CLASlite Projects

Forest degradation is the reduction of the capacity of a forest to provide goods and services. In the Brazilian Amazon region, degraded forests are dominant in the pioneer front landscapes where the economic growth and agricultural expansion has converted the primary forest to a mosaic of pastures, crop lands and forests in various stages of degradation. Understanding spatial structure and temporal trajectories of forest degradation is important to support forest conservation and management, aiming at reducing greenhouse gas emissions, and retrieving biomass and biological diversity. We present a method to evaluate and map different levels of forest degradation in the eastern Amazon in Paragominas district. We used Landsat time series (2000-2015) and the CLASlite software. We calculated the bands of Photosynthetic Vegetation (PV), Non-Photosynthetic Vegetation (NPV) and bare Soil (S), for each year to produce annual maps of forest degradation and we have analyzed the variance of each CLASlite band (PV, NPV, S) along the 2000-2015 period. Field observations of degraded forests, based on indicators of forest structure were used for the validation. The results showed that the CLASlite bands variances are a good indicator of the forest degradation process.

The results showed that the CLASlite bands variances are a good indicator of the forest degradation process.

Figure 1. The location of study area in Brazil.

  • Université du Maine, Avenue Olivier Messiaen, France
  • Université du Maine, Avenue Olivier Messiaen, France
  • UR Forests and Societies, 34398 Montpellier, France
  • Université du Maine, Avenue Olivier Messiaen, France

In the Brazilian Amazon, multiple logging activities are undergoing, involving different actors and interests. They shape a disturbance gradient bound to intensity and frequency of logging, and forest management techniques. However, until now, few studies were carried out at the landscape scale taking into account these multiple types of logging and this disturbance gradient. Here we address this issue of how to account for the multiple logging activities shaping the current forest landscape. We developed an inexpensive and efficient remote sensing methodology based on Landsat imagery to detect and track logging activity based on the monitoring of canopy opening. Then, we implemented a set of remote sensing indicators to follow the different trajectories of forest disturbance through time. Using these indicators, we emphasized five major spatial and temporal disturbance patterns occurring in the municipality of Paragominas (State of Pará, Brazilian Amazon), from well-managed forests to highly over-logged forests. Our disturbance indicators put in evidence a significant difference between legal and illegal patterns, with some illegal areas having suffered more than three explorations in fifteen years. They also clearly underlined the efficiency of RIL techniques applied under FSC guidelines to reduce the logging impacts in terms of canopy opening. For these reasons, we argue the need to promote legal certified logging to conserve forests, as without them, many actors mine the forest resources without any concerns for future stocks. Finally, our remote tracking methodology, which produces easy to interpret disturbance indicators, could be a real boon to forest managers including for conservationists working in protected areas and stakeholders dealing with international trade rules such as RBUE or FLEGT.

  • CIRAD, UPR Forests and Societies (F&S), Campus International de Baillarguet, France
  • National Institute for Space Research (INPE), Amazon Regional Center (CRA), Science and Technology Park of Guamá, Brazil.
  • EMBRAPA Amazônia Oriental, Brasil

Figure 1. Localization of the studied plots in the municipality of Paragominas.

In Belize, the lack of deforestation, forest degradation, socioeconomic and erosion data results in the inability of management organizations to make timely assessments and decisions for sustainable resource management in southern Belize. This study uses CLASlite algorithms, statistical analysis, social surveys and the Revised Universal Soil Loss Equation to identify erosion hotspots, drivers, measure, analyze and map deforestation, and forest degradation that occurred in southern Belize. In Toledo, land and institutional policy, distance to markets and lack of alternative livelihoods are the main drivers of deforestation and forest degradation. The results of the deforestation and forest degradation analysis indicate that in 2009-2011 and 2011-2012 the annual rates of deforestation were 0.75% (2,480 ha) and 1.17% (3,834 ha) respectively and the annual rates of forest degradation in 2009-2011 and 2011-2012 were 0.09% (307 ha) and 0.33% (1,110 ha) respectively. Moreover, along the Belize-Guatemala border, forest declined from 96.9% to 85.7% in Belize and from 83.15% to 31.52% in Guatemala. The Mann-Whitney U test identified significant differences between leaders and stakeholders regarding the ranking of challenges faced by management organizations in the Belize-Guatemala border, except for the lack of assessment and quantification of deforestation (LD, SH: 18.67, 23.25, U = 148, p > .05). Finally, in Toledo's Rio Grande watershed erosion hotspots were located in the upper-mid reaches of the watershed near the communties of Crique Jute, Naluum Ca, San Pedro Columbia and San Miguel. The Mann-Whitney U test identified significant difference in the ranking of erosion drivers (cattle ranching, logging, and clearing of slopes) between communities.

