Estimating the Above-Ground Biomass in Miombo Savanna Woodlands (Mozambique, East Africa) Using L-Band Synthetic Aperture Radar Data

The quantification of forest above-ground biomass (AGB) is important for such broader applications as decision making, forest management, carbon (C) stock change assessment and scientific applications, such as C cycle modeling. However, there is a great uncertainty related to the estimation of forest AGB, especially in the tropics. The main goal of this study was to test a combination of field data and Advanced Land Observing Satellite (ALOS) Phased Array L-band Synthetic Aperture Radar (PALSAR) backscatter intensity data to reduce the uncertainty in the estimation of forest AGB in the Miombo savanna woodlands of Mozambique (East Africa). A machine learning algorithm, based on bagging stochastic gradient boosting (BagSGB), was used to model forest AGB as a function of ALOS PALSAR Fine Beam Dual (FBD) backscatter intensity metrics. The application of this method resulted in a coefficient of correlation (R) between observed and predicted (10-fold cross-validation) forest AGB values of 0.95 and a root mean square error of 5.03 Mg·ha−1. However, as a consequence of using bootstrap samples in combination with a cross validation procedure, some bias may have been introduced, and the reported cross validation statistics could be overoptimistic. Therefore and as a consequence of the BagSGB model, a measure of prediction variability (coefficient of variation) on a pixel-by-pixel basis was also produced, with values ranging from 10 to 119% (mean = 25%) across the study area. It provides additional and complementary information regarding the spatial distribution of the error resulting from the application of the fitted model to new observations.

[1]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[2]  James H. Torrie,et al.  Principles and procedures of statistics: a biometrical approach (2nd ed) , 1980 .

[3]  Manabu Watanabe,et al.  ALOS PALSAR: A Pathfinder Mission for Global-Scale Monitoring of the Environment , 2007, IEEE Transactions on Geoscience and Remote Sensing.

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

[5]  Casey M. Ryan,et al.  Carbon sequestration and biodiversity of re-growing miombo woodlands in Mozambique , 2008 .

[6]  J. V. Soares,et al.  Distribution of aboveground live biomass in the Amazon basin , 2007 .

[7]  Richard E. Lewis,et al.  At the heart of REDD+: a role for local people in monitoring forests? , 2011 .

[8]  I. Woodhouse,et al.  Quantifying small‐scale deforestation and forest degradation in African woodlands using radar imagery , 2012 .

[9]  Leo Breiman,et al.  Classification and Regression Trees , 1984 .

[10]  Trevor Hastie,et al.  The Elements of Statistical Learning , 2001 .

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

[12]  Bruce M. Campbell,et al.  Beyond Copenhagen: REDD+, agriculture, adaptation strategies and poverty , 2009 .

[13]  Damien Sulla-Menashe,et al.  MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets , 2010 .

[14]  John A. Richards,et al.  Remote Sensing with Imaging Radar , 2009 .

[15]  J. Carreiras,et al.  Assessing the extent of agriculture/pasture and secondary succession forest in the Brazilian Legal Amazon using SPOT VEGETATION data , 2006 .

[16]  Birger Solberg,et al.  Estimation of biomass and volume in Miombo Woodland at Kitulangalo Forest Reserve, Tanzania , 1994 .

[17]  Ha Henny Romijn.  Land clearing and greenhouse gas emissions from Jatropha biofuels on African Miombo Woodlands , 2011 .

[18]  Casey M. Ryan,et al.  Above‐ and Belowground Carbon Stocks in a Miombo Woodland Landscape of Mozambique , 2011 .

[19]  S. Popescu,et al.  Satellite lidar vs. small footprint airborne lidar: Comparing the accuracy of aboveground biomass estimates and forest structure metrics at footprint level , 2011 .

[20]  I. Woodhouse.  Introduction to Microwave Remote Sensing , 2005 .

[21]  R. Houghton,et al.  Characterizing 3D vegetation structure from space: Mission requirements , 2011 .

[22]  David A. Landgrebe,et al.  A survey of decision tree classifier methodology , 1991, IEEE Trans. Syst. Man Cybern..

[23]  L. Montanarella,et al.  Estimating forest soil bulk density using boosted regression modelling , 2010 .

[24]  L. Breiman.  Arcing classifier (with discussion and a rejoinder by the author) , 1998 .

[25]  J. Friedman.  Stochastic gradient boosting , 2002 .

[26]  C. Brodley,et al.  Decision tree classification of land cover from remotely sensed data , 1997 .

[27]  J. Chambers,et al.  Tree allometry and improved estimation of carbon stocks and balance in tropical forests , 2005, Oecologia.

[28]  Eric Bauer,et al.  An Empirical Comparison of Voting Classification Algorithms: Bagging, Boosting, and Variants , 1999, Machine Learning.

[29]  N. H. Ravindranath,et al.  2006 IPCC Guidelines for National Greenhouse Gas Inventories , 2006 .

[30]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[31]  A. Marshall,et al.  Carbon storage, structure and composition of miombo woodlands in Tanzania's Eastern Arc Mountains , 2011 .

