The SCIPE CI-SIP (CyberInfrastructure Summer Immersion Program) is a month-long, virtual, and synchronous immersion combining development of key skills and abilities in Cyberinfrastructure (CI) with a research project applying CI skills. The program will include 4 weeks of immersion during which participants leverage CI resources and methods to support research. Participants will also access self-paced introductions to data science modules and have opportunities to develop their skills in advanced computing, data science, analytics and visualization. Participants will have opportunities to come up with a research question, collaborate with mentors, identify datasets, clean data and perform data analysis. Participants will present their learning experience at a virtual research showcase on June 25, 2027.

Program Dates

June 1 to June 25, 2027 with an orientation on May 26, 2027.

Monday-Friday, 9:00 am to 5:00 pm HST

Eligibility Requirements

• Undergraduate students enrolled at any institution in Hawai’i and the Pacific are encouraged to apply.
• All majors are eligible, non-computing majors are encouraged to apply.
• To be eligible to participate and receive the stipend per NSF regulations, students must be citizens of the United States, nationals of the United States (as defined in section 101(a) of the Immigration and Nationality Act), or lawfully admitted to the United States for permanent residence. All participants will be required to provide documentation of tax compliance prior to receipt of the stipend.

Stipend

$3,250

Application Deadline

February 28, 2027

*Applications received after the deadline will be added to a waitlist.

Questions?

Email: datasci@hawaii.edu

The post SCIPE CI-SIP 2027 Application appeared first on Hawai‘i Data Science Institute.

The SCIPE CyberInfrastructure Research Experience Opportunity (CI-REO) is a semester-long, mentored research program that supports undergraduate students through exposure to Research Cyberinfrastructure projects with researchers and cyberinfrastructure professionals during the academic year at 10 hours per week on a stipend. Research mentors are currently based on UH Manoa, UH Hilo and the University of Guam. Students not on based on UH Manoa, UH Hilo or the University of Guam will be offered remote options for participation.

Cyberinfrastructure (CI) refers to the integrated computing systems, data storage and management facilities, advanced instruments, visualization environments and the people who operate and manage them. They are all linked together by software and high-performance networks to improve research productivity and enable breakthroughs not otherwise possible.

It is highly recommended that applicants read the following information before applying to the program:

In your application, please indicate your interest in one or more currently available research projects. At this time, the program does not include a project focused on cybersecurity.

List of currently available research projects:

1. Estimating bird call density of any species of Hawaiʻian bird using Google Perch Bird Vocalization classifier: https://lohelab.org/ Project requires student to be physically at Hilo to participate, with an interest in ecology and conservation in Hawaiʻi. In the LOHE lab at UH Hilo, Jetstream2 is leveraged to run Google Perch Bird Vocalization classifier. This allows us to estimate bird call density of any species of Hawaiʻian bird from months of continuous passive recordings. Students can run validations on the Perch output, with the ultimate goal of estimate changes in bird abundance across the landscape over time. Research mentor: Dr. Patrick Hart.
   
2. Developing natural language processing models for Hawaiʻian: The student will build neural network models to support the revitalization of the Hawaiian language. This involves collecting data from Hawaiian books, newspapers, and other existing sources, and training and evaluating machine learning models for tasks such as language modeling and machine translation. The ideal student for this project would have strong programming skills in Python and prior experience with machine learning and neural networks (PyTorch). Knowledge of the Hawaiian language is desired but not absolutely necessary. Research mentor: Dr. Winston Wu.
   
3. Developing an LLM for astronomy, specifically to predict stellar parameters from spectra. The ideal student for this project would be a junior or senior with strong programming skills in Python and prior experience with machine learning and neural networks (PyTorch). The student would possess domain knowledge in astronomy. This model would take in spectra at different resolutions from different telescopes, e.g. Gaia, TESS, Gemini, and would be applied to classifying the properties of newly discovered stars and identifying new exoplanets. This is a collaboration with Dr. Cicero Lu at Gemini Observatory who is an expert in exoplanets and star formation. The student would gain experience training and fine-tuning large language models, and running computational experiments with multiple GPUs using Koa. Due to the large amount of data required, this project may also need Koa Research Storage. Research mentor: Dr. Winston Wu.
   
