PROJECT: UAV-based wave height estimation for downscaling NASA-SWOT satellite products in the Pacific coast margin
PI: Dr. Sanjay Pyare, Professor, UAS
Through collaboration with NASA Jet Propulsion Lab (JPL), this seed project seeks to improve inference and applicability of NASA remote sensing products in nearshore coastal zones, with a case-study focus on the NASA’s Surface Water Topography Mission (SWOT) mission and corresponding UAV-based lidar technology to estimate wave height in Alaska and Hawai’i coastal areas and by further conducting research on effective approaches for the downscaling of satellite data using machine learning models. The pilot study would advance development of 2 collaborative proposals with JPL and partners: the NASAROSES solicitation, Integrated SWOT Water Field Campaign; and the NSF EPSCoR RII-FEC, or “Track 2”, grant program for cross-jurisdiction EPSCoR cooperation.
The overarching goal is to improve inference and applicability of NASA remote sensing products in nearshore coastal zones where data are coarse or gaps may exist, with a case-study focus on sea-level (wave height) estimation via the NASA SWOT platform. Specific objectives of the project are to:
- Investigate the utility of UAV-based lidar remote sensing techniques (sensor parameters, sampling regimes, post processing techniques) for estimating wave height under variable environmental condition; and further validate measurement accuracy with 3 other in-situ estimation techniques (pressure transducer on a nearby buoy, photogrammetry based elevations, and GNSS-RTK elevational data of ocean surfaces).
- Through collaboration with JPL’s remote sensing program, compare UAV-based remote sensing data to SWOT satellite estimates and investigate effective analytical methods, workflows, and appropriate parameters for downscaling satellite data. We will also investigate the use of machine learning to improve performance of downscaling methodology.
- Leverage these collaborations, pilot data, analytical techniques and scientific products toward upcoming proposal RFPs that have direct relevance to Alaska NASA EPSCoR.
PROJECT: Development of additively manufactured advanced antenna systems for Unmanned Aerial Vehicle (UAV) operations in Alaska
PI: Kapil Sharma, Assistant Professor, UAF
Unmanned Aerial Vehicles (UAVs) are crucial to remote areas of Alaska for several applications such as communication, delivery of packages, remote sensing, emergency management (e.g., fighting forest fires), forecast of disasters (e.g., avalanches), environmental monitoring, surveillance and reconnaissance, aerial photography, and traffic monitoring to name a few. In the past, several UAV systems have been used in remote regions of Alaska. These UAV systems have utilized conventional as well as additively manufactured antennas with different performance characteristics. Advanced antennas with higher performance characteristics such as high gain, wide frequency bandwidth, reconfigurability, and dual polarization are required for the wide area and frequency coverage remote areas of Alaska demand. This project aims to develop additively manufactured advanced antennas to be used in UAV systems which will be cost-effective and will provide superior performance compared to the existing antennas being used in UAV systems. The proposed additively manufactured antennas will have several advantages such as reduction in size, weight, shorter lead time, and low cost while also exhibiting desirable features such as low power consumption, operational capability in extreme environments, and superior radio frequency (RF) performance.
- Major activities: A stacked microstrip patch antenna was designed using commercial simulation software packages, namely, COMSOL and CST MWS. Prototype of the designed antenna was fabricated at UAF Makerspace lab using a 3D printed PETG substrate.
- Specific objectives accomplished: An advanced antenna for UAV operations was designed using commercial simulation softwares namely, COMSOL and CST MWS and fabricated using additive manufacturing technology. The performance and reliability of the fabricated prototype is being evaluated using several testing procedures.
- Significant results, including research capabilities developments: The developed antenna exhibits several desirable performance characteristics such as high gain, wide impedance bandwidth, circular polarization, and high efficiency. The design achieved gain greater than 8 dBi, impedance bandwidth greater than 10%, circular polarization over the impedance bandwidth, radiation efficiency higher than 70%, and a radiation pattern with side lobe levels less than -20 dB.
- Key outcomes: The successful outcomes of this project include:
- Development of additively manufactured advanced antennas for UAV operations
- Evaluation of the performance and reliability of the fabricated advanced antenna prototype
- Collaboration initiative to develop partnerships with NASA researchers for future advanced antenna systems development for UAV operations
PROJECT: Leveraging NASA SWOT for Classification of Hydrokinetic Energy Potential in Alaska
PI: Dr. Erin Trochim, Research Assistant Professor, UAF
This project aims to significantly improve hydrokinetic energy assessment in Alaska by harnessing the power of NASA’s Earth observation data. Leveraging data from the Surface Water and Ocean Topography (SWOT) mission, the project will develop new approaches to analyze river characteristics and identify optimal locations for hydrokinetic energy projects. By increasing our understanding of Alaska’s vast renewable energy potential, it enables decisionmaking and supports empowering rural communities to transition towards cleaner and more sustainable energy solutions.
The primary goal of this project was to advance the assessment of hydrokinetic energy potential across Alaska by leveraging NASA Earth observation data, specifically from the Surface Water and Ocean Topography (SWOT) mission. The project aimed to develop an innovative, data-driven framework to analyze river systems and identify optimal locations for hydrokinetic energy development, with the broader objective of supporting renewable energy transitions in rural and underserved Alaskan communities.
A central objective was to design and validate an unsupervised classification method for categorizing river segments by their hydrokinetic energy potential. This approach sought to move beyond traditional site-specific assessments by enabling scalable, statewide analysis using remotely sensed variables such as river width, discharge, slope, and water surface elevation derived from SWOT data.
To ensure robustness and reliability, the project also aimed to integrate supervised machine learning techniques to validate classification outputs. By comparing model results against existing hydrokinetic datasets and in situ measurements, the project intended to refine classification accuracy and establish confidence in the methodology. A key measurable objective was to achieve at least 80% classification accuracy across at least three hydrokinetic potential categories.
Another major goal was to produce a comprehensive hydrokinetic potential map for Alaska. This deliverable was intended to serve as a decision-support tool for stakeholders, including government agencies, researchers, and local communities, enabling more informed planning and investment in renewable energy infrastructure. By improving the accessibility and usability of hydrokinetic resource information, the project aimed to reduce reliance on diesel fuel in rural areas and promote sustainable energy development.
Finally, the project sought to contribute to Alaska’s research and technical capacity by demonstrating the application of advanced machine learning techniques to NASA Earth observation data. This included fostering collaboration with NASA programs and building methodological foundations that could support future research, funding opportunities, and operational applications in renewable energy assessment.
This project successfully advanced the assessment of hydrokinetic energy potential in Alaska by developing a scalable, data-driven classification framework. Major activities included:
- acquisition and preprocessing of large-scale hydrologic and geomorphologic datasets
- development of an unsupervised machine learning pipeline
- validation and refinement of classification outputs
- generation of statewide analytical products to support decision-making