Environmental Risk and Big Data workshop

Last week, I left the rurality of Cranfield and headed to the lofty heights of London town.

I was there to attend a NERC-funded workshop so named ‘Big Data to Solutions – Environmental Risks’. The workshop was organised and run by Cambridge University’s Earth System Science Doctoral Training Program alongside the Cambridge Institute for Sustainability Leadership (CISL). The latter organisations aim being to help leaders understand global changes and challenges.

The venue was Central Hall, Westminster which proved to be a great setting (see picture below) – although, I was actually hoping for the Parliament building when I set off! Following a welcome by Professor Simon Redfurn (University of Cambridge, Earth Sciences) and Eliot Whittington (CISL), the workshop was split into a series of talks, group discussions and a team activity to end the day in between numerous coffee breaks which contained a plentiful supply of biscuits! The aim of this post is to provide you with an overview of the days events.

Central Hall, Westminster

Central Hall, Westminster

Talk 1: James Stacey (CISL), Changing Global Context

In the first talk of the day, James Stacey (CISL, Corporate Fellow) considered the changing global context which we are currently facing. He identified that businesses are interested in identifying the potential environmental risks and opportunities, or changes, and how these might impact their respective businesses.

He pointed us to a recent Accenture survey which revealed that of 1,000 CEOs across 103 countries and 27 industries, 93% regarded sustainability as important and a further 78% regarded sustainability as an opportunity for growth. This has certainly been supported by Ian Cheshire, CEO for the Kingfisher Group. James also explained how the Bank of England’s Mark Carney has suggested that the next financial crisis may be triggered by environmental systemic risk and in particular how climate change will significantly impact the insurance industry. With specific relation to soils and agriculture, it was pointed out that the global crop production needs to double by 2050 to keep up with population demand!

James then moved on to the issue of climate change terminology. We have seen the change of ‘buzzwords’, for example from global warming through to the now more accepted climate change. A recent report by the World Bank called Turn down the heat  has anticipated this lack of terminological understanding.

Turn Down the Heat - World Bank

Turn Down the Heat – World Bank

James commented that especially in business, people require significant help in understanding the science of climate change. Furthermore, he suggested that a divide exists between experts and academics and that of industry. Importantly, the potential for significant impact lies with the industrial and financial sector(s). Improved dissemination of environmental risk information and data is therefore needed between the two sectors to ensure greater sustainability and resilience in the long term.

Talk 2: Emily Shuckburgh (British Antarctic Survey)

The second talk was given by Emily Shuckburgh from the British Antractic Survey. Emily is currently seconded to the Department for Energy and Climate Change (DECC), which meant that she was well placed to discuss the ‘Application of environmental data in business and policy’.

As the days events were centred around big data, her first slide covered the challenges of using environmental big datasets. After a quick brainstorm amongst the group, we identified these main challenges with big datasets;

  • Uncertainty
  • Accessibility
  • Wrong variables – e.g. climate models only output specific variables
  • Missing data points
  • Unstructured
  • Incomplete metadata
  • Data analysis tools
  • Data presentation

Emily provided us with an example of how environmental big data and expert opinion is used in policy-making; using Defras (Department for Food and Rural Affairs) policy cycle (see below).

Defra_Cycle

Defras policy cycle

Defras policy cycle is evidence driven, and highlights the importance of discussion between people building the evidence and those setting policy. However, the linear nature of the policy struggles to account for uncertainty. Therefore, it can only handle specific types of data.

This brings us on to the tackling of uncertainty in environmental datasets. Emily firstly stated that there are many different types of uncertainty. But more importantly, the question was asked about how we embody all of these uncertainties into the decision-making process.It was argued that many regard the issue of uncertainty as a sign of ‘ignorance’ on the part of the data collector or modelling expert.

An example of the Intergovernmental Panel on Climate Change method of uncertainty, and its weaknesses, were brought forward. For example, a poll revealed that 25% of people thought that the IPCC’s ‘very likely‘ scenario was actually indicative of less than 70% probability whereas, it actually represents the 90% probability. This links back to James’ earlier talk regarding terminology uncertainty in climate change discussions.

Probability

IPCCs uncertainty classifications

A further example, of the denominator bias, got us thinking about how journalists often grab us with their headlines and how environmental data could become misleading or even promote scaremongering amongst the general public. Denes-Raj and Epstein (1994), in their psychological study, discuss the use of the denominator bias in affecting peoples opinions on cancer patients. For example;

Cancer Kills 2,414 people out of 10,000

or

Cancer Kills 24.14 people out of 100

Which looks the more startling fact to you? The same denominator bias could be applied to a specific environmental factor.

Finally, Emily addressed the notion of real-world constraints to managing environmental risk. We are all human, and policy and scientific analysis is carried out by real people. Each of whom have a particular bias and concerns. Emily pointed us to the IPCC’s WG3 Chapter 2 Report which states that intuitive and deliberative (systematic) analysis is undertaken by people.

IPCC WG3

Intuitive and deliberative process characteristics (Source: IPCC, WG3)

We are often good at judging risks that we have had direct and repeated experience of. For example, in the UK this may be flooding or high winds. In contrast, we can often be poor at judging other risks (e.g. smoking) and other environmental risks which are likely (or currently are) to face us in the coming years and decades (see picture below).

risk_perception_and_actual_hazards_onwhite1

Perception and actual hazards (Source: ourchangingclimate.wordpress.com)

Emily summed up by suggesting that successful policy making requires both engagement and cooperation within an interactive setting between scientists, economists and policy makers to ensure we use big-data correctly when assessing environmental risks.

