Evaluating statistical and combine method to predict stand above-ground biomass using remotely sensed data

Bulut, Sinan; Sivrikaya, Fatih; Günlü, Alkan

doi:10.1007/s12517-022-10140-3

Cited by 8 publications

(1 citation statement)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, it allows working in larger areas with less cost and labor. Band reflectance, vegetation indices and various remote sensing data are commonly used to estimate carbon stock with remote sensing (Aricak et al, 2015;Günlü & Ercanlı, 2020;Günlü et al, 2021;Keleş et al, 2021;Turgut & Günlü, 2022;Bulut et al, 2022;Oktian et al, 2022;Sivrikaya & Demirel, 2022).…”

Section: Introductionmentioning

confidence: 99%

Estimating Above-Ground Carbon of Taurus Cedar Stands Using Sentinel-2 Satellite Image: A Case Study of Elmalı Forest Enterprise

Demirel,

Günlü,

Sakıcı

2023

3rd International Congress on Engineering and Life Science Proceedings Book

View full text Add to dashboard Cite

Most terrestrial carbon (C) is stored in forests, an important source of fiber and fuel for humans. Therefore, forests play an essential role in mitigating the effects of climate change by reducing the carbon level in the atmosphere. Field measurements and remote sensing techniques determine the stored above-ground carbon (AGC). This study used Sentinel-2 satellite image to estimate the amount of AGC in pure Taurus cedar (Cedrus libani A. Rich.) stands in Elmalı Forest Enterprise. Regression models were developed for AGC estimation with the reflectance and vegetation indices obtained from the Sentinel-2 satellite image. Within the scope of the study, the field measurement data obtained from 120 sample plots were used and AGCs of their corresponding stands were estimated with an allometric equation. The sample plots data was randomly divided into modeling (70%, 84 sample plots) and control data (30%, 36 sample plots) to fit the regression models and to test the accuracy of the models, respectively. Multiple linear regression analysis were conducted to develop the models, and three goodness-of-fit statistics (R2, RMSE and MAE) were used to compare the success of these models. When the achievements of the models were evaluated, it was revealed that the model containing the MSR vegetation indice gave more successful results (R2=0.488). Consequently, it was determined that the developed models were moderately successful in estimating AGC.

show abstract