4. AI for zero trust architecture (ZTA) Basic fluency with Python is required. Experience using Jupyter, Python and AI/ML are desirable. Project involves 1) collecting appropriate data then 2) training and 3) evaluating an ensemble of classifiers of attributes of devices, users and workflows with the objective of producing a binary meta classifier over (trusted|untrusted) suitable for use as a component of a ZTA Policy Decision Point (PDP). The student will pursue the above research objective using available, on-premise CI that offers Jupyter notebooks with GPUs. Research mentor: Dr. Curt Dodds.
   
5. Building a data management & processing platform for scientific data, as the traditional platforms are not optimized for those data, especially when those data are associated with AI techniques. The ideal student would have (1) programming experience in Python or C++/C, and (2) experience in Machine Learning or Database Systems development. The proposed platform will be optimized specifically for AI applications on scientific data, which will improve the application data processing efficiency and save cost and leverage existing techniques. This project relies on cyberinfrastructure resources heavily. The main development is conducted on Koa and Jetstream2 currently, and the data is also stored there. Research mentor: Dr. Yifan Wang.
   
6. Volcano Data Analysis and AI-Driven Modeling using CyberInfrastructure Resources This project uses advanced AI and autonomous multi-agent systems to analyze large multimodal datasets from Hawaiian volcano monitoring stations, including seismic, gas emission, satellite, and thermal data. Students will leverage Jetstream2 and Koa resources to build collaborative agentic workflows for eruption forecasting, deploy computer vision agents for imagery analysis, and utilize LLM APIs (e.g., OpenAI, Anthropic) to autonomously synthesize scientific field reports and hazard bulletins. The work also involves creating interactive dashboards where AI agents dynamically retrieve and correlate predictive outputs with visualizations for hazard communication. The ideal student has strong Python programming skills, experience with deep learning libraries or LLM APIs (e.g., PyTorch, Hugging Face, OpenAI/Anthropic), and an interest in earth sciences and AI. Familiarity with GIS or geospatial data is a plus. Research mentor: Dr. Sukhwa Hong.
   
7. Extending the Hawai’i Climate Atlases: https://www.hawaii.edu/climate-data-portal/hawaii-climate-atlases/
   The Hawaiʻi Climate Atlases are a predecessor to the current Hawaiʻi Climate Data Portal (HCDP). These worked with several static datasets and are being modernized. This project is being developed using TypeScript, a typed superset of JavaScript using React as a framework. Currently, the climate atlas for rainfall data has been created. This needs to be extended to several other variables using this interface as a template. Research mentor: Jared Mclean
   
8. Hawai’i Climate Data Portal (HCDP) and Mesonet: https://www.hawaii.edu/climate-data-portal/hawaii-mesonet/
   The student will work closely with a team of climate scientists and cyberinfrastructure experts to develop creative ways to identify “bad” data in meteorological data from the Hawaiʻi Mesonet (HM), assist with software updates and data management on the Hawaii Climate Data Portal (HCDP), assist with front end web-design, and work collaboratively to develop interactive decision support tools that leverage Hawaiʻi Mesonet and HCDP data. Research mentor: Dr. Ryan Longman.
   
9. Protein-Protein Interaction Modeling Pipeline: This project builds a modular cyberinfrastructure pipeline for discovering, ranking, and structurally modeling protein-protein interactions. The student will first develop a literature and database mining component that queries NCBI, PubMed, and pathway analysis tools to identify protein pairs with experimental evidence of interaction but no existing structural model of the complex. Interactions will be ranked by the strength of experimental evidence (e.g., co-immunoprecipitation with mass spectrometry, yeast two-hybrid, antibody-based co-IP). The student will then use HPC resources including Koa to run AlphaFold Multimer and protein-protein docking software (e.g., HADDOCK, ClusPro) to predict binding interfaces and generate ranked interaction poses. Crystal structures from the PDB will be incorporated where available. Results will be deposited into a curated database for review. Basic fluency in Python is required. Experience with structural biology databases, molecular visualization tools (e.g., PyMOL, ChimeraX), and HPC job scheduling is desirable. The student will gain experience with large-scale computational biology workflows, HPC resource management, and AI-driven structural prediction on research cyberinfrastructure. Research mentor: Dr. Alex Stokes.
   
