Everyone has a “map” to tell? Translating stories of participatory scenario narratives into maps of spatially explicit information

In research, stories are often created using scenario planning to understand future land use and land cover (LULC) changes. With scenario narratives, decision-makers can proactively consider uncertainties when choosing between different policy options. Many global scenarios need downscaling to the local level to make an assessment of potential futures possible for particular landscapes. While this approach ensures high-resolution LULC, it lacks context regarding local circumstances. On the other hand, scenarios developed on a local scale lack spatially explicit, quantitative information while providing opportunities for stakeholders to engage in decision-making processes. As few studies out there tackle these limitations by translating qualitative narrative scenarios at the landscape level into quantitative LULC maps, Duguma et al. (2022) propose a new, five-step approach.

---

A landscape in southwestern Ethiopia. Photo credit: Girma Shumi.

1. Development of the narrative scenarios
2. Current land cover mapping
3. Translation of narrative into qualitative spatially explicit rules
4. Scenario maps generation
5. Contrasting plausible future changes in quantitative terms between different social-ecological groups (e.g. municipalities)

---

Duguma et al. applied this approach in a case study in southwestern Ethiopia. Southwestern Ethiopia is a worldwide known biodiversity hotspot. Here, the landscape is made up of a forest-agricultural mosaic with large Afromontane forests and Coffee plantations, providing numerous ecosystem services to the population. The study area consisted of three districts or woredas (Gera, Gumay, and Setema; divisions of woredas are called kebeles and contain a minimum of 500 households) in the Jimma Zone, Oromia Region.

In the first step, four narrative scenarios were used that had been developed together with stakeholders, and were described in detail in a previous publication by Jiren et al. (covered on this blog here). In the first scenario, “Gain over grain”, farmers focus on growing and commercializing cash crops like coffee, khat, or eucalyptus, and traditional food cropping is abandoned. In the “Mining green gold” scenario, coffee production is intensified using modernized production approaches facilitated by large investors. In “Coffee and conservation” a biosphere reserve is established that combines sustainable agriculture, coffee production, and tourism. In the “Food first” scenario, food production is prioritized by developing conventional, commercial agriculture.

In the second step, Duguma et al. created a baseline map of current land uses. It contained 12 land use and land cover classes: forest, woody vegetation in farmland, arable land, pasture, cultivated wetland, grazed wetland, coffee plantation, eucalyptus plantation, khat, fruits and vegetables, rural settlement, and towns. These categories were derived by first using ground control points within supervised image classification and then increasing the thematic resolution to add categories based on the author’s knowledge of the study area.

Fig. 1 Baseline and scenario land cover maps. Arrows in the map indicate the plausibility of land
cover change from the current landscape to the four scenarios.

In the third step, Duguma et al. extracted qualitative rules from each scenario that were then translated into spatially explicit rules. For this, the authors used a combination of land cover classes, biophysical elements like altitude or slope, and the distances from the forest edge to create context-specific rules that directly link to the narrative scenarios. Additionally, transition rules were established to make sure all land use changes could be expressed in spatially explicit quantitative rules. Next, Duguma et al. used a proximity-based scenario generator based on the established rules to process the baseline map into four different spatially explicit scenarios of LULC (see Fig. 1). Lastly, the authors clustered the kebeles into four different groups. This way, no important spatial patterns of LULC change get obscured compared to using the entire study area. Duguma et al. used nine baseline variables to identify these distinct groups: First, areas of woody vegetation, pastures, and arable land; second, present levels of khat, eucalyptus, and honey production; third, altitude and kebele remoteness; and lastly wealth as a socioeconomic variable.

With their approach, Duguma et al. successfully integrated land-use mapping with participatory, narrative-based scenario research. Not only can this assess alternative social-ecological scenario outcomes, but also create an opportunity for decision-makers or stakeholders to proactively manage LULC changes, biodiversity, and ecosystem services. The authors stress how working with context-specific, spatially explicit maps of LULC change can aid in more discernable approaches to managing or adapting to socioeconomic changes at the landscape scale. With this methodological development, scenarios are advanced toward being a proactive tool, telling not only stories but also maps.

If you want to read about the details and results of the study, check out the paper here.

Advertisement