Volume 11, Issue 21 (8-2023)
ifej 2023, 11(21): 75-87 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Karimi S, Heydari M, Mirzaei J, Karami O, Mosavi A. (2023). Modeling Post-Fire Vegetation Recovery using Satellite and Environmental Data in Zagros Forest Ecosystem, Ilam. ifej. 11(21), 75-87. doi:10.61186/ifej.11.21.75
URL: http://ifej.sanru.ac.ir/article-1-475-en.html
URL: http://ifej.sanru.ac.ir/article-1-475-en.html
Ilam University
Abstract: (1685 Views)
Extended Abstract
Introduction and Objective: The occurrence of fires is one of the important factors that determine the different characteristics of many terrestrial ecosystems. For a long time, fires have severely affected forest areas, and sometimes their negative effects remain for several years after the occurrence of the fire, so that the state of vegetation does not return to its previous state. The aim of this study is to model the restoration of vegetation in Zagros forests (Ilam province) following fire.
Material and Methods: We used various climatic and environmental data as independent variables (vegetation at the time of fire (NDVI+1), burn severity index, temperature and precipitation anomaly, average temperature, annual precipitation, slope, aspect, and elevation) and NDVI +5 as dependent variable for the modeling (using random forest, decision tree and gradient boosting) the vegetation recovery following fire. Landsat satellite images were used to prepare indices indicating vegetation density status and burn severity, and after preprocessing the images, these indices were prepared by spectral ratio. Climatic variables (precipitation, average temperature, minimum temperature and maximum temperature) were also estimated according to the regression relationships between these variables and the elevation in the study area. Finally, three machine learning algorithms, including decision tree, random forest, and gradient boosting, were used for modeling, and also the accuracy of these models were evaluated.
Results: The results showed that among the various variables investigated, the annual precipitation, average annual temperature, normalized vegetation difference index (NDVI) and burn intensity index at the time of fire were the most important factors affecting the vegetation restoration post fire in these forests. The precipitation and temperature were the most important factors affecting the restoration among the mentioned factors. Also, the results showed that among the different models, the gradient boosting algorithm with R2 = 0.66 models vegetation restoration better than other models. In this model, the climatic factors were the most important in the vegetation recovery.
Conclusion: According the relationships between the NDVI and other studied factors and the results of the modeling; it is possible to explain the effective role of climate factors in the vegetation restoration in the study area.
Introduction and Objective: The occurrence of fires is one of the important factors that determine the different characteristics of many terrestrial ecosystems. For a long time, fires have severely affected forest areas, and sometimes their negative effects remain for several years after the occurrence of the fire, so that the state of vegetation does not return to its previous state. The aim of this study is to model the restoration of vegetation in Zagros forests (Ilam province) following fire.
Material and Methods: We used various climatic and environmental data as independent variables (vegetation at the time of fire (NDVI+1), burn severity index, temperature and precipitation anomaly, average temperature, annual precipitation, slope, aspect, and elevation) and NDVI +5 as dependent variable for the modeling (using random forest, decision tree and gradient boosting) the vegetation recovery following fire. Landsat satellite images were used to prepare indices indicating vegetation density status and burn severity, and after preprocessing the images, these indices were prepared by spectral ratio. Climatic variables (precipitation, average temperature, minimum temperature and maximum temperature) were also estimated according to the regression relationships between these variables and the elevation in the study area. Finally, three machine learning algorithms, including decision tree, random forest, and gradient boosting, were used for modeling, and also the accuracy of these models were evaluated.
Results: The results showed that among the various variables investigated, the annual precipitation, average annual temperature, normalized vegetation difference index (NDVI) and burn intensity index at the time of fire were the most important factors affecting the vegetation restoration post fire in these forests. The precipitation and temperature were the most important factors affecting the restoration among the mentioned factors. Also, the results showed that among the different models, the gradient boosting algorithm with R2 = 0.66 models vegetation restoration better than other models. In this model, the climatic factors were the most important in the vegetation recovery.
Conclusion: According the relationships between the NDVI and other studied factors and the results of the modeling; it is possible to explain the effective role of climate factors in the vegetation restoration in the study area.
Type of Study: Research |
Subject:
سنجش از دور
Received: 2022/08/23 | Accepted: 2022/10/17 | Published: 2023/08/1
Received: 2022/08/23 | Accepted: 2022/10/17 | Published: 2023/08/1
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