10. Intelligent Data Exploring Assistant (IDEA):
    This project gives an undergraduate researcher hands-on experience exploring multi-agent AI orchestration for coding tasks, working within the Intelligent Data Exploring Assistant (IDEA), an open-source agentic framework being developed at UH Mānoa to help oceanographers analyze complex datasets through natural language. The student will investigate how specialized AI agents — for data retrieval, code generation, analysis, and validation — can be composed into reliable pipelines using frameworks like LangGraph, benchmarking orchestration strategies against real oceanographic coding problems on UH’s Koa and Jetstream2 systems. Through a full research investigation — forming hypotheses, implementing and testing agentic workflows in Python, and benchmarking results — the student will gain experience in cutting-edge, time-relevant AI research while directly contributing to IDEA’s ongoing development. No prior oceanography background is required; Python fluency is recommended. Research mentor: Dr. Roderick Tabalba.
    
11. High Performance Computing (HPC) at Guam:
    Learn alongside mentors at the University of Guam to deepen your understanding of HPC workflows and infrastructure. Research mentor: Mr. Jed Henson.
    
12. Cyberinfrastructure-enabled workflows for learning and decision making at Guam:
    This project focuses on developing cyberinfrastructure-enabled workflows for learning and decision making under limited information. Many real-world systems, including environmental, ecological, and infrastructure systems, are only partially observed, resulting in incomplete, sparse, or noisy data. The student will learn how advanced cyberinfrastructure resources can be used to support data analysis, computational modeling, and uncertainty quantification. Students considering future graduate study in data science, applied mathematics, computer science, artificial intelligence, or related quantitative fields may find this experience especially valuable. Prior experience with Python, R, GitHub, scientific computing tools, research presentations, or poster presentations is beneficial but not required. Students should be willing to learn computational tools, engage in independent learning, and participate in regular mentor meetings. The most successful participants are typically proactive, responsive, and willing to communicate their progress, questions, and challenges throughout the research experience. Activities may include (1) accessing and utilizing cyberinfrastructure resources such as ACCESS allocations, TACC systems, Tapis, and cloud-based computing environments, (2) developing computational workflows for data preprocessing, visualization, and analysis, (3) implementing and testing state estimation methods, including Kalman filtering and state-space modeling, on partially observed datasets and (4) supporting reproducible research through computational notebooks, documentation, and workflow development. Research mentor: Dr. Jaeyong Choi.
    

Access this link to learn more about Research Cyberinfrastructure resources and services available to students and instructors.

Complementary to the mentored research component, students will meet weekly for workshops offered through the Hawaiʻi Data Science Institute, professional development and student cohort activities.

Spring 2027 Program Dates

January 11 to May 5, 2027

Eligibility Requirements

• Undergraduate students currently enrolled at the University of Hawaiʻi & the University of Guam, interested in advanced computing, integrated computing systems, data storage and management facilities, advanced instruments and visualization environments.
• All majors are eligible to apply.
• Basic fluency in Python is recommended.
• Students enrolled at the University of Hawaiʻi & the University of Guam should use their university- affiliated email accounts to apply.
• To be eligible to participate and receive the stipend per NSF regulations, students must be citizens of the United States, U.S. nationals (as defined in section 101(a) of the Immigration and Nationality Act), or U.S. permanent residents. All participants will be required to provide documentation of tax compliance prior to receipt of the stipend.

Stipend

$3,200 for full session participation

Application Deadline

November 30, 2026

Questions?

Email: datasci@hawaii.edu

The post SCIPE CI-REO Spring 2027 Application appeared first on Hawai‘i Data Science Institute.

Among the notable observations, Hawaiʻi Mesonet stations recorded storm rainfall totals exceeding 20 inches at three locations on Hawaiʻi Island, while a United States (U.S.) Geological Survey gauge at Mount Waiʻaleʻale on Kauaʻi recorded more than 38 inches.

“This website allows people to see how a major storm affected different parts of Hawaiʻi, not just through a few headline numbers, but in considerable detail as the event unfolded,” said Cherryle Manzano, who developed the new website with fellow HCDP climate data analyst Matthew Lucas.

Interactive abilities

The Hurricane Nolo website brings together observations from the more than 80 weather stations of the Hawaiʻi Mesonet, other weather monitoring networks, and satellite imagery to provide a detailed picture of the storm’s impacts.

