As climate experts gather in Sharm El-Sheikh for COP27, there are new advances emerging in the world of AI and machine learning that can help them effectively tackle climate change
When Hurricane Ian swept through Florida this past fall, residents flocked to social media like TikTok and Instagram to document their harrowing experiences. And the world watched.
While social media can bring the world together, and even form an online community for people to help each other, these posts serve an even larger purpose: through Natural Language Processing (NLP), they give climate change analysts crucial information that allows them to make future projections that can inform disaster response plans more accurately.
It’s a new and emerging field, and Novacene is well positioned to grow this essential – and often overlooked – space of climate research.
Machine Learning: An Emerging Tool for Climate Change Preparedness
In the past few years, climate change experts have begun to take notice of AI and machine learning’s (ML) value in their work and reports from major publications, such as National Geographic, have touted the technologies’ use in tackling climate change.
In June 2019, a group of global researchers released a paper called Tackling Climate Change with Machine Learning which discussed ways in which ML can be applied to climate action. They were motivated by the recognition that there were significant opportunities at the intersection of ML and climate change, but that relatively few of them had gained widespread attention.
This research team eventually became Climate Change AI (CCAI), which has since created a global movement in climate change and ML that brings together researchers, engineers, entrepreneurs, investors, policymakers, companies, and NGOs. They believe that NLP has many applications for climate research and can be leveraged to tackle climate change and inform decision makers in a variety of domains – for example, climate policy, climate finance and economics, or social and behavioural science.
Natural Language Processing: An Important Element for Climate Change Research
Globally, tens of thousands of climate action plans and reports have been generated by corporations, cities, states, and national governments. Whether they’re done voluntarily or in response to regulatory pressure, these documents include climate assessments, climate legislation, agency reports, regulatory filings, and corporate ESG (Environmental, Social, and Governance) and CSR (Corporate Social Responsibility).
Daniel Spokoyny, NLP researcher and organizer of the CCAI Summer School, a program that brings together climate and AI specialists from around the world, says information from these types of documents is being disclosed in a variety of formats, which makes it difficult to evaluate and to compare reports across different sources.
He says NLP can help “read” and organize the information and present it in a standardized format.
“Although there are efforts for standardized disclosure, the vast majority of information is yet unprocessed,” he says. “One way NLP can help is by building better search tools and information extraction systems. Tools that make access to information significantly easier and faster help experts in their research, policy making, and building better assessments of areas which need focus.”
Maria João Sousa, also a member of the CCAI Summer School team, adds that most areas of society (such as governments and corporations) play a role in climate change mitigation and/or adaptation, and we are witnessing a digital transition with torrents of information generated at an exponential rate.
“ML approaches, and particularly NLP, can be an instrumental tool to tackle such challenges with use-cases such as summarization, translation, sentiment analysis,” she says. “Nonetheless, ML should not be regarded as a silver bullet but an important element for multidisciplinary collaboration in climate change research.”
Lived Experiences: How Novacene’s Expertise Can Help Inform Future Responses
As this space continues to grow, NLP is also emerging as a new tool to help climate experts understand an area of climate change that is not often talked about: people’s lived experiences of extreme weather events.
When people post online about their personal experiences during an extreme weather event, NLP can take that information and provide experts with a high-level view of how people are adapting to extreme weather events and what they are doing to mitigate the event’s impact on their homes and lives. This information can be used to help strengthen or better coordinate adaption plans or disaster response plans.
During CCAI’s summer program, Novacene’s data scientist Dr. Hengameh Hajipour worked with fellow participants from top institutions in a team project to test their theory that natural language processing (NLP), the area of AI focused on the analysis of text data, can be used to analyze tweets to help better understand people’s responses in the aftermath of a major hurricane. This proposal was the culmination of a team project where she joined forces with participants across different fields to develop ideas for how machine learning can be used to solve climate-relevant problems.
Dr. Hajipour and her team ultimately found that Twitter analysis can support climate experts in many ways – including providing a better understanding of how affected people respond to hurricanes, what they might need to respond more effectively, and who is most vulnerable. Through NLP, they were able to better understand how affected people respond during/after an extreme weather event, especially with regards to characterizing vulnerability, identifying and tracking adaptive behaviours, and observing changes over time. This critical information can be integrated into climate education, disaster risk communication, disaster response, and adaptation plans.
Dr. Hajipour is continuing to build on this research and lead Novacene’s innovation in this essential space that has the potential to help climate experts, governments, corporations, and others to better inform their climate reports and goals.