Mitchard, ETA, Saatchi, SS, Woodhouse, IH, Nangendo, G, Ribeiro, NS, Williams, M, Ryan, CM, Lewis, SL, Feldpausch, TR and Meir, P (2009) Using satellite radar backscatter to predict above-ground woody biomass: A consistent relationship across four different African landscapes. Geophysical Research Letters, 36. ISSN 0094-8276Full text available as:
Mitchard_2009_Using satellite radar backscatter to predict above-ground biomass Africa.pdf
Available under License : See the attached licence file.
Regional-scale above-ground biomass (AGB) estimates of tropical savannas and woodlands are highly uncertain, despite their global importance for ecosystems services and as carbon stores. In response, we collated field inventory data from 253 plots at four study sites in Cameroon, Uganda and Mozambique, and examined the relationships between field-measured AGB and cross-polarized radar backscatter values derived from ALOS PALSAR, an L-band satellite sensor. The relationships were highly significant, similar among sites, and displayed high prediction accuracies up to 150 Mg ha−1 (±∼20%). AGB predictions for any given site obtained using equations derived from data from only the other three sites generated only small increases in error. The results suggest that a widely applicable general relationship exists between AGB and L-band backscatter for lower-biomass tropical woody vegetation. This relationship allows regional-scale AGB estimation, required for example by planned REDD (Reducing Emissions from Deforestation and Degradation) schemes.
|Copyright, Publisher and Additional Information:||© 2009 American Geophyical Union. Reproduced in accordance with the publisher's self-archiving policy.|
|Keywords:||carbon stocks, sar, vegetation, Australia, woodlands, savanna|
|Academic Units:||The University of Leeds > Faculty of Environment (Leeds) > School of Geography (Leeds)|
|Depositing User:||Symplectic Publications|
|Date Deposited:||12 Jan 2012 16:11|
|Last Modified:||08 Feb 2013 17:36|
|Publisher:||American Geophysical Union|
Actions (login required)