Aboveground Forest Biomass Estimation with Landsat and LiDAR Data and Uncertainty Analysis of the Estimates

Landsat Thematic mapper (TM) image has long been the dominate data source, and recently LiDAR has offered an important new structural data stream for forest biomass estimations. On the other hand, forest biomass uncertainty analysis research has only recently obtained sufficient attention due to the difficulty in collecting reference data. This paper provides a brief overview of current forest biomass estimation methods using both TM and LiDAR data. A case study is then presented that demonstrates the forest biomass estimation methods and uncertainty analysis. Results indicate that Landsat TM data can provide adequate biomass estimates for secondary succession but are not suitable for mature forest biomass estimates due to data saturation problems. LiDAR can overcome TM’s shortcoming providing better biomass estimation performance but has not been extensively applied in practice due to data availability constraints. The uncertainty analysis indicates that various sources affect the performance of forest biomass/carbon estimation. With that said, the clear dominate sources of uncertainty are the variation of input sample plot data and data saturation problem related to optical sensors. A possible solution to increasing the confidence in forest biomass estimates is to integrate the strengths of multisensor data.

Download Full-text

Forest biomass estimation from airborne LiDAR data using machine learning approaches

Remote Sensing of Environment ◽

10.1016/j.rse.2012.07.006 ◽

2012 ◽

Vol 125 ◽

pp. 80-91 ◽

Cited By ~ 144

Author(s):

Colin J. Gleason ◽

Jungho Im

Keyword(s):

Machine Learning ◽

Forest Biomass ◽

Airborne Lidar ◽

Biomass Estimation ◽

Lidar Data ◽

Learning Approaches ◽

Airborne Lidar Data

Download Full-text

Deep Learning Based Retrieval of Forest Aboveground Biomass from Combined LiDAR and Landsat 8 Data

Remote Sensing ◽

10.3390/rs11121459 ◽

2019 ◽

Vol 11 (12) ◽

pp. 1459 ◽

Cited By ~ 22

Author(s):

Linjing Zhang ◽

Zhenfeng Shao ◽

Jianchen Liu ◽

Qimin Cheng

Keyword(s):

Remote Sensing ◽

Deep Learning ◽

Mean Squared Error ◽

Forest Biomass ◽

The Other ◽

Biomass Estimation ◽

Lidar Data ◽

Landsat 8 ◽

Squared Error ◽

Forest Aboveground Biomass

Estimation of forest aboveground biomass (AGB) is crucial for various technical and scientific applications, ranging from regional carbon and bioenergy policies to sustainable forest management. However, passive optical remote sensing, which is the most widely used remote sensing data for retrieving vegetation parameters, is constrained by spectral saturation problems and cloud cover. On the other hand, LiDAR data, which have been extensively used to estimate forest structure attributes, cannot provide sufficient spectral information of vegetation canopies. Thus, this study aimed to develop a novel synergistic approach to estimating biomass by integrating LiDAR data with Landsat 8 imagery through a deep learning-based workflow. First the relationships between biomass and spectral vegetation indices (SVIs) and LiDAR metrics were separately investigated. Next, two groups of combined optical and LiDAR indices (i.e., COLI1 and COLI2) were designed and explored to identify their performances in biomass estimation. Finally, five prediction models, including K-nearest Neighbor, Random Forest, Support Vector Regression, the deep learning model, i.e., Stacked Sparse Autoencoder network (SSAE), and multiple stepwise linear regressions, were individually used to estimate biomass with input variables of different scenarios, i.e., (i) all the COLI1 (ACOLI1), (ii) all the COLI2 (ACOLI2), (iii) ACOLI1 and all the optical (AO) and LiDAR variables (AL), and (iv) ACOLI2, AO and AL. Results showed that univariate models with the combined optical and LiDAR indices as explanatory variables presented better modeling performance than those with either optical or LiDAR data alone, regardless of the combination mode. The SSAE model obtained the best performance compared to the other tested prediction algorithms for the forest biomass estimation. The best predictive accuracy was achieved by the SSAE model with inputs of combined optical and LiDAR variables (i.e., ACOLI1, AO and AL) that yielded an R2 of 0.935, root mean squared error (RMSE) of 15.67 Mg/ha, and relative root mean squared error (RMSEr) of 11.407%. It was concluded that the presented combined indices were simple and effective by integrating LiDAR-derived structure information with Landsat 8 spectral data for estimating forest biomass. Overall, the SSAE model with inputs of Landsat 8 and LiDAR integrated information resulted in accurate estimation of forest biomass. The presented modeling workflow will greatly facilitate future forest biomass estimation and carbon stock assessments.

