您所在的位置:首页>>报告>>报告详情

Multi-temporal threshold algorithm in forest fire detection using MSG satellite: The case of Zimbabwe

基于MDG卫星的森林火灾探测多时相阈值算法:以津巴布韦为例

【作       者】:

{{d.作者}}

【机       构】: 非洲技术政策研究网
【承研机构】:

【英文机构】: African Technology Policy Studies Network
【原文地址】: https://atpsnet.org/wp-content/uploads/2017/05/wps76.pdf
【发表时间】:

2013-03-01

摘要

Forest fires have the potential to increase the amount of carbon in the atmosphere which is one of the greenhouse gases responsible for global warming. Climate change may also lead to increased outbreak of forest fires due to increased dryness conditions especially in tropical Africa. This indicates the need for improved methods for detection, monitoring and management of forest fires so as to protect the fragile ecosystems. Remote sensing has been widely used in active forest fire detection; however there are some limitations in operational contextual algorithms as they are greatly affected by clouds and different land cover types such as land and water with inherent temperatures that may be included in the 3 x 3 kernel or matrix used in estimating the possibility of fire in the centre pixel. Therefore this working paper evaluates the accuracy of the multi-temporal threshold algorithm in Zimbabwe based on

the hypothesis that both multi-temporal threshold algorithm and contextual algorithm (MSG fire product) have equal performance on forest fire detection. Unlike contextual algorithms, the multi-temporal threshold algorithm estimates the mean background temperature of a point (pixel) over a number of days and any temperatures above the mean are attributed to fire. The error matrix and McNemar’s test were used for accuracy assessment and testing the hypothesis of this study respectively. The preliminary results of this study have shown that the multi-temporal threshold algorithm has a higher forest fire detection rate (50.5%) as compared to MSG fire product (24.1%) which uses the contextual algorithm. There is a significant difference in the performance of these algorithms (McNemar’s test statistic (x2) =4.7, df=1, p-value=0.0295). However this is still at a preliminary stage and further validation will be performed using 2010 and 2011 data to support and give value to these results as the multi-temporal threshold algorithm is expected to perform better than contextual algorithm.This may give more confidence to the potential users and any stakeholders interested in adopting the multi-temporal threshold algorithm in their forest fire management systems in Zimbabwe and around the globe.

标签: {{b}}
展开

相关报告

  • 暂无相关数据