This research provides significant information on the drivers, deforestation, forest degradation and erosion that will aid stakeholders to garner community support, develop and implement sustainable management practices in southern Belize.

Figure 1. Surveyed communities and protected areas.

  • Nagasaki University, Japan

Climate change projections have predicted more frequent and severe droughts that may lead to major loss of trees and subsequent range shifts. Drought-induced tree mortality leaves both dead & live trees intermixed that remain standing rather than leaving clearings that result from acute disturbances such as fire. Thus this disturbance is difficult to detect at a regional scale, but is a harbinger of range shifts so its detection is high priority. During the summer of 2011, the southwestern US including Texas was impacted by an extreme drought. Statewide tree mortality was observed and thus provided an opportunity to test the efficacy of moderate to coarse resolution remotely- sensed indicators to detect, map and enumerate drought-induced tree mortality. Calibration models of 250-m ΔNDVI and 1-km VegDRI with 599 field data plots of tree mortality were developed to produce predictive maps. ΔNDVI, ΔPV, and NPV mortality indices were derived from 30-m Landsat 7 and compared to each other and 250-m ΔNDVI. ΔNDVI predicted tree mortality best (Khat = 0.15), with an estimate of 9% mortality that was primarily concentrated in East and Central Texas. However at 30-m resolution for East Texas, ΔPV matched the validation data best (Khat = 0.21).

Maximum entropy models were used with the field data to test the relative importance of 2011 drought conditions versus historical climate drivers of the distribution of drought-induced tree mortality. 2011 drought conditions explained 57% of the resulting model (AUC = 0.84) and bioclimate variables explained 43%. Mean annual precipitation explained 17% of tree mortality, followed by 2011 isothermality (16%). Models were run to test the contribution of edaphic, biotic, and climatic factors toward explaining dead tree distribution, and also test of effects of scale and location (East vs. Central Texas).

Climate was the highest contributor at the state scale (42%) and also in Central Texas (48%). In East Texas, edaphic factors were the major driver (47%). As drought frequency and intensity increase as predicted, a refinement of detection techniques and understanding of the drivers of tree mortality are needed to understand and predict the nature of drought consequences for forests.

Figure 1. FIA Regions and distribution of field plot locations.

  • Texas A&M University, United States

Selection of robust, accurate and cost efficient forest biomass monitoring and carbon mapping methods are essential to carry out Reducing Emission from Deforestation and Degradation (REDD+) program of the United Nations Framework Convention on Climate Change (UNFCCC). There are numerous tools applied for Forest Resource Assessment (FRA) ranging from traditional intensive field measurement to sophisticated airborne Light Detection and Ranging (LiDAR) technology. The key driving force behind the advancement of FRA methods in course of time is to obtain accurate forest information at low cost.

  • Tribhuvan University, Nepal

Figure 2. Operational principle of airborne LiDAR (Source: Arbonaut, 2011).

Tropical forests provide important ecosystem functions (e.g. primary production, carbon and nitrogen cycling, the hydrogen cycle, etc.) as well as ecosystem services (e.g. pharmaceuticals, timber, air and water purification, climate stabilization, etc.) (Tallis et al. 2013; Virginia and Wall 2013). Despite their important role in providing such functions and services, tropical forests and the species within them are increasingly being lost, due in part to increasing deforestation (Hansen et al. 2013). One recent study, for instance, places 36%-57% of Amazon tree species at risk of extinction (ter Steege et al. 2015). Where forests are communities of trees, i.e. inter-species associations (Chave 2009; Duckworth et al. 2000; Helmer et al. 2012)) and not merely random collections of individuals, there is therefore a need to understand the patterns of the diversity of tree communities in the tropics. Understanding the patterns of tree diversity is ultimately also necessary for adequate management and conservation of tropical forests, and such information can also aid in evaluating how forests will respond to future environmental perturbations.

Figure 1. Left: Mean MODIS (MCD43A4)-derived reflectance imagery of French Guiana for the month of September; right: location of French Guiana in South America.

