Keeping forests in mind while planning for agricultural and infrastructure expansion in Central Africa: Presenting an open-source and cloud-based geospatial tool for risk-informed land use planning with scenarios at five sites in Congo Basin

Keywords: Remote sensing, modeling, risk, GIS, land use, planning, deforestation drivers, food security, agriculture, Congo Basin, Africa
Abstract Type: Paper Abstract

Abstract

Central Africa’s Congo Basin forests constitute a globally critical continuous tract of forest yet are subject to overlapping proximate and underlying drivers of deforestation and forest degradation, such as industrial and artisanal agriculture, forestry, and mining, and infrastructure. Interactions between drivers at different scales also influence deforestation rates. For example, local and regional food security challenges in Central Africa combine with global commodity market pressures to drive forest conversion to industrial and artisanal agriculture throughout the Congo Basin that is neither sustainably planned nor risk-informed. This study presents Geoinformatics for Land Use Planning (Geo4LUP), a Google Earth Engine-based module in System for Earth Observations, Data Access, Processing & Analysis for Land Monitoring (SEPAL), an open-source, cloud-based computing environment operated by the Food and Agriculture Organization of the United Nations (FAO). Geo4LUP uses statistically- and ground-validated drivers data points from FAO’s Central African Forests Initiative and random forest modelling to produce a 30m-resolution deforestation risk output layer for a selected area of interest. The layer identifies risk of deforestation for intact forests in agriculture and infrastructure land use planning scenarios. This study demonstrates the tool using scenarios at five sites in Cameroon and Democratic Republic of Congo. At each site, we propose hypothetical roads and artisanal agriculture plots. Alternative solutions to avoid potential deforestation are suggested based on expected risk and driver pressures by site. We thus present and apply an open-source, cloud-based predictive geospatial tool to enable adaptive land use planning to address food security and deforestation risk simultaneously.