The Forests 2020 project has supported partner countries to develop improved data products to assess the risk of deforestation in different areas. Ecometrica analyst, Dr Paula Nieto, explains why understanding risk is important, and some of the technical outputs of the project.
Figure 1 – Deforestation Risk Map for the year 2019 covering the national forest extent of Indonesia. Produced by DAEMETER (March 2020) Consulting, funded by Forests 2020.
The causes of deforestation are multiple and varied, and addressing them involves understanding the complex processes that affect them, and their relationships. This includes political, institutional, environmental, economic, and cultural factors, some of which are challenging to predict.
Deforestation risk mapping is based on identifying and mapping drivers of deforestation within an impact area by modelling deforestation factors. Risk maps do not attempt to predict the absolute quantity or location of deforestation, but instead map the location where forest loss is most or least likely to occur. They recognise that the risk of loss is not spatially uniform, it doesn’t occur homogeneously across the landscape, and does this by indicating different risk classes indicative of the amount of forest expected to be lost. These maps should be adapted locally whenever possible to ensure that all region specific drivers are accounted for, preferably using appropriate expert knowledge and local datasets.
Although estimating the drivers of deforestation is very complex (due to local and temporal circumstances) deforestation risk maps provide a useful, simple and flexible tool to identify areas at high risk of deforestation.
They can be can be used to:
- Inform project design – ensuring that interventions are targeted at high risk areas and at the correct scale
- Monitor and evaluate projects – assessing the expected and actual impact of forest protection interventions to identify successes and failures using methods such as the Hectares Indicator
- Comply or operate deforestation free supply chains – highlighting locations of high risk within supply chain areas or farms to target resources
- Forecast expected deforestation – across given regions which have remained at high risk levels for extended periods of time
- Estimate future carbon emissions – linking to carbon density values
Through the Forests 2020 project deforestation risk maps have been developed by local mapping experts for 7 tropical forest countries – Ghana, Kenya, Brazil (Figure 2), Mexico, Colombia, Belize and Indonesia (Figure 1). Forests 2020 is a major investment by the UK Space Agency, led by Ecometrica, to help protect and restore up to 300 million hectares of tropical forests by improving forest monitoring in seven partner countries (Ghana, Kenya, Brazil, Mexico, Colombia, Belize and Indonesia) through advanced uses of satellite data. One of the objectives of the project was to increase the availability and accuracy of deforestation risk maps in tropical forest nations so that they can be used to assess the impact of forest protection interventions funded by the International Climate Fund (ICF) and to help government departments better target resources to reduce deforestation.
Ecometrica has reviewed these maps and developed a document describing the methods, drivers, and main conclusions behind partner countries’ risk maps. A key finding is that accessibility was proven as one of the main factors driving deforestation in most countries, with more accessible areas (e.g. areas close to roads) being at higher risk of forest loss. Some of the deforestation risk maps have been validated by Forest2020 partners, and in general a good agreement was found between the risk map and the actual forest loss, with deforestation occurring more in the very high and high-risk areas.
Figure 2: Map covers the Cerrado biome for 2018, known as the region of the agriculture-forest frontier of Brazil, produced by the National Institute of Space Research, INPE.
Ecometrica has also created a simple accuracy assessment method that can be easily applied to any deforestation risk map and for any specific area. This accuracy assessment demonstrates if a map possesses a satisfactory range of accuracy for its application. This method can be used globally, is easily applicable and repeatable, and can be used to assess on how well the probabilities of deforestation associated to each risk level apply and inform future improvements and updates of the map.