We have an exciting and diverse set of challenges for the GC event in 2022. Each challenge is presented by one of our partners and captures a genuine problem related to the core operations of their organisation. Titles and brief descriptions are given in the table below – further details will be provided closer to the event.



World Wildlife Fund

Can we make the nature stripes?

Inspired by the “climate stripes, this challenge will seek to create “nature stripes” to communicate the urgency of biodiversity loss. Effective visualisation can capture the global trends in a simple form and help mobilise public support – but what should the stripes represent? Which data should be used? What species or geographic regions? How could such visualisations be used in a communication campaign?

Attributing environmental impacts to their corporate owners

Satellite imagery can capture the environmental damage caused by human activities such as mining and industrialisation, but it is hard to link these observations to the corporate entities responsible for them. This challenge will use natural language processing and geospatial methods to identify the owners of fixed assets (e.g. mines, factories) so that they can be held to account for their environmental impacts.

Institute for Strategic Dialogue

Fighting climate denial: Identifying opposition narratives ahead of COP27

There remains an active “opposition” to any action to prevent climate change, who use narratives of denial and delay to weaken public support and hinder effective policy-making. Institute for Strategic Dialogue led a coalition of environmental organisations to counteract such activities during the COP26 negotiations in Glasgow. Looking ahead to COP27 in Sharm El Sheikh, this challenge will use natural language processing and news/social media data to identify the most prominent opposition actors and the narratives they might use to disrupt the COP27 negotiations – so that ISD can stage an effective counter-campaign.


Friends of the Earth

Assessing potential for onshore renewable energy in England and Wales

Renewable energy is essential for decarbonisation and climate mitigation, but it is very difficult to find suitable land to build new wind and solar facilities. Energy demand must be balanced against food and conservation needs, taking into account prevailing weather and the preferences of the local community – all in an uncertain future environment. This challenge will combine data from diverse sources (climate, land use, planning, social) to determine the most suitable locations for renewable energy and help local authorities make effective decisions.

Ethical guidelines for “sustainable AI”

Artificial intelligence is having widespread impacts on society and a growing public debate is considering some of its harmful effects – but what about the implications for sustainability? High performance computing is incredibly power-hungry, while hardware requires increasing amounts of scarce mineral resources. AI systems are often blind to environmental concerns and might generate unsustainable strategies in the many decision-making processes where they are now used. This challenge will consider these complex ethical issues and seek to develop guidelines for “sustainable AI”.

Global Carbon Project

Where should we plant trees for carbon sequestration?

Tree-planting is a simple way to soak up excess carbon from the atmosphere, but to be effective a huge number of trees must be planted – requiring a huge amount of land to plant them on. Alongside the ecological suitability of a given location, there are competing demands for land from agriculture, housing, conservation and renewable energy provision. This challenge will explore the potential for tree-planting in the UK, calculating the carbon sequestration opportunity for different locations balanced against other demands.


Real-time carbon budget estimation

The GCP produces annual reports on global carbon budgets and emissions of greenhouse gases. These numbers are produced using complex models that are expensive to run. It would be hugely beneficial to be able to estimate carbon budgets at shorter timescales and at local/regional scales. This challenge will explore whether machine learning models can produce useful estimates at high temporal frequency, leading towards the ambitious goal of real-time information.


Choose your own adventure!

We are open to other challenges suggested by students – if you have an idea, please get in touch with the organisers.