ESR is committed to Earth Action by partnering with industry leaders and educational institutions to develop solutions for the myriad issues facing our planet. Read about two of these projects below.
Our staff have deep geophysical and climate expertise. We have experience producing actionable information and analysis. We routinely work with remote sensing and in situ data from a variety of platforms, as well as develop and deploy instrumentation in the field. If you have a problem that requires geophysical- or climate-related components, ESR may have the expertise to support or improve your solutions or decision-making. Contact Michael Town or Julian Schanze to explore leveraging our expertise in data product development and analysis to collaboratively solve problems.
In 2023, ESR entered into a mission-aligned partnership with Reflective Earth (RE) to help promote solar reflective solutions to climate change. Michael Town leveraged his expertise in radiative transfer to develop satellite-based tools and models for RE to estimate the radiative impact of a solar radiation management (SRM) solution. Town worked with community partners to test his tools over experimental plots like parking lots and urban neighborhoods. As part of this work, Town also participates in the Shine On Collaborative, lending his radiation and climate expertise to discussions of equivalence between greenhouse gas emissions and impacts of SRM interventions.
During the summer of 2024, ESR staff (Schanze, Town, Anderson) led a group of data science and data engineering interns from Northeastern University (NU) through a geophysical problem with machine learning potential. A brainchild of Julian Schanze, the goal was to make a better prediction of Arctic sea ice extent in the Bering Sea using a new feature variable: sea surface salinity (SSS). Arctic sea ice extent predictions will be very valuable to industry and other geopolitical interests seeking high fidelity knowledge of the Northwest Passage in a changing Arctic. Using a physics-informed, supervised machine learning approach, the NU students found that SSS alone could match the predictive skill of almost all environmental variables typically used in this sort prediction.