  • Institut de recherche pour le développement, France
  • Technische Universität Dresden, Germany

Toledo, the southernmost district, is the hub of Belize's Mayan population, descendants of the ancient Mayan civilization. The Toledo District is primarily inhibited by Kekchi and Mopan Mayans whose subsistence needs are met by the Milpa slash-and-burn agricultural system and the extraction of forest resources. The poverty assessment in the country indicates that Toledo is the district with the highest percentage of household an individual indigence of 37.5 % and 49.7 % respectively. Forest cover change in the area can be attributed to rapid population growth among the Maya, together with increase in immigration from neighboring countries, logging, oil exploration and improvement and construction of roads. The forest cover change analysis show that from 2001 to 2011 there was a decrease of Lowland broad-leaved wet forest of 7.53 km sq, Shrubland of 4.66 km sq, and Wetland of 0.08 km sq. Forest cover change has resulted in soil erosion which is causing the deterioration of soils. The land cover types that are contributing the most to total erosion in the Rio Grande watershed are no-forest, lowland broad-leaved wet forest and submontane broad-leaved wet forest. In this study the Revised Universal Soil Loss Equation (RUSLE) was employed in a GIS platform to quantify and assess forest cover change and soil erosion. Soil erosion vulnerability maps in Toledo's Rio Grande watershed were also created. This study provides scientifically sound information in order to understand and respond effectively to the impacts of soil erosion in the study site.

  • Department of Civil Environmental Engineering, Nagasaki University, Japan

Figure 1. Study Site.

Land use change is a significant factor in environmental conservation and climate change which may be positive or negative depending on how it occurred. It aimed at examining the land use changes that took place across the catchment from 1984 to date. Landsat images of the area were downloaded from USGS database and processed using CLASlite, a forest monitoring application developed by Carnegie Institute for Science (CIS) and ENVI software for land use classification and change detection. Over the study period, the area was observed to have experienced different modifications, most notable one being deforestation in the upper part of the catchment where Mount Kenya Forest extended. Most of the deforestation was understood to have taken place in the 1980s and 1990s with 2.5% forest depletion between 1984 and 1995. That may be attributed to illegal logging because up to 60% of the lost forest area during the period was taken by uncultivated land.

  • Department of Civil Engineering and Construction Management, Pan African University, Kenya
  • Biomechanical and Environmental Engineering Department, Jomo Kenyatta University of Agriculture and Technology, Kenya
  • Department of Civil Engineering and Construction Management, Jaramogi Oginga Odinga University, Kenya

Figure 1. The study area with two key land use classes digitized from Survey of Kenya topographical maps (1973).

Optical remote sensing is used in different manners in the context of tropical forests monitoring: Land-cover characterization (forest types) Measuring deforestation (change detection) Estimation of degradation (under development). Objectives of this presentation are: (1) show the potential of optical remote sensing for land cover characterization in space and time; (2) check the deforestation information already available; (3) make a review of what is done in terms of degradation; (4) present perspectives in the scaling-up workflow we develop.

  • Cirad, France
  • Office National Des Forets, France
  • Universite Rennes, France

Figure. Map showing in green the undisturbed forest area, in red the deforested area from 2000 to 2010, in dark grey the deforested area prior to 2000 and in light grey non-forested areas.

Identifying Malaria Transmission Risk in the Peruvian Amazon: A Geospatial Approach

Peru has endured a long history with malaria, an infectious disease caused by the mosquito-borne transmission of the Plasmodium parasite. Throughout the 20th century, disease prevalence has varied tremendously with a number of factors including Peru's growth and development, variable support for malaria control measures, and the migration of immunologically naïve populations. However, many researchers believe that anthropogenic deforestation is at the root of a recent resurgence of malaria in the Peruvian Amazon. Deforestation creates favorable conditions for disease transmission by increasing mosquito habitat and placing humans in close proximity to more abundant disease vectors. In addition, rural communities often lack the resources to combat malaria due to the prohibitive cost of conventional technologies and lack of access to health care. Using data derived from field collections and remotely sensed images in the Loreto department of Peru, this study proposes a new method for characterizing malaria risk in the Peruvian Amazon. A variety of novel geospatial and remote sensing techniques were used to develop environmental layers from satellite imagery and produce the species distribution model. A geospatial risk model synthesized the predicted mosquito habitat and associated community risk factors into an assessment of malaria exposure risk. The threat model developed from this study can be used to create maps that will help local communities manage their malaria risk.

  • Duke University, United States

Management efforts, such as the reduction of available mosquito breeding habitat, can be concentrated in areas identified as high-risk for malaria exposure.

Figure 3-1. Study Area. The extent and boundaries of the study area are shown in red. The study area is located in the Loreto Department of Peru. Yellow dots indicate transect locations for mosquito collections.

CLASlite Course support: (claslite_coursesupport@carnegiescience.edu)

CLASlite Technical support: (claslite_techsupport@carnegiescience.edu)


Carnegie Institution for Science

260 Panama Street • Stanford, CA 94305, USA