Global Forest Change Dataset Research Articles

AI-based Validation of Deforestation Using High-Resolution Satellite Imagery in the Brazilian Amazon

Abstract. Forests play an important role in the Earth systems for carbon sequestration and climate change mitigation, yet they have been increasingly disturbed by deforestation and forest degradation at an unprecedented pace. The Brazilian Amazon, for instance, experienced a 140% rise in deforestation from 2012 to 2020, with a record loss of 13,200 km2 between August 2020 and July 2021. Alarmingly, 87% of 2019 deforestation alerts occurred on private properties, with 61% in legally restricted areas. Existing deforestation monitoring systems, such as PRODES and the Global Forest Change dataset, use about 30m resolution satellite imagery, which is insufficient for operational validation at fine scales. The Deforestation Alert System by Imazon leverages high-resolution PlanetScope data (3-4m) but faces challenges due to fewer spectral bands and variations in reflectance values across different satellite sensors and dates. As a result, current validation is based mainly on manual inspection which is highly time-consuming. To address the challenges and automate the validation process, this work develops a system based on deep learning – known for its ability to capture complex texture patterns in high-resolution images – to inspect and confirm new deforestation sites. Specifically, the system takes inputs from potential deforestation sites suggested by coarse-resolution products and uses a pair of PlanetScope images before and after the change at each site to determine new deforestation (excluding existing deforestation). Our results demonstrate that the new system achieves robust and high-quality accuracy under various test conditions.

Conservation and Avoided Deforestation: Evidence from Protected Areas of Tanzania

Tanzania dedicates a substantial proportion (38%) of its territory to conservation, with a large number of Protected Areas (PAs) managed under various regimes. Nevertheless, the country still experiences high rates of deforestation, which threaten the ecological integrity and socio-economic benefits of its forests. We utilized the Global Forest Change Dataset (2012–2022) and implemented a Propensity Score Matching (PSM) approach followed by a series of binomial logit regression modeling. Our objectives were to evaluate (1) the likelihood of PAs in avoiding deforestation compared with unprotected forest landscapes, (2) the variability in effectiveness among the different PA management regimes in avoiding deforestation, (3) evidence of leakage, defined here as the displacement of deforestation beyond PA boundaries as a result of protection inside PAs. Our findings reveal that, despite ongoing deforestation within and outside of PAs, conservation efforts are, on average, three times more likely to avoid deforestation compared with unprotected landscapes. However, the effectiveness of avoiding deforestation significantly varies among the different management regimes. National Parks and Game Reserves are nearly ten times more successful in avoiding deforestation, likely because of the stringent set of regulations and availability of resources for implementation. Conversely, Nature Forest Reserves, Game Controlled Areas, and Forest Reserves are, on average, only twice as likely to avoid deforestation, indicating substantial room for improvement. We found little evidence of the overall leakage as a consequence of protection. These results highlight the mixed success of Tanzania’s conservation efforts, suggesting opportunities to enhance the effectiveness of many less protected PAs. We conclude by proposing potential strategic pathways to enhance further the climate and ecosystem benefits of conservation in Tanzania.

Deforestation detection from spaceborne full-waveform laser altimetry, incorporating terrain effects: A case study in Porto Velho, Brazil

