Satellite Technology Used by UK Space Agency’s Forests 2020 Project to Monitor Tropical African Forests and Cocoa Industry

Supply chains in Africa have caused deforestation by illegal cocoa farming, damaging protected rainforests and creating damage to the very viable cocoa industry. Using satellite derived information from the UK Space Agency’s Forests 2020 Project, led by Ecometrica, the Ghana Forestry Commission has been supported in the development of a landscape-level map that separates cocoa from forestry, which is critical to measure how cocoa is driving deforestation. According to company materials, African Governments and the world’s cocoa companies look to UK Space Agency’s International Partnership Programme’s Forests 2020 to provide reliable and accurate maps that map forest cover change and differentiate cocoa farms from natural forests.

Fermented cocoa beans drying in Ghanaian sunshine (pic: Lewis Rattray)

With over two million small-scale farmers growing cocoa in the Ivory Coast and Ghana – the world’s biggest producers – it has proven difficult to track where each bean is coming from and therefore exert pressure on suppliers to end unsustainable practices.

’Big cocoa’ has pledged to stamp out unsustainable farming methods that involve the destruction of protected rainforests in West Africa via the Cocoa & Forests Initiative (CFI), but it has proved a difficult task for an industry served by millions of small farmers in the region. In a further development, Ecometrica, the downstream space information company, which leads the Forests 2020 project to monitor tropical forests using satellite technology, is bringing together its innovative platform with new, more detailed land cover map to enable cocoa companies to securely plot their supply chain and assess their impact on protected areas.

Ecometrica's Paula McGregor, a member of the Space Programme team, spoke with GISCafe Voice about the program to monitor tropical forests using satellite imagery.

Cocoa Segregation – Land Use Map V2

What are sustainable sources of cocoa, as mentioned in the press release?

Sustainably sourced cocoa in Ghana is produced by smallholder farmers who earn a fair income, engage in responsible labour practices and employ environmentally friendly farming practices (such as cocoa agroforestry) that safeguard the environment, conserve biodiversity and that do not cause deforestation or forest degradation.  Many cocoa companies and traders recognise the need for sustainable, zero deforestation supply chains and have signed up to initiatives like the Cocoa and Forests Initiative (CFI), but have so far struggled to monitor the effectiveness of their commitments. The Ecometrica led, project, Forests 2020, funded by the UK Space Agency’s International Partnership Programme is supporting the Government of Ghana’s Forestry Commission to create reliable and accurate maps that map forest cover change and differentiate cocoa farms from natural forests.  These maps are made available to the cocoa companies and traders to help them monitor their supply chains.

What tools are used to successfully gather data on cocoa farms, if cocoa is grown under the forest canopy and can’t easily be seen with satellite monitoring?

Our Forest 2020 Ghana partners, Ghana Forestry Commission’s Resource Management Support Centre (RMSC), Kwame Nkrumah University of Science and Technology (KNUST) with our UK partners at the University of Leicester, have worked closely to develop two processing chains; a fusion of Sentinel 1 and 2 and a fusion of ALOS PALSAR and Planet, that segregate monocropped and unshaded cocoa from natural forest areas. These tools are successfully gathering data on the spread of unshaded, monocropped cocoa production in West Ghana.  A draft cocoa map has been produced which is undergoing expert review. The crucial point is that unshaded, monocropped cocoa can easily be detected using satellites because it is not hidden under the canopy of forest trees, as is the case with shaded cocoa agroforests.

The remaining challenge is to be able to differentiate shaded cocoa agroforests from natural forests.  This is important because a substantial amount of cocoa, particularly in Ghana and Côte d'Ivoire, is being produced illegally inside protected areas (including forest reserves and national parks), which should be conserved as natural forests and it is difficult to detect using satellites because the upper forest canopy masks the cocoa trees.

The Forests 2020 project is testing the use of Light Detection and Ranging (LIDAR) and Structure from Motion (a photogrammetric technique for estimating three dimensional structures) from long range unmanned aerial vehicles (UAVs), otherwise known as drones, to detect cocoa growing under the forest canopy in Ghana. LIDAR provides high resolution data and, based upon a trial survey in Scotland, should be able to detect shaded cocoa growing under the forest canopy.

Can you use Radio Detection and Ranging (RADAR) to track cocoa production?

Forests 2020 has used RADAR data collected from Sentinel-1 in combination with optical data collected from the Sentinel-2 satellite to track cocoa production with varying levels of accuracy. The maps produced by RMSC have been reviewed by experts and have been scaled up to an area covering Ghana’s Western region. The outputs show that RADAR is capable of tracking cocoa production and differentiating between mono-cropped and shaded cocoa agroforests, but requires expert review and validation.

