Leveraging Machine Learning to Minimize Climate Change Impacts

Leveraging Machine Learning to Minimize Climate Change Impacts

  • The Results
Project completed!

The World Resources Institute (WRI) has been seeking to understand how regional and global Nature-Based Solutions (NbS) (e.g. forest and landscape restoration) can be leveraged to address and minimize climate change impacts. More than 50 Omdena AI changemakers applied machine learning and Natural Language Processing (NLP) to develop an interactive dashboard, which visualizes the impact of climate change in various areas as well as showcases mitigating Nature-Based-Solutions.

 

The problem

In recent years many coalitions of organizations for NbS have been created. The African Forest Landscape Restoration Initiative (AFR100), whose goal is to place 100 million hectares of land into restoration by 2030, Initiative 20×20, which aims to begin restoring or protecting 50 million hectares by 2030, and Cities4Forests, where leading cities partner to protect and restore forests, are leading examples. These platforms promote the use of forest and landscape restoration and tree cover to enhance human well-being.

How Nature-Based Solutions help societies and ecosystems adapt to climate impacts is understudied and underutilized.

 

The solutions

The objective of the project was to assess coalition websites and their respective members’ websites. Using various AI techniques, the Omdena teams looked into uncovering if and how these platforms are approaching/mentioning climate adaptation. Also, the developed solutions help to understand how these platform are matching key climate risks to adaptation measures like NbS. In summary, the teams addressed the following questions:

  • How are some NbS platforms addressing climate hazards?
  • What type of NbS solutions are platforms and coalitions adapting?
  • What barriers and opportunities exist for the successful adoption?

 

One part of the final dashboard (see below) was to develop choropleth maps, which use colors on a diverging scale to represent a changed situation. The analysis considers yearly data of country-level climate and landscape parameters, such as land type cover, temperature, and soil moisture. The team also looked into deforestation evaluation. In addition, the engineers created heat maps that represent the level of relevance of each NbS platform and how each of the climate risks matches with the NbS intervention. Another element of the dashboard is a Knowledge-based Question & Answering NLP system. The system answers questions through text data from the NbS platform and lists PDF documents available on platform websites. Finally, a recommendation system uses content-based filtering or “wisdom of the crowd” to recommend items.

 

Machine Learning Climate Change

Dashboard demonstration

 

This challenge is hosted with our friends at

Articles from the challenge