Forest structure, notably the canopy height above the ground, can be obtained from the full-waveform data acquired by satellite laser altimetry. On sloped terrain, the laser pulse can simultaneously reflect from both the vegetation and the ground within their intersecting vertical extents, causing their signals to overlap. This overlap complicates the separation of canopy and ground in the analysis. Consequently, accurately obtaining canopy height and its variation, while considering the influence of terrain factors, poses a significant challenge.In this paper, to counter this, we introduce an innovative approach to remove terrain effects for more reliable identification of canopy and ground. Overlapping footprints from the Ice, Cloud, and land Elevation Satellite (ICESat)/Geoscience Laser Altimeter System (GLAS) and Global Ecosystem Dynamics Investigation (GEDI) campaigns are extracted as coincidental observations. Utilizing the different footprint sizes and by minimizing the center and width of each decomposed Gaussian component pair, optimized ground returns can be obtained. Validation using airborne LiDAR data from the Bartlett Experimental Forest, New Hampshire, USA, indicate that the proposed approach can reduce the ground elevation root-mean-square error (RMSE) by approximately 3 m, compared to the prevalent methods and standard products. While the GEDI product algorithms exhibit a certain accuracy in canopy height retrieval, the proposed approach presents a more constrained error band, signifying less impact from ground elevation discrepancies.While the proposed method enhances canopy height precision, its reliance on overlapping footprints means sacrificing some observations. However, the proposed method is effective in scenarios such as change detection that require repeat observations. As a case study, we applied the proposed method in Porto Velho, Rondônia, Brazil, which is a deforestation hotspot. Using the GlobeLand30 dataset, we investigated the land-cover dynamics from co-located laser footprints, identifying significant forest loss areas and validating our findings with the Hansen Global Forest Change dataset, achieving an accuracy of 83.81 %.In conclusion, this research emphasizes the significance of terrain topography effects in vegetation structure analysis, and positions satellite laser altimetry as instrumental for forest monitoring.

Insight into the imbalance of forest cover change at county level in mainland China during 2000–2020: From the perspective of subdividing forest cover change into forest gain and loss

The significant growth of China's forest resources presents a remarkable case in global vegetation dynamics. However, there is spatial imbalance in forest cover changes, which is not only reflected in regional imbalances, but also between total and net changes. There is still a lack of knowledge in this imbalance. In our comprehensive study, we subdivided forest cover change into forest gain and loss, and calculated the total and net forest change across 2850 county-level administrative units in mainland China, utilizing the Global Forest Change Dataset. The univariate linear regression and the Sen + Mann-Kendall test is utilized to explore the temporal trends in forest loss. Additionally, spatial autocorrelation models were employed to examine the spatial differentiation of forest cover change. Our findings unveiled a wide spectrum of forest loss rates ranging from 0% to 27%, and gain rates ranging from 0% to 10%, with the highest rates observed in the southern counties. We observed that the forest loss trends in southeastern counties showed significant and slight increases, while the northwestern counties experienced significant and slight decreases, with the Huhuanyong Line serving as the demarcation. In terms of spatial patterning, quite a number of the southern counties displayed a “large net change, large total change” state, which contrasted with diverse patterns observed in certain western regions. Spatial autocorrelation analysis showed that the High-high clusters of both forest gain and loss were primarily observed in southern counties, with the Low-low clusters in the northwest. This detailed analysis provides valuable insights into the dynamics of forest cover changes, offering a robust scientific foundation for developing informed and region-specific conservation policies.

A Combined Deep Learning and Prior Knowledge Constraint Approach for Large-Scale Forest Disturbance Detection Using Time Series Remote Sensing Data

The scale and severity of forest disturbances across the globe are increasing due to climate change and human activities. Remote sensing analysis using time series data is a powerful approach for detecting large-scale forest disturbances and describing detailed forest dynamics. Various large-scale forest disturbance detection algorithms have been proposed, but most of them are only suitable for detecting high-magnitude forest disturbances (e.g., fire, harvest). Conversely, more continuous, subtle, and gradual lower-magnitude forest disturbances (e.g., thinning, pests, and diseases) have been subject to less focus. Deep learning (DL) can distinguish subtle differences in information within time series data, offering new opportunities to capture forest disturbances in a complete and detailed way. This study proposes an approach for analyzing forest dynamics across large areas and long time periods by combining DL time series classification and prior knowledge constraint. The approach consists of two stages: (1) an improved self-attention model used for time series classification to identify sequences with forest disturbance characteristics; (2) developed skip-disturbance recovery index (S-DRI) characterizing the temporal context, using prior knowledge constraint to identify forest disturbance years in time series with disturbance characteristics. In this study, the year of forest disturbances in five study areas located in the United States, Canada, and Poland from 2001 to 2020 was mapped. A total of 3082 manually interpreted test data with different disturbance causal agents (such as fire, harvest, conversion, hurricane, and pests) were sampled from five research areas for validation. Our approach was also evaluated against two forest disturbance benchmark datasets derived from LandTrendr and the Global Forest Change (GFC) dataset. The results demonstrate that our approach achieved an overall accuracy of 87.8%, surpassing the accuracy of LandTrendr (84.6%) and the Global Forest Change dataset (81.4%). Furthermore, our approach demonstrated lower omission rates (ranging from 10.0% to 67.4%) in detecting subtle to severe causal agents of forest disturbance, in comparison to LandTrendr (with a range of 18.0% to 81.6%) and GFC (with a range of 15.0% to 88.8%). This study, which involved mapping large-scale and long-term forest disturbance in multiple regions, revealed that our approach can be applied to new areas without a requirement for complex parameter adjustments. These results demonstrate the potential of our approach in generating comprehensive and detailed forest disturbance data, thus providing a new and effective method in this domain.

