iharp > Research > Opportunities > SIP and Fellow recipients
Last updated 2025 June
The Institute for Harnessing Data and Model Revolution in Polar Regions (iHARP) announced its first collaboration opportunity with its Researchers through fellowships or a Synergy Incentives Program (SIP) in April 2024. After an extensive review of applications and proposals, two (2) Polar Informatics Graduate Fellows, one (1) Polar Informatics Visiting Research Fellows, and one (1) SIP recipient were selected
2024 Awardees
Polar Informatics Graduate Fellows
Polar Informatics Graduate Fellow is a Graduate Student who is co-mentored by an experienced researcher from Polar Science and Data Science. The student will bring in experience from the two disciplines of data science or AI and polar science.
Nidhin Harilal (University of Colorado Boulder), is working under the mentorship of with iHARP’s Co-PI Dr. Aneesh Subramanian at the University of Colorado Boulder.
Project Abstract: Temporal interpolation is critical for enhancing the temporal resolution of geospatial datasets, especially in atmospheric sciences where observational frequency is often limited. While traditional video interpolation methods rely on optical flow to model pixel motion, such assumptions are poorly suited to the complex and heterogeneous dynamics of atmospheric data. We introduce STint, an unsupervised temporal interpolation framework that operates at arbitrary temporal granularity without requiring motion estimation or ground truth labels. Applied across diverse atmospheric and geospatial datasets, STint consistently outperforms existing vision-based methods and standard bicubic interpolation in both reconstruction accuracy and structural similarity. Furthermore, analysis of the model’s latent representations reveals its ability to capture meaningful cyclic patterns linked to known atmospheric phenomena.
Donglai Yang (Georgia Institute of Technology), will work under the mentorship of iHARP’s Co-PI/ Co-Director, Dr. Mathieu Morlighem (Dartmouth).
Project Abstract: Accurate prediction of ice sheet mass balance relies on understanding basal conditions, particularly ice-bed interface temperature, but is limited by uncertain boundary conditions and sparse validation data. Ice-penetrating radar attenuation offers a promising proxy for depth-averaged ice temperature. We present two methods integrating observed attenuation with thermomechanical modeling for improved basal temperature estimation: (1) Gaussian Process Regression with 3D model ensembles, and (2) generative AI with 3D ensembles. Application to West Antarctica reveals widespread thaw near Pine Island Glacier and heterogeneous conditions upstream of Thwaites Glacier, highlighting a significant flow boundary.
Polar Informatics Visiting Research Fellow
Polar Informatics Visiting Research Fellow is a Researcher who will work with iHARP researchers as a visiting researcher at an iHARP participating institution. They may also mentor graduate students and undergraduate scholars through continued partnership beyond their visit.
Dr. Zhibo Zhang (UMBC)
Project Title: Exploring ML-based cloud masking and cloud top height retrieval in polar regions using PACE HARP-2 multi-angular observations
Collaborating with Dr. Jianwu Wang
Synergy Incentives Program (SIP) Proposals
iHARP-SIP is designed to bring interdisciplinary teams working together to strengthen links to HDR ecosystem towards identifying collaboration opportunities and leveraging distinct strengths via cross-institute coordination, communication, and new research seeding.
Project Title: Causality-at-Scale for Polar Regions (CSPR)
Project Team Members: Dr. Sahara Ali, Dr. Aneesh Subramanian, Ms. Sikan Li