The Potential of Multisource Remote Sensing for Mapping the Biomass of a Degraded Amazonian Forest

Abstract

In the agricultural frontiers of Brazil, the distinction between forested and deforested lands traditionally used to map the state of the Amazon does not reflect the reality of the forest situation. A whole gradient exists for these forests, spanning from well conserved to severely degraded. For decision makers, there is an urgent need to better characterize the status of the forest resource at the regional scale. Until now, few studies have been carried out on the potential of multisource, freely accessible remote sensing for modelling and mapping degraded forest structural parameters such as aboveground biomass (AGB). The aim of this article is to address that gap and to evaluate the potential of optical (Landsat, MODIS) and radar (ALOS-1 PALSAR, Sentinel-1) remote sensing sources in modelling and mapping forest AGB in the old pioneer front of Paragominas municipality (Para state). We derived a wide range of vegetation and textural indices and combined them with in situ collected AGB data into a random forest regression model to predict AGB at a resolution of 20 m. The model explained 28% of the variance with a root mean square error of 97.1 Mg·ha −1 and captured all spatial variability. We identified Landsat spectral unmixing and mid-infrared indicators to be the most robust indicators with the highest explanatory power. AGB mapping reveals that 87% of forest is degraded, with illegal logging activities, impacted forest edges and other spatial distribution of AGB that are not captured with pantropical datasets. We validated this map with a field-based forest degradation typology built on canopy height and structure observations. We conclude that the modelling framework developed here combined with high-resolution vegetation status indicators can help improve the management of degraded forests at the regional scale. IntroductionDeforestation and forest degradation are major sources of greenhouse gas emissions [1,2], contributing to forest carbon losses [3], global climate change, affecting biodiversity [4] and the entire forest ecosystem. While deforestation refers to the rapid conversion from forest to non-forest areas, degradation implies changes in the forest structure with no change in land use [5,6]. In Amazonia, over the last decades, deforestation and forest degradation have shaped the rural landscape, resulting in a complex mosaic of fragmented forests associated with agricultural lands [7]. A total area of 766,448.5 km 2 was cleared in 2015 [8], representing 20% of Amazonia [9,10].Since 2005, deforestation in Brazil has drastically decreased thanks to coercive measures taken by the Brazilian government associated with private initiatives (soy and beef moratoria), among other factors [11]. However, these measures are not effective for reducing forest degradation [12,13]. Most of the remaining forested lands are degraded due to the accumulation over time and space of severe degradation processes mainly triggered by anthropogenic impacts through unsustainable logging practices, fire, shifting cultivation ...

Estimation of Forest Above-Ground Biomass by Geographically Weighted Regression and Machine Learning with Sentinel Imagery

“…By comparing the results of the correlation analysis, the coefficients from GWR, and the attribute importance from RF, it was indicated that texture characteristics of Sentinel-1 had great potential for estimating AGB, which was also shown in previous studies [98,99]. Additionally, it was a pioneering finding that the vegetation biophysical variables of Sentinel-2 were very helpful for AGB estimation using a local regression, which was found previously by non-parametric prediction [55].…”

Estimation of Forest Biomass in Beijing (China) Using Multisource Remote Sensing and Forest Inventory Data

“…Synthetic Aperture Radar (SAR) data, such as L-band Advanced Land Observing Satellite/Phased Array L-band Synthetic Aperture Radar (ALOS/PALSAR) [19] and X-band TerraSAR-X data, are widely used in the estimation of forest biomass [20,21]. SAR is not affected by illumination and climate conditions and it can penetrate vegetation to obtain information, covering relatively large areas in a short period of time [20].At the meso-scale, many studies have demonstrated the potential of optical and radar remote sensing-derived indicators to estimate forest biomass [22,23]. However, there is a large range and many uncertainties of remote sensing.…”

“…At the meso-scale, many studies have demonstrated the potential of optical and radar remote sensing-derived indicators to estimate forest biomass [22,23]. However, there is a large range and many uncertainties of remote sensing.…”

Cumulative disturbances to assess forest degradation using spectral unmixing in the northeastern Amazon

“…Two field missions in May and August 2015 made it possible to build a typology of degraded forests using GPS data acquisition and qualitative descriptions (see further details in Bourgoin et al., ). Based on the spatial distribution of the data, Google Earth images and Esri World Imagery (July 2016), the plot reference database was completed by visual interpretation of disturbance impacts.…”

“…In the same area, which represents an old Amazon agricultural frontier with a mosaic of degraded forest, among a number of other variables, Bourgoin et al. () classified Landsat spectral unmixing using CLASlite and mid‐infrared variables as the most robust and most explanatory variables of the aboveground biomass remaining in degraded forest. With these advances, there is now a need to study the cumulative impacts of forest damage over time.…”