Testing the Effect of Relative Pollen Productivity on the REVEALS Model: A Validated Reconstruction of Europe-Wide Holocene Vegetation

Reliable quantitative vegetation reconstructions for Europe during the Holocene are crucial to improving our understanding of landscape dynamics, making it possible to assess the past effects of environmental variables and land-use change on ecosystems and biodiversity, and mitigating their effects in the future. We present here the most spatially extensive and temporally continuous pollen-based reconstructions of plant cover in Europe (at a spatial resolution of 1° × 1°) over the Holocene (last 11.7 ka BP) using the ‘Regional Estimates of VEgetation Abundance from Large Sites’ (REVEALS) model. This study has three main aims. First, to present the most accurate and reliable generation of REVEALS reconstructions across Europe so far. This has been achieved by including a larger number of pollen records compared to former analyses, in particular from the Mediterranean area. Second, to discuss methodological issues in the quantification of past land cover by using alternative datasets of relative pollen productivities (RPPs), one of the key input parameters of REVEALS, to test model sensitivity. Finally, to validate our reconstructions with the global forest change dataset. The results suggest that the RPPs.st1 (31 taxa) dataset is best suited to producing regional vegetation cover estimates for Europe. These reconstructions offer a long-term perspective providing unique possibilities to explore spatial-temporal changes in past land cover and biodiversity.

Выявление уникальных озер с использованием геоинформационных систем на примере Ненецкого автономного округа

The possibilities of using geographic information systems (GIS) and cluster methods for identifying anomalies to identify unique lakes in the Nenets Autonomous Okrug (NAO) are demonstrated. The following tasks have been solved: 1) determination of the morphometric characteristics of lakes using methods of remote sensing of the Earth; 2) identification of anomalous morphometric characteristics by mathematical methods; 3) expert evaluation of the lakes resulting from the analysis to confirm their unique characteristics for the purpose of subsequent research and assigning them a special status. The relevance of the work is caused by the vastness and inaccessibility of the northern regions, which leads to the need for preliminary identification of objects that are most interesting for expeditionary research. In protected areas, objects that differ in their anomalous characteristics may be of particular interest. The test region of the study was limited by the boundaries of specially protected natural areas of the NAO. The deciphering of the lakes was carried out using the Global Forest Change data set. Raster processing and extraction of areal characteristics of water bodies were carried out in the QGIS software environment. The entire data set was divided into several groups according to the genetic category of the surface, which were also analyzed when identifying anomalies. This approach makes it possible to identify anomalous objects within a particular landscape. For data processing, the IBM SPSS Modeler software application was used, where the anomaly search is based on a two-stage clustering model. The search for anomalies by cluster methods is based on the fact that if the values of an instance are removed from the center of the cluster by more than a certain amount, then the object is considered an anomaly. As a result of applying the TwoStep Cluster algorithm to the sample of morphometric parameters of lakes, 42 anomalous objects were identified. The expert assessment confirmed that the identified lakes are of interest for further research. The final set included such well-known lakes as Golodnaya Guba, Peschanka-To, Kuznetskoe-To, as well as a number of small water bodies that stand out sharply for their peculiar characteristics in comparison with most of the lakes in the study region. For sparsely populated and logistically complex northern territories, the use of such an approach is an important element of field work planning.

