Remote Sensing Technology and Application, Volume. 39, Issue 5, 1261(2024)
Prediction of Forest Burned Area based on MODIS-EVI2 and Ensemble Learning
Figures & Tables(11)
Fig. 1. Forest fires in Australia in 2016， the red dot is the location of the fire
Fig. 2. Processing the satellite image flow of MOD13A1 products using ENVI5.3
Fig. 4. The prediction results of different models using EVI2 and NDVI as explanatory variables
Fig. 5. The accuracy of different models using EVI2 and NDVI as explanatory variables
Fig. 6. Experimental procedure of predicting Fire Burned Area by Stacking-XRSK Model
Fig. 7. The REC curves for the Stacking-XRSK model and base models
Table 1.
Detailed description of each variable View tableView in ArticleTable 1.
Detailed description of each variable 数据 数据描述 来源 火灾发生位置 经度 LON GFA1 纬度 LAT 火灾发生时间 时间 MON GFA1 气象因素 平均温度 TEMP NOAA2 平均风速 WDSP 最大持续风速 MXSPD 最高温度 MAX 最低温度 MIN 降雨量 PRCP 地形因素 海拔 ALT GE3 植被因素 两波段增强植被指数
归一化差值植被指数
EVI2
NDVI
ESDS4 过火面积 燃烧面积 SIZE GFA1
Table 2.
The parameter setting of four kinds of regression models View tableView in ArticleTable 2.
The parameter setting of four kinds of regression models 模型 参数 XGBoost max_depth=3；learning_rate=0.003
n_estimators=1000；booster=’gbtree’；min_child_weight=1；early_stopping_rounds=10；eval_metric=’rmse’
RF max_depth=5；n_estimators=500； min_samples_split=2；min_samples_leaf=1 SVM C=3；kernel=’rbf’；degree=3；tol=0.001 KNN n_neighbors=7
Table 3.
Evaluation results of the eight models with EVI2 as explanatory variable and NDVI as explanatory variable View tableView in ArticleTable 3.
Evaluation results of the eight models with EVI2 as explanatory variable and NDVI as explanatory variable MAE MSE XGBoost EVI2 0.068 2 0.012 9 0.811 1 NDVI 0.076 1 0.016 8 0.753 5 RF EVI2 0.066 3 0.017 7 0.740 6 NDVI 0.075 1 0.021 8 0.680 1 SVM EVI2 0.090 1 0.019 0.721 6 NDVI 0.098 5 0.021 8 0.680 9 KNN EVI2 0.079 5 0.022 3 0.672 9 NDVI 0.084 1 0.024 0.648 3
Table 4.
Evaluation results of Ensemble Learning Model and single Base Model View tableView in ArticleTable 4.
Evaluation results of Ensemble Learning Model and single Base Model MAE MSE Stacking-XRSK 0.063 1 0.012 2 0.815 5 XGBoost 0.084 7 0.015 2 0.769 6 RF 0.067 7 0.014 1 0.787 2 SVR 0.092 7 0.017 0.743 2 KNN 0.079 0.019 5 0.704 9
Tools
Get Citation
Copy Citation Text
Junchen FENG, Hao DONG, Peng HAN, Yuanbin LI, Jingyu LIU, Yunhong DING. Prediction of Forest Burned Area based on MODIS-EVI2 and Ensemble Learning[J]. Remote Sensing Technology and Application, 2024, 39(5): 1261
Paper Information
Category:
Received: Nov. 13, 2022
Accepted: --
Published Online: Jan. 7, 2025
The Author Email:
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20