African drylands in the Anthropocene: a land cover change GIS-based modelling approach for untangling the society-environment nexus.

Land cover changes in African savannahs will be quantified and long-term, multi-temporal and high-resolution maps at regional scales will be delivered. For the first time, the project will link these changes with land degradation and underlying ‘Anthropocene’-related societal drivers, including the management of natural resources, land use and ownership, adaptation to climate change and conflict.

The ‘Anthropocene’ concept provides a conceptual framework which encapsulates the current global situation in which society has an ever-greater dominating influence on Earth System functioning. Land cover change and land degradation in the drylands of sub-Saharan Africa need to be studied within this context, since these regions have become the epicentre of a number of on-going debates on the nature and scale of human influence on the environment.

Intertwined with the issue of land degradation are social issues on the management of natural resources, land use and ownership, adaptation to climate change and conflict. Continental and global-scale land cover change assessments with satellite imagery can be used to address these issues. Moreover, land cover change models are increasingly being used to understand Earth system dynamics and provide early warning scenario analyses and evaluations of environmental management policies.

However, current models do not fully reflect the typical characteristics of the Anthropocene, such as the societal influences and interactions with natural processes, feedbacks and system dynamics, tipping points and thresholds. Remote sensing and GIS technology can provide powerful tools for investigating the complex interactions between multiple parameters, necessary in order to untangle the society-environment nexus.

The utilisation of high spatial resolution satellite imagery (e.g. Landsat) to study large-scale African savannah changes has, to date, been very limited. To overcome such limitations this study will apply, for the first time, advanced compositing, image-texture description and machine-learning classification algorithms to monitor land cover change in African dryland regions. This will enable the development of a unique high resolution, multi-temporal database on savannah land cover change.

An integrated numerical approach will then be employed to enable the investigation and modelling of the interrelationships between environmental change, land degradation and a range of socio-political themes. Based on the analysis of model representations of Anthropocene dynamics, the project will identify new ways to enhance the role of modelling tools to further our understanding of Anthropocene dynamics and address the sustainability of African dryland savannahs.

The project will achieve the following objectives:
• Months 1-3: Identification of the optimum imagery from the 40-year Landsat RS archive with regard to seasonality and cloud contamination.
• Months 3-12: Employment of state-of-the-art compositing algorithms, image-texture description and machine-learning classification techniques (e.g. local binary patterns, restricted Boltzman machines and random forests) to map land cover and land cover change.
• Months 12-30: Combination of the land cover change estimates with three key socio-political parameters to further our understanding of the socio-environmental nexus in African drylands. Specific interactions that will be the foci of the project are: (i) Land tenure and environmental change (months 12-18); (ii) War and conflict (months 18-24); (iii) Grazing policy (months 24-30).
• Months 30-36: Development of a land-cover change model for disentangling the complex suite of social, economic and biophysical forces that influence the rate and spatial pattern of land-cover change. The model will be used to support the exploration of future land-cover changes under the latest Intergovernmental Panel on Climate Change (IPCC) scenario conditions (i.e. the Representative Concentration Pathways or RCPs). Socio-environmental indicators will be developed and threshold values set for the adaptation of mitigation actions.


The PhD candidate would join a growing team of researchers in MMU’s Environmental Science Research Centre. It is expected that the findings of the PhD research will form the basis of a series of publications in high impact, international journals. Excellent written, communication, and analytical skills are therefore essential.

Desirable skills include experience in programming languages (e.g. Python, IDL, statistical language R) to enable analyses in proprietary (e.g. ArcGIS, ENVI) or open-source GIS and image processing software (e.g. Grass, Q-GIS)