This project proposes using historical sea ice data from the identified sources, along with MATLAB data mining and digital image processing approaches, to understand changing weather patterns and climate conditions and their relationship with external factors such as human contributions to global warming. Specifically, in the Circumpolar North, the countries/regions of the Arctic and Sub-Arctic, the impacts of sea ice receding can be fatal. While polar ice recession is a well-known, frequently studied problem, some regions in the Circumpolar North are particularly vulnerable and urgently need to understand changing ice patterns. This includes Alaskan Native communities such as Kivalina. Kivalina faces an imminent threat of drowning in the Arctic Ocean, and the community is currently considering migrating inland. Many Alaskan Arctic communities, such as the Inupiat in the North Slope and in the Arctic North West, have also resorted to building seawalls or using sandbags along the Arctic Shore to prevent flooding as the ice melts at the start of summer. The impact of such erosion-mitigation studies needs to be considered in the local context, including how they change the melting ice pattern and the icescape. A time-series analysis of the past data may allow us to extrapolate the previous images to predict future sea ice patterns and their melting rate.
Rotten ice movement patterns [5] can be formulated by comparing image sets across subsequent dates and then employing a time-series data analysis. These patterns can improve predictions of Arctic coastal flooding [6] and erosion.
This project assumes students are already familiar with MATLAB at a basic level or are willing to self-learn using MATLAB tutorials. There will be no MATLAB sessions during class hours beyond the initial MATLAB introduction in an early class. Students are encouraged to clarify the problem and scope further with the instructor when in doubt. Students have significant freedom in this project-based learning activity. No prior experience with MATLAB is expected beyond a simple introduction provided to the students, with the expectation that they will install MATLAB and become familiar with MATLAB before attempting this project by going through the resources (included at the end of this activity). If a student is unfamiliar with MATLAB, they can use resources provided by the MATLAB community as part of their learning goal. While mastering MATLAB is not a learning outcome of this course, MATLAB is a great tool for data mining and image processing. As such, this is a useful skill to acquire in data mining. This activity helps students learn and appreciate MATLAB's complex features and functionalities, as well as its various toolboxes, in data mining and polar science.
The MATLAB Image Processing Toolbox [7] can be useful for processing these sea ice images. MATLAB can also be used with Python [8]; as such, student developers familiar with Python can choose to use both languages in conjunction for this activity. Previous research highlights how satellite image processing in the circumpolar north can help us understand the climate crisis [9]. This proposed MATLAB-based project aims to demonstrate the endless possibilities for using a series of images in data mining to find insights into the climate crisis and polar science.

The student will work on an interdisciplinary domain of polar science and image processing research, with impactful applications such as the climate crisis and Arctic coastal erosion, using MATLAB. The project will also provide exposure to diverse datasets, including data federation and data mining, as well as the challenges of executing digital image processing. However, the critical research challenge is observing ice images, usually from satellites, processing changes over time using digital image processing techniques, and identifying useful patterns across the images.

The students are senior-level undergraduates in computer science and computer systems engineering or MSc students in AI and Data Engineering. They are familiar with programming. However, no prior experience with MATLAB is expected for this project beyond a simple introduction provided to the students, with the expectation that they will install MATLAB and become familiar with MATLAB by going through the resources (included at the end of this activity). This project is part of the stacked course "Data Mining," an upper-level undergraduate elective that is also stacked as a graduate course. This is a take-home project. This project is worth 10 points (10% of the total course score).

This project has these proposed tasks to be performed with MATLAB:
1. Analyzing and identifying imaging data on polar ice, specifically in the Arctic Ocean.
2. Performing image processing across the identified images to spot short-term and long-term patterns in receding ice and ice movements.
3. Using time-series analysis to predict future movements of Arctic ice and their impact on the coastlines.
Our overarching goal is to use MATLAB digital image processing techniques to find changes and patterns across a series of polar sea ice images. The student may find additional image collections beyond the samples provided in this proposal. These are historical data. However, satellites and sensors can provide real-time data.

MATLAB is provided to students and educators free of charge under the education license. Students should have MATLAB installed on their laptops and have a basic understanding of MATLAB. The proposed tasks are only sample activities. Students may choose alternative tasks. Students can use the provided resources as required reading/learning materials before attempting this activity to ensure smooth progress.

While this activity is designed for a Data Mining upper-level elective course and a required course for an MSc in AI and data engineering, it can be easily adapted for other computer science upper-level courses, such as Machine Learning or Image Processing, or other graduate school programs, such as computer science. Students are free to choose alternative projects of similar nature, using MATLAB.

