Measuring the Impact of Monitoring: How We Know Transparent Near-Real-Time Data Can Help Save the Forests

Global Forest Watch (GFW) is an online platform that distills satellite imagery into near-real-time forest change information that anyone can access and act on. Like other open-data platforms, GFW is based on the idea that transparent, publicly available data can support the greater good—in this case, reducing deforestation. By its very nature, the use of freely available data can be difficult to track and its impact difficult to measure. This chapter explores four approaches for measuring the reach and impact of GFW, including quantitative and qualitative approaches for monitoring outcomes and measuring impact. The recommendations can be applied to other transparency initiatives, especially those providing remote-sensing data.

Strategic Application of Topoclimatic Niche Models in Managing Forest Change

Forest management traditionally has been based on the expectation of a steady climate. In the face of a changing climate, management requires projections of changes in the distribution of the climatic niche of the major species and strategies for applying the projections. We prepared climatic habitat models incorporating heatload as a topographic predictor for the 14 upland tree species of southwestern Colorado, USA, an area that has already seen substantial climate impacts. Models were trained with over 800,000 points of known presence and absence. Using 11 climate scenarios for the decade around 2060, we classified and mapped change for each species. Projected impacts are extensive. Except for the low-elevation woodland species, persistent habitat is rare. Most habitat is lost or threatened and is poorly compensated by emergent habitat. Three species may be locally extirpated. Nevertheless, strategies are described that can use the projections to apply management where it is likely to be most effective, to facilitate or assist migration, to favor species likely to be suited in the future, and to identify potential climate refugia.

Deforestation Detection with Fully Convolutional Networks in the Amazon Forest from Landsat-8 and Sentinel-2 Images

The availability of remote-sensing multisource data from optical-based satellite sensors has created new opportunities and challenges for forest monitoring in the Amazon Biome. In particular, change-detection analysis has emerged in recent decades to monitor forest-change dynamics, supporting some Brazilian governmental initiatives such as PRODES and DETER projects for biodiversity preservation in threatened areas. In recent years fully convolutional network architectures have witnessed numerous proposals adapted for the change-detection task. This paper comprehensively explores state-of-the-art fully convolutional networks such as U-Net, ResU-Net, SegNet, FC-DenseNet, and two DeepLabv3+ variants on monitoring deforestation in the Brazilian Amazon. The networks’ performance is evaluated experimentally in terms of Precision, Recall, F1-score, and computational load using satellite images with different spatial and spectral resolution: Landsat-8 and Sentinel-2. We also include the results of an unprecedented auditing process performed by senior specialists to visually evaluate each deforestation polygon derived from the network with the highest accuracy results for both satellites. This assessment allowed estimation of the accuracy of these networks simulating a process “in nature” and faithful to the PRODES methodology. We conclude that the high resolution of Sentinel-2 images improves the segmentation of deforestation polygons both quantitatively (in terms of F1-score) and qualitatively. Moreover, the study also points to the potential of the operational use of Deep Learning (DL) mapping as products to be consumed in PRODES.

Agriculture is the primary driver of tree cover loss across the Forestière region of the Republic of Guinea, Africa

Biodiversity hotspots are conservation priority areas that feature exceptionally high levels of species endemism and high levels of habitat loss. The Guinean Forests of West Africa hotspot, home to a quarter of all the mammal species of Africa, has experienced high levels of forest loss within its protected areas. Here, we analyzed tree cover loss and its proximate drivers within Guinée Forestière, a high biodiversity region within the Guinean Forests of West Africa hotspot, both inside and outside protected areas. Using Landsat analysis ready data and a regionally calibrated, annual forest change detection model, we mapped tree cover loss occurring across this region from 2000 to 2018. We quantified the area of tree cover loss and identified proximate drivers using a statistical sample of reference data. The total tree cover loss in Guinée Forestière between years 2000 and 2018 was 10,907 km2 (SE 889 km2), which consists of approximately 25% of the region’s total land area. Of this total loss, 364 km2 (SE 91 km2) occurred within protected areas of high biodiversity value. Tree cover loss was not consistent across high biodiversity areas and did not appear to be related to protected area classification. Smallholder agriculture (subsistence and cash crop farming) was the primary driver of tree cover loss across Guinée Forestière. This research provides multitemporal spatial data on tree cover dynamics that is required for effective implementation of sustainable management and biodiversity conservation strategies within the broader socioecological landscape of Guinée Forestière. We also highlight important limitations to consider and address when using remote sensing to automate change detection across landscapes.

Global Commodity Markets, Chinese Demand for Maize, and Deforestation in Northern Myanmar

This paper makes a significant contribution to understanding the logic of deforestation in Northern Myanmar and connects global trends and regional political economy with local environmental changes. Methodologically, through a combination of remote sensing GIS analysis, for which we use a newly available Myanmar Forest Change dataset produced by TerraPulse and the Smithsonian Conservation Biology Institute, as well as on-the-ground field research observations and interviews with farmers, this paper examines how the expansion of maize plantations in the northern part of Myanmar has implications for deforestation in the region. It argues that a combination of global commodity price shock around 2011–2012 plus easy market access to China generated strong incentives for local farmers to increase the cultivation of maize. The paper contributes to how we understand the environmental impacts of Chinese demands for agricultural products in Southeast Asia.