Planet OS Datathon Highlights — Part 2
Last month, Planet OS and Berkeley Haas’ AMENA Center for Entrepreneurship and Development hosted a 2-day datathon that sought to leverage Planet OS’s Datahub to develop innovative solutions to the environmental challenges in the Middle East and North Africa (MENA). The event brought together more than 100 UC Berkeley undergraduate and graduate students, Planet OS data engineers, and mentors from across the Bay and we are very excited to showcase their projects.
In our previous blog, we covered our first place winners, Team 1, and their project to address water scarcity in Jordan with solar-powered water pumps and storage tanks. In this blog, we will be highlighting the second and third place teams and their projects that address flash flooding in MENA region and the relationship between climate issues and political instability.
2nd Place Winner — Team 10: Combatting Flash Floods with Algorithms
Unlike Team 1 which analyzed six climate variables, Team 10 applied a predictive algorithm from a recent research paper to strengthen flash flood preparedness. They recommended the use of atmospheric data such as potential vorticities and water vapor rather than historical rainfall data to accurately predict the location and timing of floods.
With a 70% accuracy, the algorithm may not be ideal for weather forecasting which typically requires above 90% confidence. However, as climate data becomes more accessible and available in this region, we anticipate that the algorithm’s precision will improve and become instrumental in building resilience to the risk of flash flooding in the MENA region.
3rd Place Winner — Team 15: The Relationship between Agriculture and Sociopolitical Stability
Team 15 set upon the very complex task of untangling the connection between environmental conditions and political stability in Iraq. Unlike Team 1 or Team 10, Team 15's solution drew on dozens of factors.
To measure the state of Iraq’s environment, Team 15 utilized the PlanetOS Datahub data to gather data on temperature and rainfall. Then, they compared those sources with World Bank on agricultural yields, malnourishment, and other macroeconomic indicators. Finally, they compared these results with data on a host of social and political issues including migration, Islamic State of Iraq and Syria (ISIS) attacks, and fatalities. To understand the complex relationships between these factors, Team 15 created a correlation matrix to easily identify closely correlated variables.
Ultimately, Team 15 found that four factors — ISIL attacks, the percentage of arable land, rainfall, and food production — were the most highly correlated and together drove the greatest impact on Iraq and recommended that the diversion of funding from sustainable energy development towards water irrigation projects that were damaged by recent political instability.
The Next Challenge
We hope that, regardless of whether one was a student or professional, everyone walked away from last month’s datathon having learned something new. We would like to thank every participant for spending their weekend to think about global environmental challenges in fresh and exciting ways. Each team brought a unique idea to the table and created an insightful visualization and actionable solution that has the potential to change how communities prepare to weather events across the world. We look forward to how UNICEF will apply these analyses and recommendations to assist the MENA region through the forthcoming years.
Stay tuned as we hope to continue hosting more open to the public datathons in the months and years to come.
If you’d like to be notified when new content becomes available, follow Planet OS on Facebook and Twitter or subscribe to the Planet OS newsletter.