Detection of Land-Use Change and Rapid Recovery of Vegetation after Deforestation in the Congo Basin

Abstract The Congo Basin is severely understudied compared to other tropical regions; this is partly due to the lack of meteorological stations and the ubiquitous cloudiness hampering the use of remote sensing products. Clustering of small-scale agricultural deforestation events within the basin may result in deforestation on scales that are atmospherically important. This study uses 500-m MODIS data and the Global Forest Change dataset (GFC) to detect deforestation at a monthly and subkilometer scale and to quantify how deforestation impacts vegetation proxies (VPs) within the basin, the time scales over which these changes persist, and how they are affected by the deforestation driver. Missing MODIS data meant that a new method, based on two-date image differencing, was developed to detect deforestation on a monthly scale. Evaluation against the yearly GFC data shows that the highest detection rate was 79% for clearing sizes larger than 500 m2. Recovery to predeforestation levels occurred faster than expected; analysis of postdeforestation evolution of the VPs found 66% of locations recovered within a year. Separation by land-cover type also showed unexpected regrowth, as over 50% of rural complex and plantation land recovered within a year. The fallow period in the study region was typically short; by the sixth year after the initial deforestation event, ∼88% of the locations underwent a further considerable drop. These results show the importance of fine spatial and temporal information to assess Congo Basin deforestation and highlight the large differences in the impacts of land-use change compared to other rain forests.

Biases and limitations of Global Forest Change and author-generated land cover maps in detecting deforestation in the Amazon.

Studying land use change in protected areas (PAs) located in tropical forests is a major conservation priority due to high conservation value (e.g., species richness and carbon storage) here, coupled with generally high deforestation rates. Land use change researchers use a variety of land cover products to track deforestation trends, including maps they produce themselves and readily available products, such as the Global Forest Change (GFC) dataset. However, all land cover maps should be critically assessed for limitations and biases to accurately communicate and interpret results. In this study, we assess deforestation in PA complexes located in agricultural frontiers in the Amazon Basin. We studied three specific sites: Amboró and Carrasco National Parks in Bolivia, Jamanxim National Forest in Brazil, and Tambopata National Reserve and Bahuaja-Sonene National Park in Peru. Within and in 20km buffer areas around each complex, we generated land cover maps using composites of Landsat imagery and supervised classification, and compared deforestation trends to data from the GFC dataset. We then performed a dissimilarity analysis to explore the discrepancies between the two remote sensing products. Both the GFC and our supervised classification showed that deforestation rates were higher in the 20km buffer than inside the PAs and that Jamanxim National Forest had the highest deforestation rate of the PAs we studied. However, GFC maps showed consistently higher rates of deforestation than our maps. Through a dissimilarity analysis, we found that many of the inconsistencies between these datasets arise from different treatment of mixed pixels or different parameters in map creation (for example, GFC does not detect reforestation after 2012). We found that our maps underestimated deforestation while GFC overestimated deforestation, and that true deforestation rates likely fall between our two estimates. We encourage users to consider limitations and biases when using or interpreting our maps, which we make publicly available, and GFC’s maps.

Estimation and Study of Forest Loss and Gain Using Spatial Dataset across Districts of Uttarakhand

Aims: To study and estimate the forest cover loss and gain across the 13 districts of Uttarakhand.
 Place and Duration of Study: Govind Ballabh Pant University of Agriculture and Technology, Pantnagar, between September 2021 and December 2021.
 Methodology: We extracted forest cover time-series data from the year 2001 to the year 2020 from Hensen Global Forest Change Dataset. This data was then mapped to the shapefile created in ARC-GIS containing all 13 districts as a Feature Collection, which was then used to individually classify each region and to estimate the size of the loss of tree cover precisely over the district boundary.
 Results: Our study shows forest loss of about (21,05,71,646 square meters) and forest gain of (6,00,79,072 square meters) cumulatively in all the districts of Uttarakhand from the year 2001 to 2020 at a spatial resolution of 30 meters where trees were identified as canopies greater than 5 meters in height.
 Conclusion: Among the districts of Uttarakhand Udham Singh Nagar, Nainital, and Champawat alone contribute to the total tree cover loss area of 15061.7513801 ha. which is about 71.5 % of Uttarakhand’s total tree cover loss. These regions require monitoring and controlling deforestation and more detailed studies like this are required to analyze and prevent the causes of such great-scale deforestation. Analyzing districts apart from those mentioned above, it is observed that the amount of tree cover loss is greater than the reforestation.

