EAGER: A Comprehensive Approach for Generating, Sharing, Searching, and Using High-Resolution Terrain Parameters

University Of Tennessee Knoxville, Knoxville TN

Investigators

Abstract

Terrain parameters quantitatively describe a landscape's surface properties (for example, slope or topographic wetness). Terrain parameters hold significant potential for advancing climate-related science and engineering efforts. As terrain parameters can be generated at different spatial resolutions, they are valuable resources for scientists working on soil moisture prediction, fire propagation, estimation of soil carbon content, soil respiration, and hydrology. These applications are critical for understanding land-atmosphere interactions and mitigating the impacts of climate change across ecosystems and landscapes. However, the process of deriving terrain parameters from digital elevation models is accurate, but expensive in terms of computational resources and time. For more efficient generation of terrain parameters, this project implements a flexible workflow to generate terrain parameter datasets at different resolutions for different regions of interest. All the products of this project (data, metadata, and software) are stored in an open-access commons to ensure they are Findable, Accessible, Interoperable, and Reusable (FAIR). The team of researchers promotes increased participation of underrepresented students, particularly women, through mentoring students in Systers (the organization for women in Electrical Engineering and Computer Science at the University of Tennessee Knoxville) and the collaboration with the Women in Data Science (WiDS) at Stanford. Terrain parameters are derived from Digital Elevation Models. High-resolution terrain parameters enable accurate spatial analyses and decision-making in climate-related science and engineering domains, but generating high-resolution data is computationally expensive, hindering the usability of terrain parameters for multiple applications. The project addresses this challenge to make terrain parameters available for climate study in three ways. First, the project implements a workflow to generate 15 terrain parameters at any resolution (from 30 km to 3 m) while preserving performance and accuracy. Performance is evaluated by measuring wall times and memory usage cloud platforms. The accuracy is validated by comparing the data with the derived terrain parameters. Second, the project uses the workflow and exploits data parallelism to generate large high-resolution datasets (i.e., down to 3 m) for North America (i.e., Canada, the United States, and Mexico). The project deliverable comprises rich metadata annotating the parameter values and Jupyter Notebooks for data search and access, reproducible data generation, accuracy validation, and performance measurement. Third, by bringing together an interdisciplinary research team of leading scientists with experience in data science and soil moisture dynamics, the project facilitates collaboration among federal agencies (including NSF, NASA, and USDA, among others) and institutions to pursue interdisciplinary research, share insights, and deliver innovative solutions for climate-related issues. 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.