Postdoctoral Positions in Forest Carbon Mapping with Planet Data

Two postdoctoral positions are currently available at Carnegie Institution for Science in a collaboration with Planet to map forest carbon stocks and emissions using a mix of LiDAR data and Planet's (0.8 - 3.7m VNIR) optical data:

Both of these positions will be based in Greg Asner’s lab at the Department of Global Ecology, Carnegie Institution for Science, on the campus of Stanford University in Stanford, CA, but will interact directly with Planet scientists and technical staff on project activities.  The positions will interact with other postdoctoral researchers, students and research assistants as part of daily activities related to the project.  Responsibilities center on research and collaboration with project partners, publishing in peer-reviewed journals, and outreach with environment professionals.

A Ph.D. in optical remote sensing or equivalent is required. Strong quantitative and English writing skills are important, and a working knowledge of computer programming for remote sensing applications is essential.


Computer Vision for Tropical Forests

A postdoctoral research position is available for a computer vision or machine learning expert to develop and apply algorithms to high resolution satellite imagery of tropical forest canopies.  The planned study involves the use of high-quality, high-spatial resolution satellite data (0.8m – 3.7m VNIR) from Planet (formerly Planet Labs) and Carnegie Airborne Observatory LiDAR data to map tropical forest stocks and emissions in support of science and conservation efforts led by project partner: the Erol Foundation.  Efforts include:

  • Building rigorous training datasets of Planet imagery or image chips with airborne LiDAR data.
  • Producing three-dimensional models of forest canopies directly from 0.8m resolution stereo pairs, and via structure-from-motion from 0.8m resolution panchromatic, full motion video from space.

Spatial Modeling for Tropical Forest Carbon Stocks and Emissions

A postdoctoral research position is available for a spatial modeling expert to develop and apply algorithms to high resolution satellite imagery of tropical forest canopies.  The planned study involves the use of high-quality, high-spatial resolution satellite data (0.8m – 3.7m VNIR) from Planet (formerly Planet Labs) to map tropical forest stocks and emissions in support of science and conservation efforts led by project partner: the Erol Foundation.  Efforts include:

  • Building automated, scalable algorithms to ingest a massively dense temporal feed of Planet imagery.
  • Produce automated maps of carbon stocks and emissions and associated uncertainties based upon existing algorithms to convert Planet data to carbon stock estimates.