ReCaREDD

Regional Forest Obervatories

Roadless

TREES-3




Roadless Forest Project

Objectives

Aerial view of a tropical humid forest.

Tropical forests protect us from climate change and are a haven for biodiversity. Most forest loss, degradation and fragmentation occurs in a “risk zone” around transport networks. Although communities need roads to access markets, hospitals and schools, roads also open up the forest to possible damage. In tropical areas, new road building is often followed by secondary roads being created, often unplanned or illegal, which then trigger further degradation and deforestation. The remoteness of some areas makes it especially hard to monitor and manage this change. The RoadlessForest pilot project aims to provide up-to-date data, to help plan infrastructure wisely and spot possible areas where the forest is in danger. The Joint Research Centre of the European Commission, with funding from the European Parliament, is building a road and forest atlas for tropical regions to support sustainable development. Sustainable development and sustainable forest management is a way of protecting biodiversity, fighting desertification and responding to climate change, whilst ensuring that forest ecosystems deliver goods and services.

Road in a degraded tropical forest landscape.

Tools and methods

By processing data from satellite (Landsat) images at fine spatial resolution, we aim at mapping pan-tropical evergreen forest cover dynamics (deforestation, degradation, regrowth) at 30-meter resolution from 1982 to 2017. With RoadlessForest, we are also piloting crowdsourced mapping to monitor roadbuilding in remote forest areas. By modelling the deforestation process with advanced statistical approaches (such as hierarchical Bayesian models), we also aim at understanding better the role of road development in driving forest cover changes.

Undisturbed evergreen forest cover and changes over the last 36 years

A methodology has been developed and implemented for processing 36 years of Landsat data (1 140 000 scenes) and mapping the tropical evergreen forest cover dynamics at 30m spatial resolution. This unprecedented work gives globally consistent and locally relevant information on the forest state. For the first time at this scale, we discriminate deforestation from degradation (where disturbances are visible over a short period) and analyze the temporal sequences of disturbances over the 36 years. Our dataset depicts the evergreen forest extent and patterns of disturbances through two complementary layers: (i) a transition map, and (ii) an annual change dataset. The transition map captures the resulting disturbance dynamics over the 36-year period, whereas the annual change dataset is a collection of 36 maps depicting - for each year between 1982 and 2017 - the spatial extents of undisturbed forests and disturbances.

Map of undisturbed evergreen forest and changes (transition map)
Map of undisturbed evergreen forest, logging roads and degraded forest (due to logging and fire) in North Congo from Landsat satellite images analysis on 36 years.

Available products

Accessibility map

The Global Map of Accessibility characterizes the connectedness of the human landscape by illustrating the amount of time it takes to access the nearest densely populated area. Mapping accessibility to cities is a useful proxy for the relative ease by which people in rural areas can access services and resources concentrated in more urban areas. Conversely, this map measures the relative inaccessibility and remoteness of the forest areas. The map is available at http://forobs.jrc.ec.europa.eu/products/gam/

Accessibility map.

Intact humid forest and forest cover change over the last 32 years

A methodology is currently being developed and implemented with Google Earth Engine for mapping the intact forest and the forest cover change in the humid tropical world for the last 32 years at 30m spatial resolution. This would be an unprecedent work giving globally consistent and locally relevant information on the forest state. The final version of the map will be available in the near future.

Map of intact humid forest and logging roads in North Congo
from Landsat satellite images analysis on 32 years.

Deforestation probability map

We develop a methodology to derive a map of the probability of deforestation as a function of several spatial factors including roads. We thus isolate the effect of roadbuilding in driving deforestation from the effect of other spatial factors such as population density or topography. With the deforestation probability map, we can also identify both forest areas with high risk of deforestation and potential intact forest areas in the future. A first deforestation probability map has been produced for the whole Africa.

References

Uchida H. and Nelson A., 2008
Agglomeration Index: Towards a new measure of urban concentration
Urbanization and development: Multidisciplinary perspectives, 41-60. World Development Report 2009. The World Bank.
Vancutsem Ch. and F. Achard, 2016
Mapping Intact and Degraded Humid Forests over the Tropical Belt From 32 Years Of Landsat Time Series
Paper 2034 - Session title: Tropical Forest and REDD+ 1. ESA living planet symposium 2016
Vieilledent G., C. Grinand, M. Pedrono, T. Rabetrano, J.-R. Rakotoarijaona, B. Rakotoarivelo, F. A. Rakotomalala, Linjanantenaina Rakotomalala, Andriamandimbisoa Razafimpahanana and Frédéric Achard.
Global trade and bad governance are responsible for the unceasing deforestation in western Madagascar, not poverty.

Project website

A dedicated project website is expected to be relaunched soon