How does LIDAR compare to RADAR?

Both LIDAR and RADAR are founded on the principle of echo-location. When mounted on airborne or satellite platforms LIDAR and RADAR sensors emit pulses of radiation towards targets and measure the time for a signal to return to the sensor.

LIDAR can produce highly accurate three dimensional models (e.g. LIDAR can model a forest’s layers such as the trees’ canopies, under-storey vegetation and the underlying Earth’s surface) with centimetre accuracy. This means that the LIDAR collected data can be potentially used to represent accurately the forests trees’ canopies and branches. These LIDAR data can then be used to calculate forests attributes such as the forests’ layers heights and their cover as well the forest’s carbon content. In addition these LIDAR data can be used to calibrate coarser resolution satellite data such as from Sentinel-1 to assist with national or regional mapping.

Synthetic Aperture Radar (SAR) is a RADAR technique which can achieve great efficiencies when collecting RADAR data from space. The data collected by RADAR sensors carry information representing the interaction between Earth targets such as a forested area and the RADAR signal. This interaction is a fundamental property of RADAR data allowing the generation of two dimensional images which can be used to model forest properties. Due to the RADAR data collection technique (i.e. data are collected using an oblique geometry) RADAR cannot replicate the LIDAR data products.

Is there an alternative to growing cocoa in the forest?

Cocoa is naturally a shade loving crop which grows best in humid tropical forest regions. In Côte d'Ivoire and Ghana, cocoa is a very important agricultural commodity; 44% and 19% of the world's cocoa is produced in Côte d'Ivoire and Ghana respectively.  About 2 million smallholder farmers in the cocoa growing belt of West Africa depend on cocoa sales as part of their livelihoods.

There are alternatives to growing cocoa in the forest, but in this context, Forests 2020 (along with cocoa traders, chocolate companies and initiatives) would not want to cause disruption to cocoa farmers lives and contribute to wider social issues, such as increasing levels of unemployment and rural to urban migration. We believe that cocoa production can be forest friendly and can make a positive contribution to climate change mitigation if cocoa is produced in shaded cocoa agroforests rather than monocropped and that the most effective strategy is to ensure that cocoa production is sustainable for both farmers and forests.

What LiDAR and other geospatial products are used?

Our Forest 2020 partner Carbomap has been testing the use of LIDAR from both UAVs and airborne surveys to produce indicators of understory vegetation. We will also be testing Structure from Motion (SFM) from UAV platforms with our Forest 2020 partners, RMSC and KNUST, in Ghana.

What steps will be taken towards reducing deforestation by gathering more accurate data using LiDAR technology?

Gathering more LIDAR data will create the opportunity to identify shaded cocoa being grown illegally under the forest canopy in forest reserves and national parks and be used as training data for satellite imagery processing chains. These data will enable Governments and cocoa traders to develop solutions for sustainable cocoa production practices. For example, the Government of Côte d'Ivoire is exploring a model where illegal cocoa farms could be legalised on the condition that shaded cocoa agroforestry is practised.

What is the main focus of Forest 2020 and how does it contribute to the overall goal of protecting the forests while also protecting the cocoa industry?

Forests 2020 advocates and develops robust, accurate and sustainable forest monitoring systems. We have made significant progress with the development of forest monitoring products for the cocoa producing regions of Ghana that will assist the Government of Ghana, private sector and zero deforestation cocoa initiatives to monitor cocoa supply chains.

Our cocoa mapping products aim to contribute to the protection of the remaining natural forests by monitoring the spread of the cocoa production and highlighting areas where deforestation has taken place. For the cocoa industry, our mapping products can be overlaid with specific areas of interest (such as a profile of cocoa farms) to show which farms are producing monocropped cocoa, which are producing cocoa in shaded cocoa agroforests and which farms are currently located within protected areas (such as forest reserves and national parks) and/or are at risk of encroaching into such areas. These products will be used to monitor the cocoa supply chain which simultaneously protects legitimate smallholder cocoa producers, natural forests and the cocoa industry.

Tags: forestry, geospatial, GIS, health, imagery, intelligence, location, mapping, maps, remote sensing, satellite imagery

Categories: agriculture, analytics, cloud, field GIS, geospatial, GIS, government, image-delivery software, location intelligence, mapping, remote sensing, satellite imagery, sensors

This entry was posted on Thursday, April 25th, 2019 at 11:17 am. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.