Close menu

Modelling scenarios in collaboration with UEA

By Dr Juan Pablo Cordero, research associate at FixOurFood, working on the metrics work package.

During the third week of July, Professor Sarah Bridle and I visited Norwich and the University of East Anglia to work with Jez Fredenburgh and Prof. Neil Ward, looking to test the capabilities of AgriFoodPy and the FixOurFood dashboard.

Predictions on how the food system is going to look into the future and what interventions must be introduced to accommodate the growing needs of the population require a careful consideration of all the interconnections between different aspects of society and the environment. Our interpretation of data defines our ability to build accurate predictions, or models. Their projected accuracy is crucial to support evidence based policy making.

The future state of the system projected by these models as a function of our actions from today to that future is what we call a “scenario”. Different scenarios have been proposed by agencies, organisations, academics and governments. Based on the constraints and limitations imposed by budgets, resource management, trading agreements, data uncertainty, and in some cases, political agendas, these scenarios are typically contrasted against goals and expectations previously defined by present needs and problems.

Early this year, the Green Alliance, a think tank and charity in charge of researching and proposing policy and strategic work for environmental development, published a report with 5 scenarios aimed at describing the required changes in land use and management to achieve net zero in the UK by 2050. We identified these scenarios as a perfect set of target interventions to test the capabilities of AgriFoodPy, the software package we have been developing for the FixOurFood agrifood dashboard.

Green alliance scenarios can be described in terms of 5 key interventions, each of which has a different degree of relative importance in achieving net zero, depending on the scenario:

  1. Reduction of meat and dairy food production for human consumption
  2. Increase alternative non-animal protein uptake
  3. Adoption of engineered carbon-capturing technologies
    1. BECCS: BioEnergy with Carbon Capture and Storage
    2. DACCS: Direct Air Carbon Capture and Storage
  4. Replace high yield traditional farming with agroecological land use
    1. Agroforestry
    2. Silvopasture
  5. Forestation of agricultural land

We partnered with Jez Fredenburg (@overthefarmgate) and Prof. Neil Ward (@NeilWard586) to work together and identify how to translate the interventions described in the Green Alliance report into the set of models we have developed for FixOurFood.

  • Jez is a Journalist and editor, researcher at the Tyndall Centre for Climate Change Research, and Knowledge Exchange Fellow for the AFN Network+.
  • Neil is Professor of Rural and Regional Development at the University of East Anglia, and co-convenor of the AFN Network+.

Describing scenarios in the AgriFoodPy framework

‘Ambition levels encode how much effort and resources are employed in a particular intervention. We use these to compute the relative impact of the intervention in the model’

The first step was to identify the scale or “ambition level” assigned to each intervention. This numerical value, typically between 0 and 100, describes the level of importance of the intervention for each scenario. In some cases this represents the percentage or fraction of a quantity being transformed (e.g transformed land, consumed weight). In other cases, it can represent the relative value between two agreed boundaries. In some cases, zero ambition can still represent some level of residual transformation.

The second step of the modelling procedure is to identify physical values to connect with each ambition level. As an example, we have developed a model to compute the additional CO2e sequestration from each hectare of agricultural land transformed to silvopasture, agroforestry and forest. Neil and Jez’s experience and familiarity with the UK system proved invaluable here. How much agricultural land is there in the UK? What fraction is arable and what fraction for pasture? What is the expected yearly sequestration per hectare of peatland or broadleaf forest?


Finally, the challenge was to translate these ambition levels and underlying models into a consistent description of the food system at all times, integrated into the user interface we have developed for the FixOurFood dashboard.

We needed to make sure, for instance, that the total percentage of reconverted land does not exceed 100%; that a reduction of production is correctly consistent with consumption figures; that land repurposing correctly impacted production and the import-export balance.





Implementation into the dashboard and impact on our environmental metrics

Implementation into the dashboard was done via a drop-down menu which can be used to select each scenario. Upon selection, the sliders representing the different ambition levels moved to their pre-defined positions. These positions can be further adjusted to finer detail to create custom scenarios close to the ones proposed by the Green Alliance.










The table below summarises the positions pre-defined for each of the 5 scenarios

1. Balance food, nature and climate2.BAU3.Agroecological Farming4. Self sufficiency5. No BECCS/DACCS
Reduction of animal food450506070
Alternative protein50001020


Perhaps one of the most striking (and expected) results is the impact that diet change alone has on emissions. By setting the slider to 45% reduction of animal origin food (including dairy, eggs and sea products), emissions are reduced by a third by 2050, from ~300 Mt CO2e / yr to 200 Mt CO2e / yr. On top of this, repurposed land from animal farming and arable crops for animal farming and increase in cultured proteins can further sequester up to 60 Mt CO2e / yr.

The first Green Alliance scenario also considers heavy investment in Direct-Air Carbon Capture and Storage technologies which, if successful, can sequester up to 10 Mt CO2e / yr.

It also proposes an increase from current ~3% to 40% agricultural land transformed to agroecology which, while marginally reducing production initially, can achieve up to 95% efficiency and a reduction of 15 Mt CO2e / yr plus reduction in emissions from production.

These reductions are partly offset by changes in the import-export balance and product demand.


Green Alliance scenario 1: Balance food, nature and climate.

The orange bar shows the emissions by 2020, with bright orange  showing additional emissions from demand growth.

Green, blue and red bars show the emission reductions and sequestration from food production, land repurposing and CCS technology.





The full implementation of this exercise is available in the scenarios branch of the agrifood dashboard ( and the source code is open and available on GitHub: