Comparison of Statistical Modelling Approaches for Estimating Tropical Forest Aboveground Biomass Stock and Reporting Their Changes in Low-Intensity Logging Areas Using Multi-Temporal LiDAR Data

Accurately quantifying forest aboveground biomass (AGB) is one of the most significant challenges in remote sensing, and is critical for understanding global carbon sequestration. Here, we evaluate the effectiveness of airborne LiDAR (Light Detection and Ranging) for monitoring AGB stocks and change (ΔAGB) in a selectively logged tropical forest in eastern Amazonia. Specifically, we compare results from a suite of different modelling methods with extensive field data. The calibration AGB values were derived from 85 50 × 50m field plots established in 2014 and which were estimated using airborne LiDAR data acquired in 2012, 2014, and 2017. LiDAR-derived metrics were selected based upon Principal Component Analysis (PCA) and used to estimate AGB stock and change. The statistical approaches were: ordinary least squares regression (OLS), and nine machine learning approaches: random forest (RF), several variations of k-nearest neighbour (k-NN), support vector machine (SVM), and artificial neural networks (ANN). Leave-one-out crossRemote Sens. 2020, 12, 1498 2 of 21 validation (LOOCV) was used to compare performance based upon root mean square error (RMSE) and mean difference (MD). The results show that OLS had the best performance with an RMSE of 46.94 Mg/ha (19.7%) and R2 = 0.70. RF, SVM, and ANN were adequate, and all approaches showed RMSE ≤ 54.48 Mg/ha (22.89%). Models derived from k-NN variations all showed RMSE ≥ 64.61 Mg/ha (27.09%). The OLS model was thus selected to map AGB across the time-series. The mean (± sd standard deviation) predicted AGB stock at the landscape level was 229.10 (± 232.13) Mg/ha in 2012, 258.18 (±106.53) in 2014, and 240.34 (sd±177.00) Mg/ha in 2017, showing the effect of forest growth in the first period and logging in the second period. In most cases, unlogged areas showed higher AGB stocks than logged areas. Our methods showed an increase in AGB in unlogged areas and detected small changes from reduced-impact logging (RIL) activities occurring after 2012. We also detected that the AGB increase in areas logged before 2012 was higher than in unlogged areas. Based on our findings, we expect our study could serve as a basis for programs such as REDD+ and assist in detecting and understanding AGB changes caused by selective logging activities in tropical forests.

[1]  D. Pitt,et al.  Parametric vs. nonparametric LiDAR models for operational forest inventory in boreal Ontario , 2013 .

[2]  D. R. Almeida,et al.  Evaluating observed versus predicted forest biomass: R-squared, index of agreement or maximal information coefficient? , 2019, European Journal of Remote Sensing.

[3]  Teuvo Kohonen,et al.  An introduction to neural computing , 1988, Neural Networks.

[4]  Anthony C. Janetos,et al.  Climate mitigation and the future of tropical landscapes , 2010, Proceedings of the National Academy of Sciences.

[5]  Sassan Saatchi,et al.  A Systems Engineering Approach to Estimating Uncertainty in Above‐Ground Biomass (AGB) Derived from Remote‐Sensing Data , 2014, Syst. Eng..

[6]  Victoria Meyer,et al.  Comparison of Small- and Large-Footprint Lidar Characterization of Tropical Forest Aboveground Structure and Biomass: A Case Study From Central Gabon , 2018, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[7]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[8]  Nicholas L. Crookston,et al.  yaImpute: An R Package for kNN Imputation , 2008 .

[9]  Nicholas C. Coops,et al.  Estimation of forest biomass dynamics in subtropical forests using multi-temporal airborne LiDAR data , 2016 .

[10]  Christopher Baraloto,et al.  Rapid tree carbon stock recovery in managed Amazonian forests , 2015, Current Biology.

[11]  G. Asner,et al.  A universal airborne LiDAR approach for tropical forest carbon mapping , 2011, Oecologia.

[12]  Paul Duffy,et al.  Long-Term Impacts of Selective Logging on Amazon Forest Dynamics from Multi-Temporal Airborne LiDAR , 2019, Remote. Sens..

[13]  Florian Hartig,et al.  Importance of sample size, data type and prediction method for remote sensing-based estimations of aboveground forest biomass , 2014 .

[14]  Daniel Mozos,et al.  FPGA implementation of the principal component analysis algorithm for dimensionality reduction of hyperspectral images , 2019, Journal of Real-Time Image Processing.

[15]  Erik Næsset,et al.  Modeling and Estimating Change , 2014 .

