The existing Liberia Forests Initiative (LFI) is a flagship World Bank effort aimed to support transparency, sustainability and good governance in the management of Liberia's forests. The LFI was developed in explicit acknowledgement of the potential of Liberia's forests to contribute to economic growth through timber production, but only if greater transparency and accountability could be achieved.
The World Bank has been providing technical assistance to Liberia's forest sector since 2004. During that period it contributed to the Forest Concession Review, forest inventory, and supported the reform of a national Forest Development Agency. It has also assisted the Government in developing the Liberia Extractive Industries Transparency Initiative (L-EITI) as well as Liberian national REDD+ (Reducing Emissions from Deforestation and Forest Degradation) strategy (through Forest Carbon Partnership Trust Fund).
The adequate geo-data regarding Liberian forest resources is necessary to achieve these and other goals related to sustainable forest management. The knowledge of Liberia forests was drastically reduced as a result of the uncontrolled logging activities, especially during the civil wars between 1989-1996 and 1999-2003. To bridge that gap the LFI has initiated mapping and survey efforts to analyze the forest activities from 2002 to 2004, and 2005-2006. In 2010 the Bank has partnered with the European Space Agency to advance natural resources analysis and has jointly developed a pilot demonstration project which focused on mapping of forest and land use cover in two areas: large forest area in North-Western Liberia (Gbarpolu & Grand Cape Mount). The project now sets the basis for creating a country-wide forest and land cover map and is aimed to contribute to the future national forest inventory as well REDD activities.
Collaboration with the European Space Agency
The Liberia forest resource mapping was conducted between April and October 2011 as a proof-of-concept study to assess the added value of using the latest satellite Earth Observation techniques to support sustainable management of commercial, conservation and community forestry.
Forest cover mapping
The assessment of the Liberia’s Gbarpolu and Grand Cape Mount forest areas was completed for the years 2002, 2007 and 2010 covering approximately 75,000 ha of forest and identifying the rate of deforestation and net re-growth between these years. The study has found that with a monitoring system based on a combination of field data collection and EO data analysis, the control of forest resource utilization can be much improved.
Within the Liberia Forestry Initiative the World Bank is:
- providing assistance to the Liberia Extractive Industries Transparency Initiative (L-EITI), the first EITI effort to include timber;
- supporting the efforts to reduce emissions from deforestation and forest degradation (REDD +) through an FCPF grant to the Government of Liberia;
- helping to establish and expand the Protected Areas Network and Community Forestry, through GEF grants to the Government; and
- continuing to work with the Government and donor communities to enhance transparency, accountability, and rule of law.
More information can be found at Liberia Country Brief website.
The analysis was based on the data derived from various sensors: Landsat TM (30m), SPOT5 (2,5m - 20m) and RapidEye (up to 5m). With the use of these latest satellite datasets and enhanced processing techniques the mapping accuracy was significantly improved: the minimal mapping unit (MMU) came down to less than 1 ha as compared to the previous best available assessment of Liberian forest dated to 2006 and conducted at the scale of 1000 ha (Figure 1).
The forest cover classification included the following classes.
Forest in flat terrain (slope derived from DEM < threshold according to forestry regulations)
• Closed dense forest (>80%)
• Medium dense forest (30-80%)
• Open forest
Forest in steep terrain (slope derived from DEM)
• Closed dense forest (>80%)
• Medium dense forest (30-80%)
• Open forest
Burnt areas (not found in the study areas)
Figure1: Land cover mapping from 2004 with a MMU of 1000 ha (left image), compared to the maps developed for the years 2010/2011 within ESA-WB collaboration which achieved MMU of 1 ha (middle image). The image on the right is a high-resolution Rapid Eye imagery used as reference to produce more accurate maps. Legend: Light green/yellow are agricultural land, green are open forest and dark green are dense forest, red are urban and rural settlements. Credit: Eoworld project, Metria/Geoville for European Space Agency/World Bank.
The results proved that rate of change in forest cover in Gbarpolu and Grand Cape Mount is relatively small with a net regrowth of approximately 0.20 %. The annual deforestation rate between 2002 and 2007 was respectively 0.15 % and 0.3 %. Deforestation, if at all, was mainly taking place close to populated areas and likely as a result of traditional agricultural practices, which did not encroach into the surrounding dense forest. Moreover, there was no evidence of timber harvesting taking place in the investigated time period, despite the presence of visible timber roads in the imagery from 2002. Such little-exploited forest areas have presumably high biodiversity and conservation value despite the fact that the presence of timber roads may indicate that some selective cuttings could have been taking place.
Satellite-based monitoring is an effective method an independent monitoring of land use changes. It makes it possible to identify the hot spot areas where violations against regulations occur, especially in remote areas. For example, the study analyzed the state of the forest resources in the areas where logging should be restricted - along the rivers' buffer zone of 200 meters. It was also estimated that almost 70% of the land within 200m from rivers is forest with a density of 30 % or higher.
Land use mapping
In addition to the assessment of forest cover, the detailed land use and land cover maps were provided for the year 2010 with 11 thematic classes revealing forest areas as well as primary roads, grassland, shrubs, urban areas, isolated rural villages, base soil, sand rocks and rivers (of minimum width of 10 meters) (Figure2). The study found that 90% of the area is covered by forest (of different densities), 5% by shrub land, and almost 5% indicates human activities such as housing, infrastructure, or farming.
Figure2: An excerpt from land cover mapping from 2010 with a MMU of 1 ha.The forest classification consists of the closed dense forest (>80%), medium dense forest (30%-80%) and open forest. Credit: Eoworld project, Metria/Geoville for European Space Agency/World Bank.
Digital Elevation Model
Forest map was enhanced with terrain mapping using Digital Elevation Model (ASTERGDEM exploited together with the SRTM DEM at 15m vertical resolution) to assess available forest area. Such elevation model gives information about topographic characteristics other then land use, such as the course of river valleys, mountain ridges, and depressions. This information is essential for an effective and efficient planning of infrastructure elements as roads or power transmission lines.