[32]  Masanobu Shimada,et al.  An Evaluation of the ALOS PALSAR L-Band Backscatter—Above Ground Biomass Relationship Queensland, Australia: Impacts of Surface Moisture Condition and Vegetation Structure , 2010, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[33]  Maxim Neumann,et al.  Assessing Performance of L- and P-Band Polarimetric Interferometric SAR Data in Estimating Boreal Forest Above-Ground Biomass , 2012, IEEE Transactions on Geoscience and Remote Sensing.

[34]  Erich Meier,et al.  Rigorous Derivation of Backscattering Coefficient. , 1994 .

[35]  Ariel E. Lugo,et al.  Biomass Estimation Methods for Tropical Forests with Applications to Forest Inventory Data , 1989, Forest Science.

[36]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[37]  E. Chidumayo,et al.  Miombo Ecology and Management: An Introduction , 1997 .

[38]  J. Carreiras,et al.  Understanding the relationship between aboveground biomass and ALOS PALSAR data in the forests of Guinea-Bissau (West Africa) , 2012 .

[39]  J. Carreiras,et al.  Greenhouse gas emissions from shifting cultivation in the tropics, including uncertainty and sensitivity analysis , 2011 .

[40]  J Elith,et al.  A working guide to boosted regression trees. , 2008, The Journal of animal ecology.

[41]  S. Goetz,et al.  Mapping and monitoring carbon stocks with satellite observations: a comparison of methods , 2009, Carbon balance and management.

[42]  Greg Ridgeway,et al.  Generalized Boosted Models: A guide to the gbm package , 2006 .

[43]  S. Goetz,et al.  Reply to Comment on ‘A first map of tropical Africa’s above-ground biomass derived from satellite imagery’ , 2008, Environmental Research Letters.

[44]  Michael G. Wing,et al.  Airborne Light Detection and Ranging (LiDAR) for Individual Tree Stem Location, Height, and Biomass Measurements , 2011, Remote. Sens..

[45]  L. Breiman.  Arcing Classifiers , 1998 .

[46]  B. Campbell.  The miombo in transition: woodlands and welfare in Africa. , 1996 .

[47]  S. Goetz,et al.  Importance of biomass in the global carbon cycle , 2009 .

[48]  Glenn De ' ath.  BOOSTED TREES FOR ECOLOGICAL MODELING AND PREDICTION , 2007 .

[49]  C. Kroeze.  N2O from animal waste. Methodology according to IPCC Guidelines for National Greenhouse Gas Inventories. , 1997 .

[50]  Corinne Le Quéré,et al.  Trends in the sources and sinks of carbon dioxide , 2009 .

[51]  M. Herold,et al.  Capacity development in national forest monitoring: Experiences and progress for REDD+ , 2012 .

[52]  F. Ulaby,et al.  Handbook of radar scattering statistics for terrain , 1989 .

[53]  Niall P. Hanan,et al.  Woody cover in African savannas: the role of resources, fire and herbivory , 2008 .

[54]  W. Cohen,et al.  Estimates of forest canopy height and aboveground biomass using ICESat , 2005 .

[55]  K. Tully,et al.  Untangling a Decline in Tropical Forest Resilience: Constraints on the Sustainability of Shifting Cultivation Across the Globe , 2010 .

[56]  J. Chambers,et al.  Regional ecosystem structure and function: ecological insights from remote sensing of tropical forests. , 2007, Trends in ecology & evolution.

[57]  Sassan Saatchi,et al.  Comment on ‘A first map of tropical Africa’s above-ground biomass derived from satellite imagery’ , 2011 .

[58]  Raymond J. Mooney,et al.  Combining Bias and Variance Reduction Techniques for Regression Trees , 2005, ECML.

[59]  Sassan Saatchi,et al.  Estimation of Forest Fuel Load From Radar Remote Sensing , 2007, IEEE Transactions on Geoscience and Remote Sensing.

[60]  K. Annan.  Center for International Forestry Research Center for International , 2001 .

[61]  R. DeFries,et al.  Classification trees: an alternative to traditional land cover classifiers , 1996 .

[62]  P. R. Bevington,et al.  Data Reduction and Error Analysis for the Physical Sciences , 1969 .

[63]  S. Quegan,et al.  Understanding Synthetic Aperture Radar Images , 1998 .

[64]  M. Herold,et al.  A step-wise framework for setting REDD+ forest reference emission levels and forest reference levels , 2012 .

[65]  Niklaus E. Zimmermann,et al.  Predicting tree species presence and basal area in Utah: A comparison of stochastic gradient boosting, generalized additive models, and tree-based methods , 2006 .

[66]  R. B. Jackson,et al.  A Large and Persistent Carbon Sink in the World’s Forests , 2011, Science.

[67]  W. Salas,et al.  Benchmark map of forest carbon stocks in tropical regions across three continents , 2011, Proceedings of the National Academy of Sciences.

[68]  J. Glenday.  Carbon storage and emissions offset potential in an African dry forest, the Arabuko-Sokoke Forest, Kenya , 2008, Environmental monitoring and assessment.