An interactive animation lets users follow the storm hour by hour, showing rainfall, wind gusts and satellite images as Nolo approached, stalled south of Hawaiʻi Island, and eventually moved away. Users can select individual weather stations to examine how conditions changed throughout the storm. A second interactive feature displays daily rainfall maps, allowing users to explore how precipitation varied across the islands.

Statewide monitoring, accessible data

“Extreme weather event analyses such as this one for Hurricane Nolo demonstrate the tremendous value of maintaining a dense, statewide weather monitoring network and making its data readily accessible to the public,” added Thomas Giambelluca, professor emeritus at UH Mānoa, Hawaiʻi Mesonet leader and HCDP co-lead, now based at Hawaiʻi Sea Grant College Program.

The Hurricane Nolo website is the fifth extreme weather event website released by the HCDP team since March 2026. Previous websites documented two Kona low storms in March and Hurricanes Lowell and Lala. These event summaries build on the routine collection, quality control and analysis of weather and climate data.

The Hawaiʻi Mesonet and HCDP provide data and information that support emergency preparedness, weather forecasting, water resource management, agriculture, scientific research and public education.

The post Hurricane Nolo’s path, rainfall impacts mapped in new interactive website appeared first on Hawai‘i Data Science Institute.

Users can visit the newly launched Hurricane Lowell website to explore interactive wind and rainfall maps, station observations and satellite storm replays.

The most striking weather signal during the storm was the wind. At Līhuʻe Airport, the highest measured hourly average wind speed reached 56 mph late on September 7. However, because Kauaʻi’s weather stations are concentrated primarily at lower elevations, experimental high-resolution modeling was used to estimate conditions across remote terrain. The models suggest that winds over exposed high-elevation ridges were much stronger, locally exceeding 100 mph as the island’s steep topography accelerated airflow over ridges and through valleys.

Rainfall on Windward Oahu, Mount Waiʻaleʻale, Kauaʻi’s north shore

Rainfall across the western islands was also substantial throughout the event. On September 6, rainfall amounts of 5 inches or more along the Koʻolau Ridge sent flood waters down Waikāne and Waiahole Valleys on Oʻahu, forcing the closure of Kamehameha Highway.

On September 7, nearly all of Kauaʻi received more than 2 inches of rain, with many areas exceeding 5 inches. High-elevation and north shore sensors show Mount Waiʻaleʻale recorded more than 23 inches of rain on September 7, accumulating nearly 36 inches over the full three-day storm period. Meanwhile, on Kauaʻi’s north shore, where rising river waters forced the closure of Hanalei Bridge, the Hawaiʻi Mesonet stations at Waipā and Lower Limahuli recorded more than 14 and 17 inches of rainfall, respectively, during the storm.

Coastal impacts and interactive storm replay

The report also references information from Pacific Islands Ocean Observing System (PacIOOS) on the significant coastal hazards brought by the storm. Large surf and storm surge impacted south- and west-facing shores of Kauaʻi as the hurricane tracked to the west.

To make these complex meteorological observations accessible to the public and emergency responders, the UH team launched a dedicated Hurricane Lowell website. The site features interactive graphics that allow users to follow Lowell’s winds and rainfall through time, compare ground station observations with mapped estimates, and replay the storm using Hawaiʻi Mesonet measurements integrated with Geostationary Operational Environmental Satellite-West satellite imagery.

The post Hurricane Lowell brought 100+ mph winds, heavy rainfall to Kauaʻi, Niʻihau appeared first on Hawai‘i Data Science Institute.

Reading time: 2 minutes

Starting July 1, the Hawaiʻi community can receive free, personalized monthly climate updates tailored specifically to their own neighborhood or coastline delivered straight to their inbox. Developed by the University of Hawaiʻi’s Hawaiʻi Climate Data Portal (HCDP), the new Monthly Climate Summary tool translates complex scientific metrics into highly localized, plain-language updates on rainfall, temperature and drought conditions.

Whether you are a rancher monitoring drought signs, a water manager tracking watershed health, a teacher looking for real-world classroom data, or a resident wanting to stay connected to a beloved valley, the tool brings high-resolution climate data directly to the community.

