SBIR Phase II: Arctic Environmental Modeling with Augmentation and Curation from an Artificial Intelligence Engine

点击翻译按钮获取中文摘要

The broader impact of this Small Business Innovation Research (SBIR) Phase II project will address the growing need for accurate models and forecasts of the Arctic. As Arctic maritime operations such as fishing, shipping, and mariculture of kelp are increasingly impacted by unprecedented climate change, traditional modeling techniques are unable to support these new demands. Ecosystems are not static, and their unpredictability hinders safe and sustainable economic development for communities. This modeling approach supports the economic competitiveness for Alaskan fisheries by increasing transparency of resources on short timescales (precision fishing) and long timescales (ecosystem modeling). The results could be used to identify new and emerging locations for fisheries, under- or over-fished locations, and differences between locations that can be restored or those that the ecosystem has shifted away from supporting. By applying dynamic ecology information, this project can provide tools to improve management of ocean resources, which could increase industry profits while simultaneously raising the total harvestable biomass.The Small Business Innovation Research (SBIR) Phase II innovation is to create and refine an Artificial Intelligence (AI) engine capable of generating custom environmental models, making new and emerging science quickly accessible to the people and communities that need the solutions. The AI-produced software will integrate multiple types of scientific techniques from the fields of nearshore bathymetry models, habitat mapping, precision fishing, and Ecosystem Based Fisheries Management (EBFM) tools. Scaling the generation of tailored models provides cost effective, adaptable, and accurate solutions to ocean environmental challenges. Successfully executing this plan will require a combination of technical and computing skills; numerical modeling expertise; significant scientific literacy in remote sensing, bathymetry, sea ice, physical oceanography, and fisheries science; and the development and training of novel neural network architectures. In Phase I work, remote sensing from satellites was used to map the nearshore in the prototype AI engine. Remote sensing from new satellite resources has enabled this effort of a deeper understanding of the world’s remote locations, like the Arctic Ocean. These data deserts can leverage satellites and pockets of Indigenous Traditional Knowledge to build productive, new economies that are resilient and adaptable.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.