Estimating fire severity and carbon emissions over Australian tropical savannahs based on passive microwave satellite observations

Xi Chen*, Yi Y. Liu, Jason P. Evans, Robert M. Parinussa, Albert I.J.M. van Dijk, Marta Yebra

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

12 Citations (Scopus)

Abstract

We investigated the use of a recently developed satellite-based vegetation optical depth (VOD) data set to estimate fire severity and carbon emission over Australian tropical savannahs. VOD is sensitive to the dynamics of all aboveground vegetation and available nearly every two days. For areas burned during 2003–2010, we calculated the VOD change (ΔVOD) pre- and post-fire and the associated loss in the above ground biomass carbon. ΔVOD agreed well with the Normalized Burn Ratio change (ΔNBR) which is the metric used to estimate fire severity and carbon loss compared well with modelled emissions from the Global Fire Emissions Database (GFED). We found that the ΔVOD and ΔNBR are generally linearly related. The Pearson correlation coefficients (r) between VOD- and GFED-based fire carbon emissions for monthly and annual total estimates are very high, 0.92 and 0.96, respectively. A key feature of fire carbon emissions is the strong inter-annual variation, ranging from 21.1 Mt in 2010 to 84.3 Mt in 2004. This study demonstrates that a reasonable estimate of fire severity and carbon emissions can be achieved in a timely manner based on multiple satellite observations over Australian tropical savannahs, which can be complementary to the currently used approaches.

Original languageEnglish
Pages (from-to)6479-6498
Number of pages20
JournalInternational Journal of Remote Sensing
Volume39
Issue number20
DOIs
Publication statusPublished - 18 Oct 2018
Externally publishedYes

Fingerprint

Dive into the research topics of 'Estimating fire severity and carbon emissions over Australian tropical savannahs based on passive microwave satellite observations'. Together they form a unique fingerprint.

Cite this