Allometric modeling for biomass estimation in remote sensing

[ABSTRACT]This research develops an allometric model for estimation of biomass based on the height and DBH of trees in the forests of Iran. Without an accurate allometric model, we can’t estimate true biomass from SAR images. After extensive field investigations, 124 trees are selected from the 4 different dominant tree species for the development of an allometric model covering the wide range of DBH and height classes. Twelve commonly used allometric models, three generic models and a proposed model are tested and the most suitable model is selected based on some of the commonly measured statistical parameters (coefficient of determination (R2), RMSE, Mean Error, Underestimated Error, and Overestimated Error). We show that the biomass estimation accuracy is improved in a multilayer perceptron neural network (MLPNN) when the density of woods and the tree measurements are used in combination compared to estimating the biomass by classic allometric models.

https://opac.ll.chiba-u.jp/da/curator/107498/

https://opac.ll.chiba-u.jp/da/curator/107498/ceresIntlSympo-20-P059.pdf

Chiba University. Center for Environmental Remote Sensing

千葉大学環境リモートセンシング研究センター

会議発表用資料

Proceedings of the CEReS international symposium = CEReS国際シンポジウム資料集

20

59

60

2013-08-08

publisher

環境リモートセンシング研究センター「CEReSシンポジウム資料集」