基于機(jī)載激光雷達(dá)的森林地上碳儲量估測1)

穆喜云 劉清旺 龐勇 胡凱龍 張秋良

(赤峰市林業(yè)科學(xué)研究院森林生態(tài)研究所， 內(nèi)蒙古·赤峰， 024000)(中國林業(yè)科學(xué)研究院資源信息研究所) (中國礦業(yè)大學(xué)(北京)地球科學(xué)與測繪工程學(xué)院) (內(nèi)蒙古農(nóng)業(yè)大學(xué)林學(xué)院)

?

基于機(jī)載激光雷達(dá)的森林地上碳儲量估測1)

穆喜云 劉清旺 龐勇 胡凱龍 張秋良

(赤峰市林業(yè)科學(xué)研究院森林生態(tài)研究所， 內(nèi)蒙古·赤峰， 024000)(中國林業(yè)科學(xué)研究院資源信息研究所) (中國礦業(yè)大學(xué)(北京)地球科學(xué)與測繪工程學(xué)院) (內(nèi)蒙古農(nóng)業(yè)大學(xué)林學(xué)院)

以內(nèi)蒙古大興安嶺生態(tài)站為研究對象，以2012、2013年的66塊樣地?cái)?shù)據(jù)和2012年同步獲取的機(jī)載LiDAR遙感數(shù)據(jù)為數(shù)據(jù)源，分別采用多元線性回歸和隨機(jī)森林回歸算法，通過對比不同算法間的估測精度差異，選擇更適于研究區(qū)的估測方法，實(shí)現(xiàn)研究區(qū)森林地上碳儲量的遙感估測。結(jié)果表明：隨機(jī)森林回歸算法的估測精度最優(yōu)，模型訓(xùn)練精度(R2為0.861，RMSE為11.133 t/hm2，rRMSE為0.279)和預(yù)測精度(RMSE為17.956 t/hm2，rRMSE為0.342，估測精度范圍40.898%～95.129%，平均估測精度76.385%)均優(yōu)于多元線性回歸的模型訓(xùn)練結(jié)果(R2為0.676，RMSE為11.846 t/ha，rRMSE為0.351)和模型預(yù)測結(jié)果(RMSE為22.703 t/hm2，rRMSE為0.636，估測精度范圍45.824%～94.752%，平均估測精度69.859%)。機(jī)載LiDAR數(shù)據(jù)的高度變量和密度變量與森林地上碳儲量均具有顯著相關(guān)性，高度變量相關(guān)性更為顯著。隨機(jī)森林回歸算法對區(qū)域森林地上碳儲量的估測結(jié)果趨于真實(shí)分布情況，效果比較理想。

機(jī)載LiDAR；隨機(jī)森林回歸；多元線性回歸；森林地上碳儲量

In the Great Khingan State Ecosysterm Research Station in Inner Mongolia, we chose a more suitable method to estimate forest aboveground carbon storage with the plots data from 2012, 2013 and the synchronously acquired airborne LiDAR data of 2012 as data sources in the study area, by comparing the model estimated accuracy of multiple linear stepwise regression and random forest regression algorithms to realize the remote sensing estimation of forest aboveground carbon storage of study area. The random forest regression algorithm was training higher accuracy (model training accuracyR2=0.861, RMSE=11.133 t/ha andrRMSE=0.279; testing accuracyR2=0.826, RMSE=17.956 t/ha,rRMSE=0.342, the estimate accuracy range is in 40.898%-95.129% and its average estimate accuracy is 76.385%) than the multiple linear stepwise regression algorithm (model training accuracyR2=0.676, RMSE=11.846 t/ha andrRMSE=0.351; testing accuracyR2=0.727, RMSE=22.703 t/ha,rRMSE=0.636, the estimate accuracy range is in 45.824%-94.752% and the average estimate accuracy is 69.859%). The percentile height and density variables of LiDAR data had significant correlation with the forest aboveground carbon storage, percentile height variable correlation is more significant. Therefore, the estimate results of total forest carbon storage on regional scale using random forest regression algorithm was closer to its true distribution with ideal effects.

森林作為陸地生態(tài)系統(tǒng)的主體，蓄儲了全球陸地生態(tài)系統(tǒng)中約80%以上的碳[1]，作為陸地生態(tài)系統(tǒng)中最大的碳庫，對吸收CO2等溫室氣體，減緩氣候變化發(fā)揮著關(guān)鍵作用。……