Fire is among the major disasters leading to biodiversity loss and climate change. Analysing high temporal resolution data over space and time can provide critical inputs for environmental management. The present study attempted to analyse fire occurrences to determine the distribution of vegetation fire hotspots in Afghanistan, Bangladesh, Bhutan, India, Nepal, Pakistan, Sri Lanka, and Myanmar. The spatial statistics and space–time pattern mining tools are used for investigating fire trends that occurred across vegetated landscapes over a period of 16 years. This study found a total of 2,352,063 active vegetation fire locations using daily MODIS data from 2003 to 2018. Mann–Kendall test has identified the statistically significant trend of fire hotspots across forests, scrub, grasslands, agriculture, and plantations. The emerging hotspot analysis detected considerable spatial variation across fire hotspots and identified the geographical clusters of fire hotspots across the South Asian countries and Myanmar. The fires showed variations with an increasing or decreasing trend or no statistically significant trend in different land cover types across the eight countries. The emerging hotspot analysis found the dominance of sporadic hotspots and persistent hotspots.

中文翻译：

---

表征缅甸和南亚国家的植被火灾动态

火灾是导致生物多样性丧失和气候变化的主要灾害之一。分析空间和时间的高时间分辨率数据可以为环境管理提供关键输入。本研究试图分析火灾事件，以确定阿富汗、孟加拉国、不丹、印度、尼泊尔、巴基斯坦、斯里兰卡和缅甸的植被火灾热点分布。空间统计数据和时空模式挖掘工具用于调查 16 年来植被景观中发生的火灾趋势。本研究使用 2003 年至 2018 年的每日 MODIS 数据发现了总共 2,352,063 个活跃植被火灾地点。 Mann-Kendall 检验确定了森林、灌木丛、草原、农业和种植园火灾热点的统计显着趋势。新兴热点分析检测到火灾热点之间的相当大的空间变化，并确定了南亚国家和缅甸的火灾热点地理集群。火灾在八个国家/地区的不同土地覆盖类型中显示出增加或减少趋势或没有统计上显着趋势的变化。新兴热点分析发现零星热点和持续热点占主导地位。