Human amplification of drought-induced biomass burning in Indonesia since 1960

Human amplification of drought-induced biomass burning in Indonesia since 1960
复制标题

DOI:
10.1038/ngeo443
复制
发表时间:
2009-03-01
期刊:
影响因子:
18.3
通讯作者:
Shen, Samuel S. P.
Shen, Samuel S. P.
中科院分区:
地球科学1区
文献类型:
--
作者:
Field, Robert D.;van der Werf, Guido R.;Shen, Samuel S. P.

文献摘要

被引文献

相似文献

全球大气二氧化碳浓度的年际变化大部分归因于生物量燃烧排放的变化(1-3)。在干旱条件下,印度尼西亚的燃烧是这些排放的不成比例的贡献者,如1997/98年烟霾灾害所示(1,4)。然而,由于缺乏1990年代中期之前对火灾监测有用的卫星数据,我们对印度尼西亚严重生物质燃烧的频率、严重程度和根本原因的了解有限。在这里,我们提出了一个连续的每月记录严重燃烧事件,从1960年至2006年使用能见度报告在该地区的机场。我们发现,这些火灾导致极差的空气质量条件,它们只发生在降水量福尔斯下降到一个明确的阈值以下的年份。从历史上看,苏门答腊至少从20世纪60年代以来就发生过大型火灾事件。相比之下,加里曼丹(印度尼西亚婆罗洲)在1980年代记录了第一次大火,尽管早些时候发生了严重的干旱。我们将这种差异归因于土地利用和人口密度变化的不同模式。印度尼西亚的火灾通常与厄尔尼诺现象(1,2,5 -12)有关,但我们发现印度洋偶极子模式也是一个重要的促成因素。
Much of the interannual variability in global atmospheric carbon dioxide concentrations has been attributed to variability of emissions from biomass burning(1-3). Under drought conditions, burning in Indonesia is a disproportionate contributor to these emissions, as seen in the 1997/98 haze disaster(1,4). Yet our understanding of the frequency, severity and underlying causes of severe biomass burning in Indonesia is limited because of the absence of satellite data that are useful for fire monitoring before the mid-1990s. Here we present a continuous monthly record of severe burning events from 1960 to 2006 using the visibility reported at airports in the region. We find that these fires cause extremely poor air quality conditions and that they occur only during years when precipitation falls below a well defined threshold. Historically, large fire events have occurred in Sumatra at least since the 1960s. By contrast, the first large fires are recorded in Kalimantan ( Indonesian Borneo) in the 1980s, despite earlier severe droughts. We attribute this difference to different patterns of changes in land use and population density. Fires in Indonesia have often been linked with El Nino(1,2,5-12), but we find that the Indian Ocean Dipole pattern is as important a contributing factor.