[16]  Lênio Soares Galvão,et al.  Combining LiDAR and hyperspectral data for aboveground biomass modeling in the Brazilian Amazon using different regression algorithms , 2019, Remote Sensing of Environment.

[17]  Adrián Cardil,et al.  Measuring Individual Tree Diameter and Height Using GatorEye High-Density UAV-Lidar in an Integrated Crop-Livestock-Forest System , 2020, Remote. Sens..

[18]  John F. Weishampel,et al.  Structural diversity indices based on airborne LiDAR as ecological indicators for managing highly dynamic landscapes , 2015 .

[19]  C. Silva,et al.  A principal component approach for predicting the stem volume in Eucalyptus plantations in Brazil using airborne LiDAR data , 2016 .

[20]  N. Altman An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression , 1992 .

[21]  Sandra A. Brown,et al.  Monitoring and estimating tropical forest carbon stocks: making REDD a reality , 2007 .

[22]  Andy Liaw,et al.  Classification and Regression by randomForest , 2007 .

[23]  D. Clark,et al.  Tropical forest biomass estimation and the fallacy of misplaced concreteness , 2012 .

[24]  Alessandro Montaghi,et al.  A performance comparison of machine learning methods to estimate the fast-growing forest plantation yield based on laser scanning metrics , 2015, Comput. Electron. Agric..

[25]  Z. Niu,et al.  Correlating the Horizontal and Vertical Distribution of LiDAR Point Clouds with Components of Biomass in a Picea crassifolia Forest , 2014 .

[26]  Xiaohuan Xi,et al.  Above-ground biomass estimation using airborne discrete-return and full-waveform LiDAR data in a coniferous forest , 2017 .

[27]  Liviu Theodor Ene,et al.  Modelling aboveground forest biomass using airborne laser scanner data in the miombo woodlands of Tanzania , 2015, Carbon Balance and Management.

[28]  R. Chazdon Tropical forest recovery: legacies of human impact and natural disturbances , 2003 .

[29]  Sandra A. Brown,et al.  Carbon emissions from tropical forest degradation caused by logging , 2014 .

[30]  Marcos Heil Costa,et al.  A comparison of precipitation datasets for the Amazon Basin , 1998 .

[31]  Darío Domingo,et al.  Comparison of regression models to estimate biomass losses and CO2 emissions using low-density airborne laser scanning data in a burnt Aleppo pine forest , 2017 .

[32]  Plinio Sist,et al.  Above-ground biomass dynamics after reduced-impact logging in the Eastern Amazon , 2010 .

[33]  S. Goetz,et al.  A meta-analysis of terrestrial aboveground biomass estimation using lidar remote sensing , 2013 .

[34]  M. dos-Santos,et al.  Quantification of selective logging in tropical forest with spaceborne SAR interferometry , 2018, Remote Sensing of Environment.

[35]  K. Moffett,et al.  Remote Sens , 2015 .

[36]  Wan Shafrina Wan Mohd Jaafar,et al.  ForestGapR: An r Package for forest gap analysis from canopy height models , 2019, Methods in Ecology and Evolution.

[37]  C. Silva,et al.  Optimizing individual tree detection accuracy and measuring forest uniformity in coconut (Cocos nucifera L.) plantations using airborne laser scanning , 2019, Ecological Modelling.

[38]  Lutz Fehrmann,et al.  Comparison of linear and mixed-effect regression models and a k-nearest neighbour approach for estimation of single-tree biomass , 2008 .

[39]  T. Breurch,et al.  A simple test for heteroscedasticity and random coefficient variation (econometrica vol 47 , 1979 .

[40]  Juan de la Riva,et al.  Estimation of Total Biomass in Aleppo Pine Forest Stands Applying Parametric and Nonparametric Methods to Low-Density Airborne Laser Scanning Data , 2018 .

[41]  G. Asner,et al.  Environmental and Biotic Controls over Aboveground Biomass Throughout a Tropical Rain Forest , 2009, Ecosystems.

[42]  Robert J. McGaughey,et al.  Monitoring selective logging in western Amazonia with repeat lidar flights , 2014 .

[43]  Christina Gloeckner,et al.  Modern Applied Statistics With S , 2003 .

[44]  Gregory Asner,et al.  Spatially-Explicit Testing of a General Aboveground Carbon Density Estimation Model in a Western Amazonian Forest Using Airborne LiDAR , 2015, Remote. Sens..

[45]  Matthew G. Hethcoat,et al.  A Machine Learning Approach to Map Tropical Selective Logging , 2018, bioRxiv.

[46]  Heng Tao Shen,et al.  Principal Component Analysis , 2009, Encyclopedia of Biometrics.

