Environmental Impact and Environmental Sourcing Risks in Humanitarian Supply Chains: A Decision Support Tool for World Food Programme in Sub-Saharan Africa

Sammanfattning

Background Climate change has become one of the leading drivers of global food insecurity, and its effects are felt most strongly in regions such as Sub-Saharan Africa. As environmental degradation and climate-related stresses increasingly affect agricultural production, large buyers of food com- modities become exposed to growing environmental sourcing risks. The United Nations World Food Programme (WFP) is the world’s largest humanitarian organisation and relies heavily on the procurement of agricultural commodities to deliver food assistance. These procurement de- cisions are currently guided primarily by price and availability. WFP has begun to integrate environmental considerations into its operations through ECODASH, a tool that monitors car- bon emissions and waste, but the scope of this work remains limited to those two dimensions. Other environmental impacts and environmental sourcing risks are not yet considered in a struc- tured way within procurement, which limits WFP’s ability to make fully risk-informed sourcing decisions. Problem Definition and Purpose The central problem addressed in this thesis is the absence of a practical decision support tool that combines environmental impact and environmental sourcing risk in a way that allows WFP to systematically assess, compare and prioritise country-crop sourcing options. The purpose of the thesis is therefore to develop, in collaboration with WFP, a decision support tool that in- forms sourcing decisions by presenting environmental impact and environmental sourcing risk indicators for country-crop combinations in Sub-Saharan Africa. Three research questions guide the work. The first concerns which environmental impact and environmental sourcing risk di- mensions are relevant for agricultural sourcing decisions in the region. The second concerns how these dimensions can be quantified and compared across country-crop combinations. The third concerns how the resulting indicators can be structured into a decision support framework that informs WFP’s sourcing decisions. Method The thesis adopts a pragmatic research philosophy and an abductive approach, combined with a multi-method design. A Design Science Research strategy was applied and structured according to the Design Science Research Methodology, with elements of Action Design Research incor- porated in the design and development phase to support iterative artifact development together with WFP. The work combined a literature search, analysis of secondary quantitative geospatial data and iterative qualitative input from WFP stakeholders. All spatial analysis was carried out in Google Earth Engine, using the Spatial Production Allocation Model (SPAM 2020) as a common crop-specific spatial reference. The tool was demonstrated by applying it to maize sourced from Benin and Tanzania, two of WFP’s largest current sourcing origins for maize in Sub-Saharan Africa. Conclusions The study identifies two analytically distinct but complementary dimensions of environmental ii concern in agricultural sourcing. Environmental sourcing risk is structured around abiotic crop stressors related to temperature, water, soil, and biotic stressors such as pests, diseases and weeds. Environmental impact is structured around five ecological consequences of crop produc- tion, greenhouse gas emissions, deforestation and land use change, biodiversity loss, water use and quality, and pesticide use. All dimensions are shown to be crop-specific and country-specific. Quantification is achieved through a shared methodological architecture anchored to SPAM 2020 and to a documented two-tier dataset selection process. Environmental sourcing risk indicators are calculated as production-weighted means across the crop production footprint in the sourcing country, while environmental impact indicators are expressed as a per-tonne intensity that can be scaled by procurement volume. Each indicator is reported both as an absolute value and as a ratio to a crop-specific African production-weighted reference mean. The final tool is struc- tured as an integrated but non-aggregated framework that presents a hazard-specific risk profile and a dimension-specific impact profile side by side, making environmental trade-offs visible without reducing them to a single score. The two profiles are presented as paired radar charts in which each indicator is plotted relative to a reference ring set at the African production-weighted mean, so that worse- and better-than-average conditions can be read directly for each sourcing option. The demonstration confirmed that the framework produces coherent, interpretable and directly comparable outputs across sourcing origins. An indicator-level plausibility check further indicated that the computed values are consistent with independent sources and documented regional patterns. Academically, the thesis connects the previously separate research streams of humanitarian logistics and sustainable supply chain risk management, and provides an opera- tional, indicator-based layer for assessing environmental sourcing risk and environmental impact together. In practice, the framework and its accompanying documentation give WFP a trans- parent and extendable basis for more risk-informed and climate-smart sourcing decisions. Contribution: This thesis has been a complete collaboration between the two authors. Each author has been involved in every part of the process and contributed equally.