“Not everyone understands how much rainfall typically falls in their area of interest,” said Ryan Longman, director of HCDP and Pacific Islands Climate Adaptation Science Center (PI-CASC) University Consortium program director at UH Sea Grant College Program (Hawaiʻi Sea Grant). “But if you let them know that it was 50% below normal, or that it was the third driest month in the last century; that’s the type of information that anyone can understand.”

Tailoring data to traditional boundaries

Subscribers can customize their summaries using traditional Hawaiian geographic boundaries—moku and ahupuaʻa—as well as standard climate divisions and watershed boundaries. This reflects the HCDP team’s commitment to making data meaningful within the state’s unique cultural and ecological landscape.

To sign up, users can visit the HCDP Climate Summary page, select an area of interest on an interactive map, and enter their email address. Users can subscribe to multiple locations. Each month, they will receive localized climate overviews alongside access to an online dashboard for digging deeper into current and historical trends.

“We wanted end users to be able to connect with the data from the places that meant the most to them,” said Cherryle Heu, HCDP climate data analyst and lead climate summary developer. “When you select a point on the map the numbers will update to tell you exactly what is going on at that location. Given that Hawaiʻi’s climate is so diverse, this type of tool allows for local insights that a single statewide average can’t capture.”

Community-backed resilience

The tool’s development was initially funded by NOAA’s National Integrated Drought Information System until the grant was canceled in December 2025.

“We were too far along to stop,” Longman said. Support from the DLNR Commission on Water Resource Management allowed the team to complete the tool, a grant from PI-CASC helped support outreach efforts, and the NSF EPSCoR Change Hawaiʻi project provided the cyberinfrastructure foundation that makes it all run.

The HCDP team consists of several grant researchers based in Hawaiʻi Sea Grant and the UH Water Resources Research Center. The Pacific Drought Knowledge Exchange Team also led this effort along with collaborators across the state who helped shape the final product.

The post UH launches free personalized climate summaries for Hawaiʻi subscribers appeared first on Hawai‘i Data Science Institute.

Reading time: 4 minutes

Researchers at the University of Hawaiʻi Cancer Center and UH Mānoa John A. Burns School of Medicine (JABSOM) were awarded more than $12 million from the National Institutes of Health (NIH) to establish the Pacific Center for Artificial Intelligence and Data Science in Medicine (PAC-AID)—a new research center dedicated to accelerating biomedical discoveries through artificial intelligence and data science to benefit people in Hawaiʻi and the Pacific region, and worldwide.

Funded through the NIH‘s Centers of Biomedical Research Excellence (COBRE) program, the five-year award will provide approximately $12 million through February 2031. The award was administered by the National Institute of General Medical Sciences, with a federal award date of June 4, 2026.

The center will be led by Principal Investigators John Shepherd and Youping Deng. Shepherd is chief scientific officer at the UH Cancer Center, and B.H. and Alice C. Beams Endowed Professor in Cancer Research at JABSOM. Deng is co-director, Genomics and Bioinformatics Shared Resource, at the UH Cancer Center; and professor and director, Bioinformatics Core Facility, at JABSOM.

PAC-AID will serve as a central hub integrating AI into biomedical research to improve health outcomes in Hawaiʻi. Over five years, the project will renovate the UH Cancer Center Data Center to establish a new Medical AI Core (MedAI Core), providing advanced, high-performance computing resources and AI expertise. Physically located within both the UH Cancer Center and JABSOM on the shared Kakaʻako campus, PAC-AID will directly fund four major inaugural research projects and establish a Pilot Projects Program to support more than eight new, locally relevant pilot studies.

“At the heart of our mission as a flagship research university is the drive to translate innovation into meaningful impact,” said Vassilis Syrmos, incoming chancellor of UH Mānoa. “PAC-AID is a vital expansion of that mission, enabling our faculty to harness the power of artificial intelligence to pioneer new avenues of biomedical inquiry that were previously unreachable, solidifying the University of Hawaiʻi at Mānoa’s role as a global leader in health innovation.”

“AI has the potential to unlock major medical breakthroughs and help people live healthier lives, and we need to take advantage of it,” said U.S. Sen. Brian Schatz, who played a pivotal role in securing the grant. “This new funding will help Hawaiʻi continue to attract top-tier talent and develop treatments and cures that will benefit people across the state.”

