Problem: The Met Office, the United Kingdom’s national weather service, needed to develop a business case for a new High Performance Computer which will enable more accurate seasonal weather forecasts.  The business case had to demonstrate how improved weather forecasts could deliver socio-economic benefits to the UK.  We were tasked with assessing the potential benefits to the UK food supply chain.

Approach: The study was structured around three key sectors of the food industry:  production, processing and retail.  We carried out in-depth desk research to understand the structure of the food industry, the impacts of weather, the predictability of significant weather events and available mitigation strategies.  This information was used to select seven representative major food types for detailed investigation, and a shortlist of target organisations to interview.

We undertook interviews with key stakeholders in each sector, to explore their current and potential use of weather forecasts to support decision making, and the potential benefits of improved forecasts to business performance.

We then analysed the results of this research to construct a series of decision threads and illustrated these using benefits maps.  Each decision thread described the dependencies between computing options, weather events, forecasts, mitigating actions and resulting benefits for a particular decision and food type.

Methodology: The qualitative analysis provided the basis for development of a quantitative benefits model, implemented in MS Excel.  A subset of the decision threads that represented a cross-section of the food products and industry sectors, and which also represented likely high-benefit examples, were selected.  These Use Cases were then examined individually in more detail, to assess the level of benefit that could be achieved by acting on a forecast and implementing mitigating actions.  For example, improved forecasts of adverse growing conditions for certain crops could be acted on to vary planting dates and/or seed varieties, thus mitigating the eventual loss of yield under those conditions.

The computing options were characterised by the forecast skill (accuracy and false alarm rate) that they would achieve.  The benefits were assessed using Monte Carlo simulation to take into account the uncertainties in the data.  The benefits model then aggregated benefits over all Use Cases.  Sensitivity runs were performed to assess the impact of key assumptions, including the likely take-up of forecast services by different organisations within the food supply chain.  The benefits model allowed results to be examined for each sector, food type and Use Case, or as a whole, over different time periods.

Outcome: The study developed a conservative and robust estimate of the socio-economic benefits resulting from improved seasonal forecasts, especially in the area of reducing food waste and better matching food production to retail demand.  The Met Office was able to use these figures to present a compelling case for an enhanced high performance computer facility.

In October 2014, the Met Office announced that funding for a £97m supercomputer had been confirmed, and that this would enable a “step change” in forecast accuracy.  “It will allow us to add more precision, more detail, more accuracy to our forecasts on all time scales for tomorrow, for the next day, next week, next month and even the next century”, commented Met Office chief executive Rob Varley (http://www.bbc.co.uk/news/science-environment-29789208)

Client comment: “The Met Office wanted to show how more accurate forecasts could deliver socio-economic benefits to the UK.  Andalus and the wider team of Jigsaw associates worked professionally and delivered the work on time and in budget.  Throughout the project they demonstrated a flexible, rigorous and intellectually challenging approach. They worked coherently as a multi-disciplinary team – project managed through a single point of reference – and their analytical strengths were demonstrated in their final report.” – Shane Vallance, Met Office Economist, February 2014.

Benefits: This study was complex and challenging both politically and technically, requiring us to engage effectively with key industry players, conduct research in the absence of recognised data sources, build a robust model with limited quantitative data, and interpret and communicate probabilistic results in a clear but rigorous manner.

By articulating and addressing the key challenges and risks from the outset, we were able to forge good working relationships with the client and other stakeholders, and to leverage the wide ranging knowledge and experience within the project team.

By communicating the project plan, with clear objectives, scope, roles and dependencies, and closely monitoring progress, we were able to make necessary adjustments in a timely and controlled fashion.

By embedding peer and client review into each step of the work, we avoided unnecessary rework, fostered effective collaboration and ensured the final report was fit for purpose.

Contact: Jane Parkin