Using Landsat-8 and Sentinel-1 data for Above Ground Biomass assessment in the Tamar valley and Dartmoor

Alboabidallah, Ahmed H. H.; Martín, José Manuel Sánchez; Lavender, Samantha; Abbott, Victor

doi:10.1109/multi-temp.2017.8035247

Cited by 4 publications

(2 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a way to overcome these constrains, remote sensing (RS)-based approaches for forest parameters estimation have gained wide attention, as an ever-growing variety of Earth Observation (EO) sensors and techniques has been developed and made available throughout the years [11] [12] [13] [14]. In this context, Light Detection and Ranging (LiDAR) systems represent the most straightforward alternative to replace direct approaches, as the height of the canopies can be directly inferred from the time-of-flight of the laser signal returns.…”

Section: Introductionmentioning

confidence: 99%

A Deep Learning Framework for the Estimation of Forest Height From Bistatic TanDEM-X Data

Carcereri,

Rizzoli,

Ienco

et al. 2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Up-to-date canopy height model (CHM) estimates are of key importance for forest resources monitoring and disturbance analysis. In this work we present a study on the potential of Deep Learning (DL) for the regression of forest height from TanDEM-X bistatic interferometric (InSAR) data. We propose a novel fully convolutional neural network (CNN) framework, trained in a supervised manner using reference CHM measurements derived from the LiDAR LVIS airborne sensor from NASA. The reference measurements were acquired during the joint NASA-ESA 2016 AfriSAR campaign over five sites in Gabon, Africa, characterized by the presence of different kinds of vegetation, spanning from tropical primary forests to mangroves. Together with the DL architecture and training strategy, we present a series of experiments to assess the impact of different input features on the network estimation accuracy (in particular of bistatic InSAR-related ones). When tested on all considered sites, the proposed DL model achieves an overall performance of 1.46m mean error, 4.2m mean absolute error and 15.06% mean absolute percentage error. Furthermore, we perform a spatial transfer analysis aimed at deriving preliminary insights on the generalization capability of the network when trained and tested on data sets acquired over different locations, combining different kinds of tropical vegetation. The obtained results are promising and already in line with state-of-the-art methods based on both physical-based modelling and data-driven approaches, with the remarkable advantage of requiring only one single TanDEM-X acquisition at inference time.

show abstract

Section: Introductionmentioning

confidence: 99%

A Deep Learning Framework for the Estimation of Forest Height From Bistatic TanDEM-X Data

Carcereri,

Rizzoli,

Ienco

et al. 2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…There are studies which use estimates of the AGB obtained with field-measured dasometric information (as a response variable) to calibrate images from sensors such as Landsat (Gizachew et al, 2016;Vargas-Larreta et al, 2017), Ikonos 2 (Phua et al, 2012), Radar (Sinha et al, 2015), ALS (Peuhkurinen et al, 2008), Sentinel (Alboabidallah et al, 2017) or a combination of sensors for estimating AGB in larger areas and in shorter periods of time than the NFI (GFOI, 2016).…”

Section: I2222 Extrapolation Of the Agb Estimationmentioning

confidence: 99%

Quantifying the uncertainty caused by sampling, modeling, and field measurements in the estimation of AGB with information of the national forest inventory in Durango, Mexico

Trucíos-Caciano¹

View full text Add to dashboard Cite

Aboveground biomass (AGB) is related to the carbon content of the forest and forest carbon is a core variable for contemporary forest management and policy decisions. A credible and precise estimate of AGB is a prerequisite for the scientifically proper implementation of commitments made regarding the reduction of forest carbon emissions. With datasets of the Mexican National Forest Inventory (MNFI), this thesis estimates AGB in the temperate forests of Durango, a federal state of Mexico, evaluating the uncertainty of the estimate of total AGB and how different error sources contribute to the overall error. In addition to AGB, also the basal area was analyzed which is closely correlated to AGB but does not have the error source of coming from the application of allometric models. SummaryVI almost be ignoredalways. However, assuming that measurement and model errors are random errors and biases are absent.Applying re-measurements was an efficient way to estimate and describe the measurement errors in DBH and TH. Through the application of the GUM Method, error propagation is decomposed into sources and processes, and it is better understood how uncertainties are combined. The Monte-Carlo simulation Method (MCM) also proves to be an effective, practical and reliable way to approximate the total AGB uncertainty estimate with acceptable ranges of probable error at scales of the MNFI. The results in the error propagation by the GUM Method and by the MCM are equivalent.We found that TH measurement errors have a greater contribution than DBH measurement errors at the tree-level. Furthermore, when the main contribution to the uncertainty at tree-level comes from uAM instead of uMes, then the total estimate of uNS at stand-level is proportional to the number of the trees. However, a ratio of uMes>uAM produces a total uNS estimate at stand level that is proportional to the contribution according to the size of the tree. Therefore, in this last relation, a greater contribution to the total NS estimate is made by the trees with the largest AGB estimated.

show abstract

DInSAR method applied to dual-pair interferograms with Sentinel-1 data: a study case on inconsistent unwrapping outputs

Téllez-Quiñones

Salazar-Garibay

Valdiviezo-Navarro

et al. 2020

International Journal of Remote Sensing

View full text Add to dashboard Cite

Using Landsat-8 and Sentinel-1 data for Above Ground Biomass assessment in the Tamar valley and Dartmoor

Cited by 4 publications

References 24 publications

A Deep Learning Framework for the Estimation of Forest Height From Bistatic TanDEM-X Data

A Deep Learning Framework for the Estimation of Forest Height From Bistatic TanDEM-X Data

Quantifying the uncertainty caused by sampling, modeling, and field measurements in the estimation of AGB with information of the national forest inventory in Durango, Mexico

DInSAR method applied to dual-pair interferograms with Sentinel-1 data: a study case on inconsistent unwrapping outputs

Contact Info

Product

Resources

About