Pine Beetle

Problem

The Southern Pine Beetle is wreaking havoc on northeastern states: as the pine beetles move further north each year, they kill trees in masses, threatening the air quality and integrity of forests throughout the US. The most effective way to successfully combat the infestation is by taking preventative measures before the beetles take hold. But how can we use resources currently available to predict future beetle movement? Maybe even more importantly, how can we communicate the risk and necessary protective measures to people it will impact directly, like farmers, and people who can help, like forest rangers?

Solution

Using statistical methods and predictive insights developed by the US Forest Service, Project Pine Beetle seeks to aggregate and visualize pine beetle data in a cohesive website that will help all users and predict trends in beetle movement. For a holistic understanding of the situation, the site offers visualizations of historical trends for key variables, such as number of pine beetles or number of spots (clumps of infected trees) observed in a specific locale within a two week period. The predictive model itself is the perfect mechanism for understanding and planning preventative measures: the system generates a model for every forest in a selected state, and displays the data in an intuitive map layout.

Impact

Project Pine Beetle will provide forest rangers and scientists with the tools necessary to understand past, present, and future trends in pine beetle movements, completing the first step in preventing pine beetle infestations, and the foundation for all future work.

DIFUSE Data Science Component

Project Pine Beetle encompasses several major fields within data science and analysis, including database management, data pipeline construction, data visualization and statistical modeling.

The project manages large amounts of raw data from academic and governmental sources over a range of 30 years and several US states that had to be cleaned and homogenized. It implements USFS statistical models that translate those data into actionable insights that any public user can take advantage of to visualize and understand beetle movement in their region. Furthermore, the data it allows users to interact with was being shuttered by the Forest Service before Project Pine Beetle made use of it.

Thanks to the NSF DIFUSE program, Project Pine Beetle’s users are able to track pine beetle infestation through tried and tested statistical models and easily available visualization tools. These online tools and the data they make use of represent an intuitive and relatable way for any interested user to become more comfortable with data science.

For more information email: difuse-pi-group@dartmouth.edu