Scholarlink

Background <p>The climate change disorder and global warming rising context of last decades has worsen wildfires hazard. Aware of that, the international community classified their occurrences, manifestations and consequences as a whole group of geoenvironmental disasters that need specific measures. This study proposes a geospatial multi-modelling that supports the ‘Fire Ready Formula’ prescribed by the United Nations Environment Program.</p> Methods <p>The split-window algorithm-based land surface temperature, SWA-LST, is computed from Landsat 8/9 data, to assess thermal dynamics. Next, a thermal flexile burn index, T<sub>h</sub>FBI, based on near infrared, shortwave infrared and thermal bands, targets the spots and degree of burned. Further, a fire disturbance model, FDM, solicits the least squares function between greenness and drought sub-models, to predict environmental damages. Complementarily, a threshold and reclassification index alternative, TRIAL, is proposed from the OTSU-segmentation and the random forest algorithm, to map spots and classes of fire, plus risk and disturbance.</p> Results <p>The pre-fire versus post-fire values range of the SWA-LST record a significant increase, up to 16.7&#xa0;°C. Besides, the T<sub>h</sub>FBI highlights all burned spots and classes, as validated by the spectral discrimination index values, ɱ: [1.003–4.615] and the sensitivity analysis, AUC: [0.812–0.864]. Concerning the FDM, the pre-fire/post-fire probability and cumulative density function graphs display each, two separates then crossed curves, that depict biophysical actual and forecasted changes. Whereas, the graphing of SWA-LST, T<sub>h</sub>FBI and FDM, show less to significant intersections, as the degree of their imbrication on each site. Finally, the TRIAL successfully depicts the charriest spots as "high risk" exposure to fire and disturbance.</p> Conclusions <p>Overall, despite some adjustments and cautions related to local parameters missing for a more precise SWA-LST, the singleness of T<sub>h</sub>FBI per site and a need for further FDM validation, this whole procedure could be of a technical support to fire management decision-makers at any scale.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

How do geoenvironmental fire disasters impact land surface thermal budget and threaten biophysical equilibrium? A geospatial multi-modelling tentative answer from triggered dynamics

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月24日

查看全文
  • Alfred Homère Ngandam Mfondoum

摘要

Background

The climate change disorder and global warming rising context of last decades has worsen wildfires hazard. Aware of that, the international community classified their occurrences, manifestations and consequences as a whole group of geoenvironmental disasters that need specific measures. This study proposes a geospatial multi-modelling that supports the ‘Fire Ready Formula’ prescribed by the United Nations Environment Program.

Methods

The split-window algorithm-based land surface temperature, SWA-LST, is computed from Landsat 8/9 data, to assess thermal dynamics. Next, a thermal flexile burn index, ThFBI, based on near infrared, shortwave infrared and thermal bands, targets the spots and degree of burned. Further, a fire disturbance model, FDM, solicits the least squares function between greenness and drought sub-models, to predict environmental damages. Complementarily, a threshold and reclassification index alternative, TRIAL, is proposed from the OTSU-segmentation and the random forest algorithm, to map spots and classes of fire, plus risk and disturbance.

Results

The pre-fire versus post-fire values range of the SWA-LST record a significant increase, up to 16.7 °C. Besides, the ThFBI highlights all burned spots and classes, as validated by the spectral discrimination index values, ɱ: [1.003–4.615] and the sensitivity analysis, AUC: [0.812–0.864]. Concerning the FDM, the pre-fire/post-fire probability and cumulative density function graphs display each, two separates then crossed curves, that depict biophysical actual and forecasted changes. Whereas, the graphing of SWA-LST, ThFBI and FDM, show less to significant intersections, as the degree of their imbrication on each site. Finally, the TRIAL successfully depicts the charriest spots as "high risk" exposure to fire and disturbance.

Conclusions

Overall, despite some adjustments and cautions related to local parameters missing for a more precise SWA-LST, the singleness of ThFBI per site and a need for further FDM validation, this whole procedure could be of a technical support to fire management decision-makers at any scale.