Students are assessed at three levels. First, the successful use of MATLAB in the proposed project or an alternative problem of the student's choice at a similar level of effort and complexity. Students should upload their code and documentation to Blackboard. Students are encouraged to upload their submission to a GitHub project and then upload just the repository URL to Blackboard. Second, the students are evaluated for their demo, and finally, the questions and answers that follow the presentation. The first part is evaluated based on the code and documentation students submit. The second and third parts are evaluated through the student demos and a follow-up Q&A session.
Students are encouraged to use the AI/LLM capabilities readily provided by MATLAB. However, they are also encouraged to learn basic MATLAB syntax if they are unfamiliar with MATLAB. MATLAB Onramp provides a comprehensive tutorial in the form of course modules. Students who are unfamiliar with MATLAB are encouraged to complete these training modules at their own pace before working on this project.
Rubric
The assignment is evaluated across three different aspects. Each aspect is judged on a scale of 0 – 3, from 0 being "Ungradable" to 3 being "Excellent.
- Excellent (3): The work is either perfect or has negligible flaws.
- Good (2.5): The work is of good quality. Only minor revisions would be needed for the work to be Excellent.
- Satisfactory (2): While the work has some room for improvement, the student has put in a good-faith effort to complete all the work and demonstrated sufficient mastery of the material. A few revisions would be needed for the work to be considered Good.
- Needs Improvement (1): The student has put in a good-faith effort to complete the work, but revealed a lack of mastery in the material that can be addressed via concrete feedback. The work could become Satisfactory or better with some major revisions.
- Ungradable (0): The student did not make a good-faith effort to complete the work. This includes not submitting the (relevant part of) work at all, but also situations like submitting only placeholder code or code segments merely taken from the textbook.
The aspects of the grading:
1. Code functionality and reproducibility: The submitted code is complete in its functionality. The program scales reasonably to the complexity of the task. As part of your documentation, please include details on how to replicate your setup and run your program. If your documentation is unclear, you are likely to receive low points for this component too, since it is difficult to reproduce a setup and confirm the code functionality when no proper documentation is included.
2. Code relevance: The submitted code connects well with the course materials (as justified through the README). i.e., the code does not merely function. But rather, serves its purpose as the learning material for this class. This ensures the project serves as a Data Mining assignment. As part of your documentation, clarify the relationship of the proposed solution to the given assignment and how it connects with the Data Mining course. This ensures your submission stays close to the learning goals of the data mining course.
3. Demo execution (including answers to raised questions): This demo shows a complete working project that showcases your data mining expertise! The demo is performed in class on the class that immediately follows the specified deadline. Students are expected to run their program and explain the outcomes. This is a demo. As such, no Powerpoint presentation is expected (although not discouraged – if you think having a slide or two will help you with explaining the topic, go for it!).
The last point is reserved for documentation completeness. The total of 1 point is for complete documentation, whereas 0 points are for a lack of documentation. A partial point (such as 0.5) is possible for incomplete documentation or documentation that leaves more to be desired.

[1] Sample Arctic satellite image: https://noaadata.apps.nsidc.org/NOAA/G02135/north/daily/images/1978/10_Oct/N_19781026_conc_blmrbl_hires_v4.0.png Accessed: September 22, 2026.
[2] Index of /NOAA/G02135/ https://noaadata.apps.nsidc.org/NOAA/G02135/ Accessed: September 22, 2026.
[3] Sea Ice Index, Version 4. https://nsidc.org/data/g02135/versions/4 Accessed: September 22, 2026.
[4] https://noaadata.apps.nsidc.org/NOAA/G02135/seaice_analysis/
[5] Junge, K., et al. "Physical and optical characteristics of heavily melted "rotten"; Arctic sea ice." Cryosph Discuss 2009 (2018): 1-30. https://tc.copernicus.org/articles/13/775/2019/tc-13-775-2019.html
[6] Irrgang, Anna M., et al. "Drivers, dynamics and impacts of changing Arctic coasts." Nature Reviews Earth & Environment 3.1 (2022): 39-54.
[7] Image Processing Toolbox https://www.mathworks.com/products/image-processing.html This page provides an overview of the Image Processing Toolbox, describing how to perform image processing, visualization, and analysis.
[8] Using MATLAB with Python https://www.mathworks.com/products/matlab/matlab-and-python.html This page provides insights into MATLAB and Python integration. This page can be helpful for students who choose to develop their projects partially in Python and then integrate them with the MATLAB components. This is particularly useful since many students are familiar with Python and Python data mining and image processing libraries, and such an integration allows them to develop certain features in MATLAB while utilizing their existing code in Python.
[9] Namir, I., Hussain, M. A., Ramanna, S., Liu, Q., & Kathiravelu, P. (2025). Satellite image processing in the circumpolar north: understanding climate crisis by predicting sea ice extent in the Arctic. Remote Sensing Applications: Society and Environment, 101797.