Talk 3: James Stacey (CISL) Business Perspectives

James once again returned to give us a talk on dealing with environmental risk from a business perspective. This mainly revolved around the reduction of Carbon emissions from our atmosphere. James pointing out that the UK government has indeed, by law, stated that it will have reduced Carbon emissions by 80% to 1990 levels by the year 2050.

coal

The idea of a circular, and greener, economy was then discussed as well as the rise of clean-tech industries. This involves the remanufacture of materials from existing products. An example that James provided was Levi jeans using plastic bottles in their jeans manufacturing process. The picture below shows another example of Levis use of recycled water in jeans manufacture.

recyclewaterlevi

Using recycled water in Levi Jeans production

Talk 4: Eliot Whittington (CISL) Government Perspectives

In our final talk of the day, Eliot gave us an overview of what government policies exist to mitigate the impact of environmental risk(s).

It was stressed that a close relationship exists between environmental stress and societal insecurity. Historically, both in the recent past and from the archaeological record we know that particularly food shortages can result in riotous behaviour, collapse of civilisations and in desperation amongst people.

Eliot asked the question, ‘Why does the [UK] government care about environmental data?’. There are several reasons why this is apparent;

  • Day to day use – e.g. weather reporting
  • Keeping a happy population
  • Maintaining of good economic supply chains
  • Management of infrastructure services
  • Public debate
  • Informing the public

With the latter, we as a public experience environmental data in our everyday lives. For example, from watching the weather report to understanding whether flooding is a risk to our property or business. This data can the ultimately lead us to make substantial commercial decisions.

One example of the governments use of environmental big data was the Environment Agency’s flood and air quality index maps available as an web-based public application (click on the map image below to see if you are at risk – UK only).

Screenshot of Environment Agency Flood Map

Screenshot of Environment Agency Flood Map

Next, Eliot discussed the policy responses of the UK government to environmental risk. He identified the UK’s Climate Change Risk Assessment (CCRA) being a step in the right direction, however, it only considers a world with a 2oC warming. Whereas, we are currently on track for a possible 4oC warming, and so does not help us plan for this possible eventuality. In particular, the CCRA has assessed the risk and opportunities posed to UK agriculture and forestry;

CCRA_RiskOpp

Climate change risk and opportunities for UK agriculture and forestry (Source: UK Climate Change Risk Assessment)

The Committee on Climate Change have taken a different approach to putting across potential environmental risk. They provide a direction of trend and the implication of trend (see picture below). However, what Eliot’s talk showed us is that although environmental data shows us that there is a problem (i.e. climatic change) we still don’t have all that we need to solve the problem at hand. Solutions are currently unclear, and dissemination of data uncertainty within policy still requires work. Eliot stated that politics is a complex and often messy process being non-linear and complicated. Perhaps we are currently at a point where our understanding of risk is not deep enough and although we find it easy to talk about the present, present actions are not so apparent.

agriculture1

Essentially, climate change is not a vote winner. Therefore, it does not always get the attention that it deserves. Furthermore, the media will tell us how much climate change will cost us, but not necessarily how much it will cost us if we don’t change and adapt – the useful graphic below gives us an indication of the costs of climate inaction versus action.

Action_vs_Inaction_500

Source: www.skepticalscience.com

Coffee growing Exercise

For the final activity for the day, we were split into a number of groups. We were then given a hypothetical situation where a temperature increase was set to hit Uganda’s coffee growing industry. The different groups were split to represent the views of a International Coffee Company, the Ugandan Ministry and the smallholder coffee growers themselves.

Within our groups, we were asked the following questions;

  1. Consider the priorities for each stakeholder group
  2. Data requirements

What this exercise revealed, at least in the capacity of a bunch of PhD students and Post-docs representing each of the respective stakeholders, was that we had similar data needs, but on differing scales.  We also all considered an increase in temperature to not just be an environmental risk but perhaps also an opportunity. For example, groups suggested the growing of different coffee types, growing a different crop altogether or exploitation of the coffee industry to tourists of Uganda.

However, the majority of the data needs that we identified were perhaps too idyllic. We were starkly reminded that data availability in a real-world situation is often a problem. Whether this is due to commercial or security reasons or simply that the data is not available for your study area. Therefore, it is something we all need to consider in our own research when trying to evaluate environmental risks and opportunities.

Summary

Overall the day proved extremely useful. It gave me a further regard for the interpretation of environmental risk from a range of stakeholder perspectives. This ranged from the general public through to the policy-makers responsible for keeping the public out of harms way of environmental change. However, both in the UK and globally, we have a responsibility for ensuring damaging climate change does not occur. We must remember, and these sentiments were expressed by a number of the days speakers, that environmental risk and opportunity should instead be regarded as environmental change. It is to this environmental change which we need to adapt to ensure a resilient society and economy. Soils and agriculture will play a key role in this.

Author: O. Pritchard

The following post represents my interpretation of the speakers key messages and not necessarily their own personal views.

Follow the story on Twitter using the hashtag: #EnvRskSolns