Deforestation contributes 6-17% of global anthropogenic CO 2 emissions to the atmosphere 1 . Large uncertainties in emission estimates arise from inadequate data on the carbon density of forests 2 and the regional rates of deforestation. Consequently there is an urgent need for improved data sets that characterize the global distribution of aboveground biomass, especially in the tropics. Here we use multi-sensor satellite data to estimate aboveground live woody vegetation carbon density for pan-tropical ecosystems with unprecedented accuracy and spatial resolution. Results indicate that the total amount of carbon held in tropical woody vegetation is 228.7 Pg C, which is 21% higher than the amount reported in the Global Forest Resources Assessment 2010 (ref. 3). At the national level, Brazil and Indonesia contain 35% of the total carbon stored in tropical forests and produce the largest emissions from forest loss. Combining estimates of aboveground carbon stocks with regional deforestation rates 4 we estimate the total net emission of carbon from tropical deforestation and land use to be 1.0 Pg C yr −1 over the period 2000-2010-based on the carbon bookkeeping model. These new data sets of aboveground carbon stocks will enable tropical nations to meet their emissions reporting requirements (that is, United Nations Framework Convention on Climate Change Tier 3) with greater accuracy.When forests are cleared, carbon stored above and below ground in leaves, branches, stems and roots is released to the atmosphere. As a consequence, forest clearing, especially in the tropics, is a major source of CO 2 to the atmosphere. Although the proportion of carbon stored in forests comprises 70-80% of total terrestrial carbon 5 , the spatial and temporal variability in carbon storage is substantial 6 . This variability arises from natural and anthropogenic disturbances, as well as differences in stand age, topography, soils and climate. Globally, soils hold two to three times more carbon than that stored above ground in forest vegetation, but with the exception of cultivation, peatland fires and thawing permafrost, much of the carbon in soils is physically and chemically protected and not easily oxidized 7 . In contrast, carbon stored in aboveground biomass is readily mobilized by disturbance processes such as fire, wind throw, pest outbreaks and land conversion 8 .Efforts to quantify the amount of carbon stored in aboveground biomass over large areas of the tropics have been fraught with uncertainty. For example, estimates of aboveground carbon storage in tropical African forests vary by over ref. 9). In turn, the lack of reliable estimates of forest carbon storage introduces large uncertainties into estimates of terrestrial carbon emissions 10-14 . In Amazonia, recent studies have suggested

Observations from the moderate resolution imaging spectroradiometer (MODIS) were used in combination with a large data set of field measurements to map woody above-ground biomass (AGB) across tropical Africa. We generated a best-quality cloud-free mosaic of MODIS satellite reflectance observations for the period 2000-2003 and used a regression tree model to predict AGB at 1 km resolution. Results based on a cross-validation approach show that the model explained 82% of the variance in AGB, with a root mean square error of 50.5 Mg ha −1 for a range of biomass between 0 and 454 Mg ha −1 . Analysis of lidar metrics from the Geoscience Laser Altimetry System (GLAS), which are sensitive to vegetation structure, indicate that the model successfully captured the regional distribution of AGB. The results showed a strong positive correlation (R 2 = 0.90) between the GLAS height metrics and predicted AGB.

Mapping and monitoring carbon stocks in forested regions of the world, particularly the tropics, has attracted a great deal of attention in recent years as deforestation and forest degradation account for up to 30% of anthropogenic carbon emissions, and are now included in climate change negotiations. We review the potential for satellites to measure carbon stocks, specifically aboveground biomass (AGB), and provide an overview of a range of approaches that have been developed and used to map AGB across a diverse set of conditions and geographic areas. We provide a summary of types of remote sensing measurements relevant to mapping AGB, and assess the relative merits and limitations of each. We then provide an overview of traditional techniques of mapping AGB based on ascribing field measurements to vegetation or land cover type classes, and describe the merits and limitations of those relative to recent data mining algorithms used in the context of an approach based on direct utilization of remote sensing measurements, whether optical or lidar reflectance, or radar backscatter. We conclude that while satellite remote sensing has often been discounted as inadequate for the task, attempts to map AGB without satellite imagery are insufficient. Moreover, the direct remote sensing approach provided more coherent maps of AGB relative to traditional approaches. We demonstrate this with a case study focused on continental Africa and discuss the work in the context of reducing uncertainty for carbon monitoring and markets.

BackgroundImproved maps of species distributions are important for effective management of wildlife under increasing anthropogenic pressures. Recent advances in lidar and radar remote sensing have shown considerable potential for mapping forest structure and habitat characteristics across landscapes. However, their relative efficacies and integrated use in habitat mapping remain largely unexplored. We evaluated the use of lidar, radar and multispectral remote sensing data in predicting multi-year bird detections or prevalence for 8 migratory songbird species in the unfragmented temperate deciduous forests of New Hampshire, USA.Methodology and Principal FindingsA set of 104 predictor variables describing vegetation vertical structure and variability from lidar, phenology from multispectral data and backscatter properties from radar data were derived. We tested the accuracies of these variables in predicting prevalence using Random Forests regression models. All data sets showed more than 30% predictive power with radar models having the lowest and multi-sensor synergy (“fusion”) models having highest accuracies. Fusion explained between 54% and 75% variance in prevalence for all the birds considered. Stem density from discrete return lidar and phenology from multispectral data were among the best predictors. Further analysis revealed different relationships between the remote sensing metrics and bird prevalence. Spatial maps of prevalence were consistent with known habitat preferences for the bird species.Conclusion and SignificanceOur results highlight the potential of integrating multiple remote sensing data sets using machine-learning methods to improve habitat mapping. Multi-dimensional habitat structure maps such as those generated from this study can significantly advance forest management and ecological research by facilitating fine-scale studies at both stand and landscape level.

Recent advances in remote sensing and ecological modeling warrant a timely and robust investigation of the ecological variables that underlie large-scale patterns of breeding bird species richness, particularly in the context of intensifying land use and climate change. Our objective was to address this need using an array of bioclimatic and remotely sensed data sets representing vegetation properties and structure, and other aspects of the physical environment. We first build models of bird species richness across breeding bird survey (BBS) routes, and then spatially predict richness across the coterminous US at moderately high spatial resolution (1 km). Predictor variables were derived from various sources and maps of species richness were generated for four groups (guilds) of birds with different breeding habitat affiliation (forest, grassland, open woodland, scrub/shrub), as well as all guilds combined. Predictions of forest bird distributions were strong (R 2 = 0.85), followed by grassland (0.76), scrub/shrub (0.63) and open woodland (0.60) species. Vegetation properties were generally the strongest determinants of species richness, whereas bioclimatic and lidar-derived vertical structure metrics were of variable importance and dependent upon the guild type. Environmental variables (climate and the physical environment) were also frequently selected predictors, but canopy structure variables were not as important as expected based on more local to regional scale studies. Relatively sparse sampling of canopy structure metrics from the satellite lidar sensor may have reduced their importance relative to other predictor variables across the study domain. We discuss these results in the context of the ecological drivers of species richness patterns, the spatial scale of bird diversity analyses, and the potential of next generation space-borne lidar systems relevant to vegetation and ecosystem studies. This study strengthens current understanding of bird species-climate-vegetation relationships, which could be further advanced with improved canopy structure information across spatial scales.