Protected areas have remarkable spillover effects on forest conservation on the Qinghai‐Tibet Plateau

AbstractAimAlthough protected areas (PAs) are assumed to reduce natural threats within boundaries, their spillover effects remain equivocal. It is necessary to determine whether PAs truly achieve conservation targets and whether they promote or inhibit natural habitat degradation in adjacent areas by blockage or leakage spillover. This study aims to choose 54 nature reserves (NRs) focusing on forest protection as a case study to assess PA conservation effectiveness and spillover prevalence.LocationPAs on the Qinghai‐Tibet Plateau (QTP).MethodsWe used matching methods to compare deforestation rates inside PAs and their 20 km buffer zones with matched control areas based on the Global Forest Change dataset from 2001 to 2019. We contrasted the effects of NRs with different management levels, ages and areas. We designed five concentric buffer zones to assess spillover change with distance and estimated potential drivers of the spillover effect to explain its directions and magnitudes.Results75.9% of the NRs were effective in preventing deforestation within their boundaries. NRs with different properties showed similar performance on forest conservation. Spillover effects were heterogeneous around NRs. One hundred and twenty‐two buffer zones had positive spillover ranging from 0.1% to 5.3%. One hundred and nineteen buffer zones had leakages from −8.84% to −0.1%. Blockages slightly outnumbered leakages at different distances, while leakages happened more frequently when we treated buffer zones as a whole spillover area. The linear model indicated NR age and population density of buffer zones were the most relevant predictors to spillover value.Main conclusionsMost PAs performed well in forest conservation. Leakages could undermine or offset PA conservation efforts and were related to multiple natural or socio‐economic factors. We recommend considering the plurality of PAs as well as spillover effect and incorporating a social‐ecological framework in further PA establishment and management.

Does leakage exist in China's typical protected areas? Evidence from 13 national nature reserves.

There is profound interest in knowing the degree to which the effectiveness of China's nature reserves, and whether leakage is common around the reserves, in the face of the most drastic conflicts between conservation and development in the world. To answer these questions, we employed the Landsat-derived Global Forest Change Dataset with 30-m resolution to examine forest change patterns during 2001 and 2017 both inside and outside of 13 typically national nature reserves in China. The average forest loss rates inside the reserves were significantly lower than those of outside the reserves (i.e., both in buffer and landscape zones), suggesting the success in protecting forest of these reserves in China. We found that the protection practice reduced approximately 10% of deforestation. Protection efficiency may be substantially overestimated (about 13-43%) if failing to control the related variables, such as altitude, climate, and human interference. The forest loss rates in the buffer zones were not significantly higher than those in the broader landscape zones, suggesting that leakage is not a frequent occurrence in the buffer zones of the reserves. However, the forest loss rates showed a slightly increasing tendency from 2001 to 2017, the loss rates increased gradually from inside to their outside buffer zones, and leakage was observed in certain zones of some years for most of the reserves. The conversions of forest to grassland and cultivated land were the primary trajectories of forest loss both inside and outside of the reserves. Though the leakage is not universal in the reserves across the country, forest loss rates are much higher in the buffer zones than those inside the reserves, resulting in increased insulation of the reserves that could undermine the provisioning of ecosystem services and the biodiversity conservation efficiency.