Download Full-text

Fusion of airborne LiDAR data and hyperspectral imagery for aboveground and belowground forest biomass estimation

Ecological Indicators ◽

10.1016/j.ecolind.2016.10.001 ◽

2017 ◽

Vol 73 ◽

pp. 378-387 ◽

Cited By ~ 48

Author(s):

Shezhou Luo ◽

Cheng Wang ◽

Xiaohuan Xi ◽

Feifei Pan ◽

Dailiang Peng ◽

...

Keyword(s):

Hyperspectral Imagery ◽

Forest Biomass ◽

Airborne Lidar ◽

Biomass Estimation ◽

Lidar Data ◽

Airborne Lidar Data

Download Full-text

Forest biomass estimation over three distinct forest types using TanDEM-X InSAR data and simulated GEDI lidar data

Remote Sensing of Environment ◽

10.1016/j.rse.2019.111283 ◽

2019 ◽

Vol 232 ◽

pp. 111283 ◽

Cited By ~ 13

Author(s):

Wenlu Qi ◽

Svetlana Saarela ◽

John Armston ◽

Göran Ståhl ◽

Ralph Dubayah

Keyword(s):

Forest Biomass ◽

Biomass Estimation ◽

Lidar Data ◽

Forest Types

Download Full-text

SUBTROPICAL FOREST BIOMASS ESTIMATION USING AIRBORNE LiDAR AND HYPERSPECTRAL DATA

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprs-archives-xli-b8-747-2016 ◽

2016 ◽

Vol XLI-B8 ◽

pp. 747-749

Author(s):

Yong Pang ◽

Zengyuan Li

Keyword(s):

Forest Ecosystem ◽

Forest Biomass ◽

Subtropical Forest ◽

Hyperspectral Data ◽

Airborne Lidar ◽

Observation System ◽

Biomass Estimation ◽

Lidar Data ◽

Forest Types ◽

Waveform Lidar

Forests have complex vertical structure and spatial mosaic pattern. Subtropical forest ecosystem consists of vast vegetation species and these species are always in a dynamic succession stages. It is very challenging to characterize the complexity of subtropical forest ecosystem. In this paper, CAF’s (The Chinese Academy of Forestry) LiCHy (LiDAR, CCD and Hyperspectral) Airborne Observation System was used to collect waveform Lidar and hyperspectral data in Puer forest region, Yunnan province in the Southwest of China. The study site contains typical subtropical species of coniferous forest, evergreen broadleaf forest, and some other mixed forests. The hypersectral images were orthorectified and corrected into surface reflectance with support of Lidar DTM product. The fusion of Lidar and hyperspectral can classify dominate forest types. The lidar metrics improved the classification accuracy. Then forest biomass estimation was carried out for each dominate forest types using waveform Lidar data, which get improved than single Lidar data source.

Download Full-text

A Comparison of Three Biomass Estimation Methods: A Case Study of Pinus tabulaeformis Forests in China

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.726-731.4237 ◽

2013 ◽

Vol 726-731 ◽

pp. 4237-4240

Author(s):

Fei Li ◽

Zhong Yu Wang ◽

Hua Yong Zhang ◽

Yi Xin Xu ◽

Lu Han

Keyword(s):

Linear Model ◽

Forest Biomass ◽

Pinus Tabulaeformis ◽

Estimation Methods ◽

Biomass Estimation ◽

Hyperbolic Model ◽

Power Model ◽

Regression Equations ◽

Relative Errors

Power model, linear model and hyperbolic model were commonly used to estimate forest biomass via stand volume, however the relative accuracy is unclear forPinus tabulaeformisforests in China. In order to compare the accuracies of these models, data from 130Pinus tabulaeformisforest stands were compiled from published literatures. Data of 100 stands were randomly selected to establish regression equations, the other 30 data were used to compare the accuracies of equations either established in this study or in previous studies. The results show that biomass ofPinus tabulaeformisforests could be well estimated by power model and linear model, while hyperbolic model is likely to result in enormous overestimation or underestimation. The mean relative errors of the power model and linear model established in this study are-0.3% and 1.8% respectively. In comparison with models established by previous studies, these two models have better prediction accuracies.

