English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This study assessed the prediction accuracy of the forest aboveground biomass (AGB) model using remotely sensed data sources (i.e. airborne laser scanning (ALS), RapidEye, Landsat), and the combination of ALS with RapidEye/Landsat using parametric weighted least squares (WLS) regression. We also analysed the AGB model using random forests, extremely randomized trees, and deep learning stacked autoencoder (SAE) network from nonparametric statistics to compare the performance with WLS regression. We also compared the widths of the 95% confidence intervals for estimates of the mean AGB per unit area using the model-based estimator. The study site in the Terai Arc Landscape, Nepal, comprised 14 protected areas extending from the southern part of Nepal to India and encompassed mosaics of continuous dense forest and tall grassland. The ALS data provided the largest prediction accuracy (0.30–0.35 relative root mean squared error (rRMSE)), whereas RapidEye and Landsat had smaller prediction accuracies (0.48‒0.54 and 0.47‒0.55 rRMSE, respectively) for the estimation of AGB. The combined use of ALS and RapidEye predictors in the AGB model reduced the rRMSE and narrowed the confidence interval compared with ALS alone, but the improvements were minor. The SAE prediction technique provided the largest prediction accuracy, with inputs of combined ALS and RapidEye predictors that yielded an R 2  of 0.80, an rRMSE of 0.30, and a confidence interval of 176‒184 compared to other tested prediction techniques. The SAE prediction technique can become more powerful than other tested prediction techniques if properly adjusted and tuned for accurate forest AGB mapping applications. To our knowledge, this is the first study assessing the performance of the SAE in AGB modelling with a range of hyper-parameter values.

Estimation of tropical forest aboveground biomass in Nepal using multiple remotely sensed data and deep learning