Identifying economically relevant forest types from global satellite data

Satellite data offer a transformative view of global forest resources but observe a highly restricted set of forest attributes, limiting both theoretical and applied research. In response to this problem, we propose the use of ‘economic forests’ as analytical units and show how key characteristics of forests that determine human decision-making can be retrieved by combining open earth observation datasets. To operationalize the economic forest approach, we draw on classical microeconomic theory (von Thünen's land use model) to identify two economically relevant variables (market accessibility and forest productivity) that can be proxied in open earth observation data. We then post-process the Global Forest Change dataset (v1.5) to produce wall-to-wall maps of forest harvest for the period 2000–2015, obtain proxy measures of our two variables for a stratified random sample, and classify the resulting dataset of global forest harvest into categories of economic forest using Gaussian Mixture Models. We map the resulting categories, demonstrating a scalable and low-cost approach to analyzing global forest resources which captures key features that determine economic use. We employ this approach to empirically test Hyde's tripartite typology of global forest resources, circumventing the data challenge posed by the theoretical importance of spatial value functions, and find evidence that an expanded typology is required. We also produce a first approximation of Earth's economically inaccessible forests, drawing on previous mapping of Intact Forest Landscapes. To demonstrate the value of our approach for knowledge-discovery, we allocate FAOSTAT-reported timber supply for 23 European countries to the categories of economic forests we identify and employ these data to analyze the Heckscher-Ohlin-Vanek relationship in forestry. This exercise reveals plausible and significant relationships between production structure and comparative advantage that are obscured by aggregated data, providing direct evidence of the value of our approach. We conclude by discussing potential applications of the economic forests concept.

Causal Analysis of Accuracy Obtained Using High-Resolution Global Forest Change Data to Identify Forest Loss in Small Forest Plots

Identifying areas of forest loss is a fundamental aspect of sustainable forest management. Global Forest Change (GFC) datasets developed by Hansen et al. (in Science 342:850–853, 2013) are publicly available, but the accuracy of these datasets for small forest plots has not been assessed. We used a forest-wide polygon-based approach to assess the accuracy of using GFC data to identify areas of forest loss in an area containing numerous small forest plots. We evaluated the accuracy of detection of individual forest-loss polygons in the GFC dataset in terms of a “recall ratio”, the ratio of the spatial overlap of a forest-loss polygon determined from the GFC dataset to the area of a corresponding reference forest-loss polygon, which we determined by visual interpretation of aerial photographs. We analyzed the structural relationships of recall ratio with area of forest loss, tree species, and slope of the forest terrain by using linear non-Gaussian acyclic modelling. We showed that only 11.1% of forest-loss polygons in the reference dataset were successfully identified in the GFC dataset. The inferred structure indicated that recall ratio had the strongest relationships with area of forest loss, forest tree species, and height of the forest canopy. Our results indicate the need for careful consideration of structural relationships when using GFC datasets to identify areas of forest loss in regions where there are small forest plots. Moreover, further studies are required to examine the structural relationships for accuracy of land-use classification in forested areas in various regions and with different forest characteristics.

A globally applicable indicator of the capacity of terrestrial ecosystems to retain biological diversity under climate change: The bioclimatic ecosystem resilience index

An important element of the Convention on Biological Diversity’s Aichi Target 15 – i.e. to enhance “ecosystem resilience … through conservation and restoration” – remains largely unaddressed by existing indicators. We here develop an indicator addressing just one of many possible dimensions of ecosystem resilience, by focusing on the capacity of ecosystems to retain biological diversity in the face of ongoing, and uncertain, climate change. The Bioclimatic Ecosystem Resilience Index (BERI) assesses the extent to which a given spatial configuration of natural habitat will promote or hinder climate-induced shifts in biological distributions. The approach uses existing global modelling of spatial turnover in species composition within three broad biological groups (plants, invertebrates and vertebrates) to scale projected changes in composition under a plausible range of climate scenarios. These projections serve as filters through which to analyse the configuration of habitat observed at a given point in time (e.g. for a particular year) – represented as a grid in which cells are scored in terms of habitat condition. BERI is then calculated, for each cell in this grid, as a function of the connectedness of that cell to areas of natural habitat in the surrounding landscape which are projected to support a similar composition of species under climate change to that currently associated with the focal cell. All analyses are performed at 30-arcsecond grid resolution (approximately 1 km cells at the equator). Results can then be aggregated to report on status and trends for any desired set of reporting units – e.g. ecoregions, countries, or ecosystem types. We present example outputs for the Moist Tropical Forest Biome, based on a habitat-condition time series derived from the Global Forest Change dataset. We also describe how BERI is now being extended to cover all biomes (forest and non-forest) across the entire terrestrial surface of the planet.