[47]  C. Silva,et al.  Estimating Above-Ground Biomass of Araucaria angustifolia (Bertol.) Kuntze Using LiDAR Data , 2019, Floresta e Ambiente.

[48]  Arto Haara,et al.  Comparing K Nearest Neighbours Methods and Linear Regression - Is There Reason To Select One Over the Other? , 2012, Math. Comput. For. Nat. Resour. Sci..

[49]  E. Næsset Practical large-scale forest stand inventory using a small-footprint airborne scanning laser , 2004 .

[50]  Sassan Saatchi,et al.  Impacts of Airborne Lidar Pulse Density on Estimating Biomass Stocks and Changes in a Selectively Logged Tropical Forest , 2017, Remote. Sens..

[51]  Jungho Im,et al.  Forest biomass estimation from airborne LiDAR data using machine learning approaches , 2012 .

[52]  Steven M. Lalonde,et al.  Multivariate Statistical Methods: A First Course , 2001, Technometrics.

[53]  V. Kapos,et al.  Reducing Greenhouse Gas Emissions from Deforestation and Forest Degradation: Global Land-Use Implications , 2008, Science.

[54]  D. Roberts,et al.  Estimation of tropical rain forest aboveground biomass with small-footprint lidar and hyperspectral sensors , 2011 .

[55]  R. Valentini,et al.  Biodiversity Mapping in a Tropical West African Forest with Airborne Hyperspectral Data , 2014, PloS one.

[56]  Guofan Shao,et al.  Improving Lidar-based aboveground biomass estimation of temperate hardwood forests with varying site productivity , 2018 .

[57]  Ovidiu Csillik,et al.  Monitoring tropical forest carbon stocks and emissions using Planet satellite data , 2019, Scientific Reports.

[58]  S. Shapiro,et al.  An Analysis of Variance Test for Normality (Complete Samples) , 1965 .

[59]  B. Koch,et al.  Non-parametric prediction and mapping of standing timber volume and biomass in a temperate forest: application of multiple optical/LiDAR-derived predictors , 2010 .

[60]  Gregory P. Asner,et al.  Forest canopy damage and recovery in reduced-impact and conventional selective logging in eastern Para, Brazil , 2002 .

[61]  M. Keller,et al.  Selective Logging in the Brazilian Amazon , 2005, Science.

[62]  O. Phillips,et al.  The changing Amazon forest , 2008, Philosophical Transactions of the Royal Society B: Biological Sciences.

[63]  Sandra Englhart,et al.  Quantifying Dynamics in Tropical Peat Swamp Forest Biomass with Multi-Temporal LiDAR Datasets , 2013, Remote. Sens..

[64]  K. O. Niemann,et al.  Simulated impact of sample plot size and co-registration error on the accuracy and uncertainty of LiDAR-derived estimates of forest stand biomass , 2011 .

[65]  R. Houghton,et al.  Aboveground Forest Biomass and the Global Carbon Balance , 2005 .

[66]  B. Nelson,et al.  Improved allometric models to estimate the aboveground biomass of tropical trees , 2014, Global change biology.

[67]  M. d'Oliveira,et al.  Estimating forest biomass and identifying low-intensity logging areas using airborne scanning lidar in Antimary State Forest, Acre State, Western Brazilian Amazon , 2012 .

[68]  R. McRoberts A model-based approach to estimating forest area , 2006 .

[69]  A. Hudak,et al.  Nearest neighbor imputation of species-level, plot-scale forest structure attributes from LiDAR data , 2008 .

[70]  Piermaria Corona,et al.  Design-based approach to k-nearest neighbours technique for coupling field and remotely sensed data in forest surveys , 2009 .

[71]  Paulo Barreto,et al.  Financial and ecological indicators of reduced impact logging performance in the eastern Amazon , 2002 .

[72]  Roberta E. Martin,et al.  Amazonian landscapes and the bias in field studies of forest structure and biomass , 2014, Proceedings of the National Academy of Sciences.

[73]  R. Betts,et al.  Climate Change, Deforestation, and the Fate of the Amazon , 2008, Science.

[74]  Barbara Koch,et al.  Status and future of laser scanning, synthetic aperture radar and hyperspectral remote sensing data for forest biomass assessment , 2010 .

[75]  Hailemariam Temesgen,et al.  A Comparison of Selected Parametric and Non-Parametric Imputation Methods for Estimating Forest Biomass and Basal Area , 2014 .

[76]  John A. Scrivani,et al.  Lidar-based Mapping of Forest Volume and Biomass by Taxonomic Group Using Structurally Homogenous Segments , 2008 .