Data Science student Greta has a passion for sustainability, so we asked her to highlight some of the ways in which Data Science and Information Technology can help us build a more sustainable future.
It’s no secret that the Information Technology sector is a huge consumer of energy. In fact, the UN environment programme estimates the tech sector is responsible for 2 to 3 percent of global greenhouse emissions [1].
However, within the sector lies a useful tool for combatting carbon emissions - Data Science.
The first way in which Data Science can aid sustainable development is by measuring the impact of climate change. This allows us to: 1) Monitor if our previous predictions match the current data, 2) Look at the relationship between global warming and socio-economic issues, and 3) Predict future climate issues.
 |
"Blue Marble (Planet Earth)"
by woodleywonderworks is licensed under CC BY 2.0. |
A 2021 article from the Geoscience and Remote Sensing Journal [2] provides a convincing explanation of how Big Data and Machine learning solutions can support 12 out of 17 of the United Nations Sustainable Development Goals.
Computer vision is an important part of Data Science, whereby a machine learning model can be trained to recognise certain objects or trends. By applying computer vision solutions to satellite images, the article suggests that algorithms can map; human presence and availability of energy, marine and coastal ecosystems, water acidity, natural resources, pollutants, areas at risk of extreme weather, and biodiversity.
On a smaller scale, local data can be used to monitor UK energy consumption and demand, aiding the efficient distribution of natural and renewable resources. An example of an interesting measurement tool is MyGridGB, a near real-time online platform which monitors where the UK is sourcing its power [3]. Sheffield has a special link to this Data Science project, as our very own Weston Park Weather Station contributes solar data [4].
 |
Weston Park Weather Station by
The University of Sheffield via Facebook |
As I write this article, the UK is sourcing most of its energy from fossil fuels, followed by nuclear energy, and wind energy. During the mid-July heat wave, the UK grid was taking a whopping 25% of its energy from solar power.
Collecting and monitoring climate data has a useful secondary purpose - predicting the future. Machine learning algorithms are excellent at predicting future trends by learning from historical data. This might include accurately predicting climate-related droughts, floods, energy shortages, crop problems, and areas of risk for wildfires. This all sounds quite negative! However, by being able to predict the future impacts of climate change we can be better prepared to tackle them. On a positive note, scientists have used data science to predict how changing our habits could slow the development of climate change [5].
Using Data Science to monitor and predict climate change is really useful. However, for me personally, the most exciting area of development is the use of Data Science to discover efficient and effective methods for tackling climate change head on. The development of carbon capture materials is a particularly exciting area of research.
Chemoinformatic modelling methods, which traditionally involve predicting the properties of medicines, have been applied to predict the carbon capture ability of certain molecules (Chemoinformatics is an interesting field in itself - check out what the Information School’s Chemoinformatics research group is working on here).
By using machine learning to predict the carbon capture capabilities of molecules, data science provides the opportunity to streamline the discovery of new and more efficient carbon capture and storage materials, thus reducing greenhouse gases and tackling climate change [6][7].
The research described in this article is only the tip of the iceberg (pardon the pun) in how data science aids environmental sustainability. So, whether it’s by monitoring satellite images, predicting crises or developing carbon capture materials, data scientists can feel proud that their field is front-line in the fight against climate change.[1] With new pact, tech companies take on climate change (unep.org)
[2] [2112.11367] Deep Learning and Earth Observation to Support the Sustainable Development Goals (arxiv.org)
[3] Dashboard - MyGridGB
[4] Weston Park Weather (@WPWeather) / Twitter
[5] Predictions of Future Global Climate | Center for Science Education (ucar.edu)
[6] Molecular investigation of amine performance in the carbon capture process: Least squares support vector machine approach | SpringerLink
[7] Toward smart carbon capture with machine learning - ScienceDirect
Comments