An Assessment of Global Forest Change Datasets for National Forest Monitoring and Reporting

Global Forest Change datasets have the potential to assist countries with national forest measuring, reporting and verification (MRV) requirements. This paper assesses the accuracy of the Global Forest Change data against nationally derived forest change data by comparing the forest loss estimates from the global data with the equivalent data from Guyana for the period 2001–2017. To perform a meaningful comparison between these two datasets, the initial year 2000 forest state needs first to be matched to the definition of forest land cover appropriate to a local national setting. In Guyana, the default definition of 30% tree cover overestimates forest area is by 483,000 ha (18.15%). However, by using a tree canopy cover (i.e., density of tree canopy coverage metric) threshold of 94%, a close match between the Guyana-MRV non-forest area and the Global Forest Change dataset is achieved with a difference of only 24,210 ha (0.91%) between the two maps. A complimentary analysis using a two-stage stratified random sampling design showed the 94% tree canopy cover threshold gave a close correspondence (R2 = 0.98) with the Guyana-MRV data, while the Global Forest Change default setting of 30% tree canopy cover threshold gave a poorer fit (R2 = 0.91). Having aligned the definitions of forest for the Global Forest Change and the Guyana-MRV products for the year 2000, we show that over the period 2001–2017 the Global Forest Change data yielded a 99.34% overall Correspondence with the reference data and a 94.35% Producer’s Accuracy. The Guyana-MRV data yielded a 99.36% overall Correspondence with the reference data and a 95.94% Producer’s Accuracy. A year-by-year analysis of change from 2001–2017 shows that in some years, the Global Forest Change dataset underestimates change, and in other years, such as 2016 and 2017, change is detected that is not forest loss or gain, hence the apparent overestimation. The conclusion is that, when suitably calibrated for percentage tree cover, the Global Forest Change datasets give a good first approximation of forest loss (and, probably, gains). However, in countries with large areas of forest cover and low levels of deforestation, these data should not be relied upon to provide a precise annual loss/gain or rate of change estimate for audit purposes without using independent high-quality reference data.

Deforestation and Household- and Individual-Level Double Burden of Malnutrition in Sub-Saharan Africa.

Although forests and forest-based ecosystems have been shown to influence health and sustainable diets, there is limited evidence on how deforestation affects the current nutrition transition and the double burden of malnutrition. We examined the relationship between deforestation and the individual- and household-level double burden of malnutrition in 15 countries in Sub-Saharan Africa. We combined data from geolocated Demographic and Health Surveys and the Global Forest Change dataset. We defined household-level double burden of malnutrition as the co-occurrence of an overweight woman of childbearing age (WCBA) and a stunted pre-school child (PSC) within the same household. We defined individual-level double burden in two ways: 1) as the co-occurrence of overweight and anemia within an individual WCBA, and 2) as the co-occurrence of overweight and stunting within a PSC. We used logistic regression analysis to examine the association between forest cover loss and these three measures after adjusting for potential confounders. We also assessed the mechanisms linking forest cover loss and nutritional status, such as livestock ownership and access to clean water. In our sample, the prevalence rates of the three measures of the double burden were: overweight and anemic WCBA: 8.4%, overweight WCBA and stunted PSC: 6.9%, overweight and stunted PSC: 2.7%. After adjusting for the confounders as well as country fixed effects and the month of the survey, forest cover loss was marginally associated with a higher odds of an overweight WCBA and stunted PSC (odds ratio (95% CI): 4.80 (0.82, 28.25)). We found no association between forest cover loss and odds of an overweight and stunted PSC (odds ratio (95% CI): 2.47 (0.80, 7.60)) or the odds of an anemic and overweight WCBA (odds ratio (95% CI): 0.71 (0.15, 3.32)). Deforestation does not seem to be an important driver of the double burden of malnutrition in SSA. However, deforestation influences several intermediate factors which, in turn, may influence the double burden. The overall weak association between forest cover loss and double burden measures mask significant heterogeneity across regions within SSA. Future research should unpack the mechanisms behind these regional differences.

