FOLLOW

FOLLOW

SHARE

4 Ways Big Data And Machine Learning Are Helping Conservation

The use of data and analytics isn't all about business

8Mar

The interdisciplinary field of computational sustainability is using machine learning algorithms to analyze and extract valuable insights from sets of data gathered from environmental fields. It’s not just about having large data sets or advanced pattern finding algorithms – it’s how we use them. The following projects highlight how machine learning and big data are helping conservation efforts of all kinds.

Earthcube Project

The ambitious Earthcube project has been in development for the past five years and aims to produce a living 3D replica of Earth to serve scientists of different disciplines. It has been built upon interconnected projects that use computer science, big data, and geoscience among various other branches of learning. There is a gold mine of data when it comes to earth sciences that have been collected over years of study and research that has the potential to benefit environmental causes greatly. Earthcube funds a variety of projects, such as the Coral Reef Science & Cyberinfrastructure-Network (CRESCYNT). By using species databases, image analysis software and 3d mapping they can monitor the decline of the coral reef’s structural changes, disturbances, coral disease, sea temperatures and coral bleaching. The research will allow for a greater understanding to help protect and preserve what we can of the coral reef.

The Great Elephant Census

Since 2006, over 12,000 elephants have been killed each year in Africa alone. The protection of these ecosystems are vital not only to wildlife but the communities around them and big data is helping. In 2014, Microsoft co-founder Paul Allen launched a 2-year survey, The Great Elephant Census with the goal of achieving a greater understanding of the numbers of elephants in Africa. Teams including 90 researchers traversed over 285,000 miles of the African continent, over 21 countries conducting this research. It’s resulted in one of the largest raw data sets of its kind. This data has informed African conservation efforts; reserves and wildlife centers have received more funding and rangers to support operations and security. The survey has shown that elephant numbers are down by 30% in seven years, showing 352,271 African elephants in 18 countries. The differentials in numbers highlight the need for ongoing monitoring to ensure better response times to emergency situations. Big data is having a positive impact on conservation efforts and will help better protect the elephant population of Africa.

eBird

Launched in 2002, the eBird app lets users record bird sightings as they come across them and input this data into the app. The aim is to help create usable datasets valuable to professional and recreational bird watchers. These data sets are shared with professionals across various disciplines such as teachers, land managers, ornithologists, biologists and conservation workers. They’ve used the data so far to create BirdCast, a regional migration forecast giving real-time predictions of bird migration for the first time ever, which uses machine learning and computer vision techniques to follow and predict migration and roosting patterns of different species of bird. This will benefit conservation efforts greatly by providing more accurate intelligence for land planning and management, and to allow areas prone to roosting bird gatherings time for necessary preparations.

Leafsnap

Leafsnap is an electronic field guide app available in North Eastern America, Canada and the UK developed by researchers from Columbia University, the University of Maryland and the Smithsonian Institution. The app uses visual recognition algorithms, derived from machine learning facial recognition techniques and allows users to identify species of trees from pictures of their leaves. The imaging system takes into consideration other signifiers such as flowers, fruits, and bark. The datasets available with Leafsnap includes 185 tree species, 23,147 lab images, and 7,719 field images, which are set to grow as the app develops. As stated on their website, Leafsnap aims to ‘build an ever-greater awareness of and appreciation for biodiversity.’ The City College of New York is finding educational benefits from using the app to support their curriculum. Data produced by app users and researchers can help conservation efforts through understanding how natural and man-made disasters can affect tree populations, distribution, and growth patterns. Through greater understanding of the natural world around us, we can work towards conserving it and data is playing a huge part in that.

Comments

comments powered byDisqus
Data culture small

Read next:

Building A Culture Of Data

i