Postdoctoral Data Scientist Biodiversity Monitoring
Updated: 16 Jun 2021
Biodiversity is a topic more important than ever before. With biodiversity loss impacting our planet, well-being and economy, we need to find innovative solutions with the use of data and technology. Within the NIEBA-ARISE programme we are looking for data science experts to make a difference. Are you interested in the integration of data derived from different types of sensors aimed at biodiversity monitoring? Then we may be looking for you.
Data & Technology: The sensors include (semi-)automated sampling and monitoring devices such as wildlife and insect cameras and eco-acoustic devices that allow the identification of species through machine-learning tools based on image and sound recognition. Additionally, radar will be used to monitor aerial biomass flows (e.g. movements of insects, bats and birds) and environmental sensors (such as weather stations) will record temperature, wind (speed & direction), light, humidity and precipitation. The integration of such data from multiple sensors requires to align and calibrate data formats from different sensors and to aggregate data at appropriate resolutions for subsequent ecological analyses.
The position is part of NIEBA ARISE (Netherlands Infrastructure for Ecosystem and Biodiversity Analysis: Authoritative and Rapid Identification System for Essential biodiversity information), a programme funded by NWO under the National Roadmap for Large-Scale Research Infrastructure and aimed at establishing a unique end-to-end infrastructure for the interpretation of evidence of the occurrence of species from digital sensors and eDNA. The aim of ARISE is to innovate and build services to facilitate rapid species identification based on various input sources such as specimens, environmental DNA, sound recordings, photographic images and radar information. This will be achieved by exploiting recent advances in DNA sequencing technologies (e.g. establishing high-throughput DNA sequencing pipelines) and machine-learning (e.g. species identification through image and sound recognition). The services to be built within ARISE will make species recognition and biodiversity monitoring accessible for both researchers, industry, policy and citizens. Having such knowledge readily available will drive the future of biodiversity and our planet.
The position is based at the Institute for Biodiversity and Ecosystem Dynamics (IBED) of the University of Amsterdam (UvA). Next to this you will spend part of your time (e.g. one day per week) at the Data Science Centre (DSC) of the UvA.
The DSC mission is to enhance the university’s research and education by developing, sharing and applying data science methods and technologies. As a coordinating hub within the UvA Library, the centre is uniquely positioned to facilitate knowledge exchange as well as the training of skills related to digital and data-driven research and thus contributes to the innovation, diffusion and cross pollination of data science across the university.
As a member of the DSC you'll will engage with data scientists, engineers, and researchers from across the university in a unique network. Knowledge exchange would include typical data science aspects such as data integration, data mining, machine learning and big data handling and visualization.
What are you going to do?
You will focus on:
- integration of species information from multiple sensors, including cameras, sound devices and radar, and environmental information (e.g. from weather stations, phenocams, soil sensors, surface water sensors or remote sensing).
This will include:
- sensor data calibration and data format alignment (e.g. time stamp information);
- linking biodiversity data (e.g. species observations derived from cameras, sound and radar) to environmental data (e.g. temperature, wind, light, humidity and precipitation) and remote sensing;
- integrating bioacoustics data with camera trap images;
- coupling of radar information with camera trap and sound data;
- connecting digital sensor information (images and sound) with traditional field monitoring methods (e.g. malaise and pitfall traps, transect counts)
38 hours per week
Science Park 904