Deforestation in Colombian protected areas increased during post-conflict periods

Protected areas (PAs) are a foundational and essential strategy for reducing biodiversity loss. However, many PAs around the world exist on paper only; thus, while logging and habitat conversion may be banned in these areas, illegal activities often continue to cause alarming habitat destruction. In such cases, the presence of armed conflict may ultimately prevent incursions to a greater extent than the absence of conflict. Although there are several reports of habitat destruction following cessation of conflict, there has never been a systematic and quantitative “before-and-after-conflict” analysis of a large sample of PAs and surrounding areas. Here we report the results of such a study in Colombia, using an open-access global forest change dataset. By analysing 39 PAs over three years before and after Colombia’s peace agreement with the Revolutionary Armed Forces of Colombia (FARC), we found a dramatic and highly significant increase in the deforestation rate for the majority of these areas and their buffer zones. We discuss the reasons behind such findings from the Colombian case, and debate some general conservation lessons applicable to other countries undergoing post-conflict transitions.

A spatially explicit database of wind disturbances in European forests over the period 2000–2018

Abstract. Strong winds may uproot and break trees and represent a major natural disturbance for European forests. Wind disturbances have intensified over the last decades globally and are expected to further rise in view of the effects of climate change. Despite the importance of such natural disturbances, there are currently no spatially explicit databases of wind-related impact at a pan-European scale. Here, we present a new database of wind disturbances in European forests (FORWIND). FORWIND is comprised of more than 80 000 spatially delineated areas in Europe that were disturbed by wind in the period 2000–2018 and describes them in a harmonized and consistent geographical vector format. The database includes all major windstorms that occurred over the observational period (e.g. Gudrun, Kyrill, Klaus, Xynthia and Vaia) and represents approximately 30 % of the reported damaging wind events in Europe. Correlation analyses between the areas in FORWIND and land cover changes retrieved from the Landsat-based Global Forest Change dataset and the MODIS Global Disturbance Index corroborate the robustness of FORWIND. Spearman rank coefficients range between 0.27 and 0.48 (p value < 0.05). When recorded forest areas are rescaled based on their damage degree, correlation increases to 0.54. Wind-damaged growing stock volumes reported in national inventories (FORESTORM dataset) are generally higher than analogous metrics provided by FORWIND in combination with satellite-based biomass and country-scale statistics of growing stock volume. The potential of FORWIND is explored for a range of challenging topics and scientific fields, including scaling relations of wind damage, forest vulnerability modelling, remote sensing monitoring of forest disturbance, representation of uprooting and breakage of trees in large-scale land surface models, and hydrogeological risks following wind damage. Overall, FORWIND represents an essential and open-access spatial source that can be used to improve the understanding, detection and prediction of wind disturbances and the consequent impacts on forest ecosystems and the land–atmosphere system. Data sharing is encouraged in order to continuously update and improve FORWIND. The dataset is available at https://doi.org/10.6084/m9.figshare.9555008 (Forzieri et al., 2019).

Assessing the Importance of Tree Cover Threshold for Forest Cover Mapping Derived from Global Forest Cover in Myanmar

Comprehensive forest cover mapping is essential for making policy and management decisions. However, creating a forest cover map from raw remote sensing data is a barrier for many users. Here, we investigated the effects of different tree cover thresholds on the accuracy of forest cover maps derived from the Global Forest Change Dataset (GFCD) across different ecological zones in a country-scale evaluation of Myanmar. To understand the effect of different thresholds on map accuracy, nine forest cover maps having thresholds ranging from 10% to 90% were created from the GFCD. The accuracy of the forest cover maps within each ecological zone and at the national scale was assessed. The overall accuracies of ecological zones other than tropical rainforest were highest when the threshold for tree cover was less than 50%. The appropriate threshold for tropical rainforests was 80%. Therefore, different optimal tree cover thresholds were required to achieve the highest overall accuracy depending on ecological zones. However, in the unique case of Myanmar, we were able to determine the threshold across the whole country. We concluded that the threshold for tree cover for creating a forest cover map should be determined according to the areal ratio of ecological zones determined from large-scale monitoring. Our results are applicable to tropical regions having similar ecological zones.