“The $12 million over five years will substantially strengthen AI and data science capabilities and support the development of the next generation of investigators,” said UH Cancer Center Director Naoto T. Ueno. “The research advances made possible by PAC-AID will further the UH Cancer Center’s work toward new understandings and treatments for cancer, to save lives in Hawaiʻi and the Pacific, and across the globe.” In addition, high paying jobs generated by this project will have a strong economic benefit for our local communities.

“At JABSOM, our mission is to improve the health of Hawaiʻi through education and research,” said Sam Shomaker, JABSOM dean. “This investment will help accelerate discoveries that address the real health challenges facing our communities and ensure that advances in AI translate into better outcomes for patients.”

Enabling workforce development

Shepherd and Deng will oversee efforts to build research capacity, as well as foster the next generation of independent investigators using artificial intelligence and advanced data science approaches to address critical medical and public health challenges. Beyond funding research, this award establishes a workforce development mechanism to support the next generation of faculty investigators.

“This COBRE award provides the critical infrastructure to bridge advanced AI computational methods with our specific clinical and community health challenges,” Shepherd said. “By fostering a collaborative environment for our investigators, we are equipping them with the technical capabilities to tackle the most persistent health disparities in our islands and turn complex data into actionable health solutions.”

“By the end of this project, we expect to have a nationally competitive Medical AI Core, four independently funded research leaders, and more than 10 pilot projects,” said Deng. “Through these efforts, alongside workshops and collaborative research opportunities, we will significantly strengthen Hawaiʻi‘s capacity for AI-enabled biomedical research and innovation to address important health challenges in our region and beyond.”

Transforming medical research with AI

Artificial intelligence is increasingly transforming healthcare and biomedical research by helping scientists analyze large and complex datasets, identify patterns that may otherwise go undetected, and accelerate the development of new approaches to disease prevention, diagnosis, and treatment.

AI has the potential to unlock major medical breakthroughs and help people live healthier lives, and we need to take advantage of it.
—U.S. Sen. Brian Schatz

Shepherd, who leads the Hawaiʻi and Pacific Islands Mammography Registry, is using a massive breast imaging database to develop AI biomarkers that predict cancer risk specifically for Asian and Pacific Islander populations.

The new PAC-AID will support similar faculty-led research projects, provide funding for pilot studies, and establish shared research resources that can be used by investigators across disciplines.

Initial funded projects include AI-driven research by UH Cancer Center researchers Kevin Cassel (using full-body imaging to triage skin lesions) and Elizabeth Nakasone (studying pancreatic cancer in Native Hawaiian and Japanese populations); UH Mānoa public health researcher Jonathan Huang (modeling environmental toxicant effects on fetal development); and JABSOM researcher Yiqiang Zhang (identifying genetic traits in congenital heart disease).

COBRE Phase 1 grants focus on the development of independent biomedical researchers and research centers, and are designed to have a 15-year cycle.

“PAC-AID is specifically designed to develop six to eight early-stage faculty members at the University of Hawaiʻi and across Pacific Island institutions who will use artificial intelligence and data science to address cancer and chronic disease outcomes in one of the nation’s most medically underserved and understudied regions,” Shepherd said.

He added that the expected impact to Hawaiʻi is substantial. “Our benchmark for success is that the funded faculty projects, four initially and two to four more when those graduate in year 3, and will later achieve independent NIH R01 (Research 01 level university) funding at an estimated $3.25 million per award. That represents a projected $19.5 million in additional federal research funding returned to the State of Hawaiʻi (six funded R01s) on top of the initial $12 million COBRE award itself — a combined potential economic and research impact of more than $31 million for Hawaiʻi‘s research and healthcare workforce. These estimates are for the first 5-year period.”

PAC-AID joins a growing portfolio of NIH-funded Centers of Biomedical Research Excellence at UH Mānoa, including the Diabetes Research Center, Integrative Center for Precision Nutrition and Human Health and the Integrative Center for Environmental Microbiomes and Human Health.

The post UH researchers awarded $12M grant to advance AI, data science in medicine appeared first on Hawai‘i Data Science Institute.

Reading time: 2 minutes

An innovative University of Hawaiʻi research team is using trail cameras and artificial intelligence (AI) to better understand how much fog contributes water to native ecosystems and groundwater recharge. Led in part by Honolulu Community College Professor John DeLay, the project focuses on Mount Kaʻala, Oʻahu’s highest peak and an essential watershed.

