Land and biodiversity policies/Data uncertainties limitations
Parts of Land and biodiversity policies/Data uncertainties limitations
| Relevant overviews |
| References |
Data, uncertainties and limitations
Uncertainties
The type of analysis described in this section contains major uncertainties that increase with longer scenario periods. All processes involved are described in the respective model-specific component, together with the related uncertainties (see Policy intervention Tables in the parts above). The key uncertainties from a systems perspective are discussed here.
An uncertainty with regard to agricultural demand is the shift in consumption and consumption patterns due to higher incomes, cultural preferences and urbanisation. Although consumption per person is not likely not to exceed 4,000 kcal per person per day, the proportion of animal products in the diet and the types of products will largely determine the land area needed per person. Thus, a key uncertainty is the current transition to more animal products in the diet of a increasing proportion of the world population. In addition, the possible increase in biomass use for energy plays a key role in total demand.
In modelling production systems, the main uncertainties are developments in productivity and yield increases. Historically, yields have increased approximately 1% annually. However, increased production and yields result from development of new technologies and breeds, and from increased adoption of existing improved management practices, both partly driven by commodity prices. Empirical data and quantitative modelling of these intensification processes are still rather poor (Hertel, 2011). In addition, the agro-economy model does not cover farmer objectives and constraints. And research and plant breeding focus on decreasing susceptibility to water scarcity, diseases and pests, but it is uncertain whether more efficient varieties can be developed. New technologies to unravel crop DNA could significantly accelerate plant breeding process (Godfray et al., 2010).
The key uncertainty in forestry is the use of fuelwood and sources of traditional biomass, mostly due to limited data availability (Component Forest management). A substantial quantity of fuelwood is still used in developing countries, and how rapidly the shift to modern energy carriers will occur is highly uncertain (see also Component Air pollution and energy policies).
In land-use regulation, the effectiveness of protected areas is not always sufficient to maintain original biodiversity values (Leverington et al., 2010). IMAGE only takes account of economic processes in the agricultural sector but the complex socio-economic interactions that drive deforestation (Lambin et al., 2001) are not included. In addition, the role of national institutions in land use and land-use change is not included, although it is well known that these institutions largely define landscape and land use developments.
The uncertainty range changes with the time horizon of the scenario, and is greater for indicators further down the modelling chain. For example, biodiversity results are largely driven by land-use change, which is sensitive to uncertainties about land allocation and production systems, and to agricultural demand.
Limitations
The IMAGE framework may be used to analyse policy issues in different ways. Most frequently, technical interventions or assumed behavioural changes are implemented, and the results from alternative scenarios are used to answer 'what if' questions. However, the costs and policy measures to bring about such changes often cannot be modelled internally, and the feasibility of such options cannot be taken into account. Some economic instruments in the agro-economy model (e.g., a meat tax) can be modelled explicitly, and also specific policies, such as those on REDD+, biofuels, or additional protected areas, can be modelled explicitly for their economic and system-wide effects. Other important transitions, such as behavioural change cannot be modelled.