate GSV at a pixel size of 0.01°. Assessment of the retrieved GSV indicated (i) the importance of comparing to reference data at the same scale, (ii) moderate agreement with reference datasets at full resolution because of the limited sensitivity of C-band SAR backscatter to GSV, (iii) strong agreement with reference datasets after aggregation at regional and national level, (iv) consistent retrieval results throughout the study area, (v) saturation at approximately 300 m3/ha.

Plain Language Summary

Knowing forest volume (the amount of wood per area) is important to forest management and to calculating amounts of biomass and carbon. But estimating forest volume is challenging in the northern hemisphere, across broad ecological zones encompassing polar, boreal, temperate and subtropical regions. To address this problem, we used radar data from the Envisat satellite to survey representative areas in these regions to estimate forest volume. This approach builds upon a previously published radar data estimation method that was developed and applied across northern boreal regions. We developed maps of forest volume from these data. Uncertainty was smallest in the boreal and temperate regions and largest in the subtropical region. The data provide an unprecedented first estimate for 2010 across three continents and four ecological zones.