“If we can keep the project going long enough, we might be able to detect a change in the lifting condensation level—affecting the extent of cloud and fog at the canopy level,” DeLay said.

Recharging the aquifer

Supported by the UH Office of Innovation and Commercialization, the system has achieved up to 93% accuracy in identifying fog conditions across different sites in both day and night conditions. This allows researchers to build long-term records of fog frequency and examine how moisture patterns vary.

“Not only is it really important for these ecosystems, which have fragile native species that need a lot of moisture and water, it’s also recharging out water budgets and our aquifer,” said Joel Nicolow, a UH Mānoa graduate research assistant in the School of Ocean and Earth Science and Technology.

Turning to AI

Historically, measuring this critical moisture required highly specialized personnel and expensive, complex equipment. The team turned to technology for a solution.

“We wanted to look at using cameras as a much more scalable, cost-effective approach,” explained researcher Dylan Giardana of the UH Mānoa Water Resources Research Center. “We use a machine learning model to classify all the images that we’re getting from these trail cameras.”

Their analysis shows that fog frequency jumps dramatically from about 10% at an elevation of 600 meters to 70% at the 1,200-meter summit. Understanding these patterns is essential as climate change and invasive species threaten natural watersheds. Long-term data may also help land managers in adapting conservation strategies.

The post How AI and field tech are clearing through the fog on Mount Kaʻala appeared first on Hawai‘i Data Science Institute.

Reading time: 2 minutes

Tori Kiliʻohu Cashman Valdez has spent her years at the University of Hawaiʻi at Hilo training in everything required to care for the ʻāina (land)—studying ʻōlelo Hawaiʻi (Hawaiian language) over late nights with fellow haumāna (students) at the campus’ Kīpuka Native Hawaiian Student Center, then analyzing aerial imagery the next morning in the geospatial research lab.

This spring 2026, she graduates with a double major in environmental science and Hawaiian studies, headed into a master’s program where the full breadth of her training meets a single research focus.

For Valdez, the choice to pursue two seemingly different degrees was a natural extension of her upbringing in Waimea, Hawaiʻi and Kailua, Oʻahu.

“We very intimately see the connection between the condition of our ʻāina and its impact on our livelihoods,” she explained.

This “ʻāina-first” perspective is what led her to UH Hilo’s Ka Haka ʻUla O Keʻelikōlani, where she found a community dedicated to the perpetuation of culture.

“We cannot care for ʻāina in Hawaiʻi without that ʻike (knowledge) about language, about culture, about people, about history,” Valdez said.

Valdez said she was deeply inspired by her ʻōlelo Hawaiʻi class with associate professor of Hawaiian language and literature, Kumu Hiapo Perreira, whose passion for language unlocked a desire for her to find even more meaningful ways to solidify the intersection of ʻike Hawaiʻi (Hawaiian knowledge) in and across all areas of life in Hawaiʻi—including her interest in science.

Valdez’s time at UH Hilo has been defined by rigorous academic achievement and high-level research. Working in the Spatial Data Analysis and Visualization Lab under geography and environmental science Professor Ryan Perroy, she has utilized aerial imagery for coconut rhinoceros beetle detection and presented her findings at the Hawaiʻi Conservation Conference.

Despite a grueling schedule—often reaching 18 credits per semester in order to complete her double major within four years—Valdez found her “home away from home” through Kīpuka.

“Hilo takes care,” she noted, reflecting on nights spent playing music and speaking ʻōlelo Hawaiʻi with friends. “Hilo makes it easier [to be away from home].”

Valdez’s journey at UH Hilo doesn’t end with her undergraduate degrees. She has already been accepted into the Tropical Conservation Biology and Environmental Science master’s program, where she will use LIDAR, a laser-based mapping technology to protect cultural historic sites from climate change.

For more go to Kūkala Nūhou.

The post Bridging tradition and technology: Meet Tori Kiliʻohu Cashman Valdez appeared first on Hawai‘i Data Science Institute.

Apply Here by May 31, 2026

The post Fall Undergraduate Student Program: SCIPE CI-REO 2026 appeared first on Hawai‘i Data Science Institute.

The post Public Health Research Symposium this Friday, May 1 (2-4 PM) appeared first on Hawai‘i Data Science Institute.