Download Full-text

SUBTROPICAL FOREST BIOMASS ESTIMATION USING AIRBORNE LiDAR AND HYPERSPECTRAL DATA

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprsarchives-xli-b8-747-2016 ◽

2016 ◽

Vol XLI-B8 ◽

pp. 747-749

Author(s):

Yong Pang ◽

Zengyuan Li

Keyword(s):

Forest Ecosystem ◽

Forest Biomass ◽

Subtropical Forest ◽

Hyperspectral Data ◽

Airborne Lidar ◽

Observation System ◽

Biomass Estimation ◽

Lidar Data ◽

Forest Types ◽

Waveform Lidar

Download Full-text

Above ground biomass estimation across forest types at different degradation levels in Central Kalimantan using LiDAR data

International Journal of Applied Earth Observation and Geoinformation ◽

10.1016/j.jag.2012.01.010 ◽

2012 ◽

Vol 18 ◽

pp. 37-48 ◽

Cited By ~ 72

Author(s):

Karin Kronseder ◽

Uwe Ballhorn ◽

Viktor Böhm ◽

Florian Siegert

Keyword(s):

Biomass Estimation ◽

Lidar Data ◽

Above Ground Biomass ◽

Forest Types ◽

Central Kalimantan ◽

Ground Biomass

Download Full-text

Tropical-Forest Biomass Estimation at X-Band From the Spaceborne TanDEM-X Interferometer

IEEE Geoscience and Remote Sensing Letters ◽

10.1109/lgrs.2014.2334140 ◽

2015 ◽

Vol 12 (2) ◽

pp. 239-243 ◽

Cited By ~ 54

Author(s):

Robert Treuhaft ◽

Fabio Gonzalves ◽

Joao Roberto dos Santos ◽

Michael Keller ◽

Michael Palace ◽

...

Keyword(s):

Tropical Forest ◽

Forest Biomass ◽

Biomass Estimation ◽

X Band

Download Full-text

Flood estimation for ungauged catchments in the Philippines using multiple archival data records

10.5194/egusphere-egu21-4905 ◽

2021 ◽

Author(s):

Trevor Hoey ◽

Pamela Tolentino ◽

Esmael Guardian ◽

Richard Williams ◽

Richard Boothroyd ◽

...

Keyword(s):

Catchment Area ◽

Daily Rainfall ◽

The Philippines ◽

Data Availability ◽

Estimation Methods ◽

Annual Maximum ◽

Ungauged Catchments ◽

Pearson Type ◽

Flood Estimation ◽

Best Fit

Assessment of flood and drought risks, and changes to these risks under climate change, is a critical issue worldwide. Statistical methods are commonly used in data-rich regions to estimate the magnitudes of river floods of specified return period at ungauged sites. However, data availability can be a major constraint on reliable estimation of flood and drought magnitudes, particularly in the Global South. Statistical flood and drought magnitude estimation methods rely on the availability of sufficiently long data records from sites that are representative of the hydrological region of interest. In the Philippines, although over 1000 locations have been identified where flow records have been collected at some time, very few records exist of over 20 years duration and only a limited number of sites are currently being gauged. We collated data from three archival sources: (1) Division of Irrigation, Surface Water Supply (SWS) (1908-22; 257 sites in total); (2) Japan International Cooperation Agency (JICA) (1955-91; 90 sites); and, (3) Bureau of Research and Standards (BRS) (1957-2018; 181 sites). From these data sets, 176 contained sufficiently long and high quality records to be analysed. Series of annual maximum floods were fit using L-moments with Weibull, Log-Pearson Type III and Generalised Logistic Distributions, the best-fit of these being used to estimate 2-, 10- and 100-year flood events, Q2, Q10 and Q100. Predictive equations were developed using catchment area, several measures of annual and extreme precipitation, catchment geometry and land-use. Analysis took place nationally, and also for groups of hydrologically similar regions, based on similar flood growth curve shapes, across the Philippines. Overall, the best fit equations use a combination of two predictor variables, catchment area and the median annual maximum daily rainfall. The national equations have R2 of 0.55-0.65, being higher for shorter return periods, and regional groupings R2 are 0.60-0.77 for Q10. These coefficients of determination, R2, are lower than in some comprehensive studies worldwide reflecting in part the short individual flow records. Standard errors of residuals for the equations are between 0.19 and 0.51 (log10 units), which lead to significant uncertainty in flood estimation for water resource and flood risk management purposes. Improving the predictions requires further analysis of hydrograph shape across the different climate types, defined by seasonal rainfall distributions, in the Philippines and between catchments of different size. The results here represent the most comprehensive study to date of flood magnitudes in the Philippines and are being incorporated into guidance for river managers alongside new assessments of river channel change across the country. The analysis illustrates the potential, and the limitations, for combining information from multiple data sources and short individual records to generate reliable estimates of flow extremes.

Download Full-text