Project 1: Better algorithms for complex differential equations

[This project is not offered in 2020.]

Project abstract:  Mathematics guide the design of improved algorithms for solving complex differential equations on modern computers. By decomposing complex structures into simpler blocks, the finite element method (FEM) is able to numerically solve partial differential equations on complicated geometries. The research of our group is aimed at developing more accurate and more efficient types of FEM through mathematical insights.
For Summer 2020, we are offering an undergraduate research  project in collaboration with a local start-up company Microstructure Engineering. The company’s goal is to develop affordable computational tools for predicting microstructural evolution and its effect on the properties of metals. The student will learn about field dislocation models, conduct parameter studies, and catalog results obtained using a certain type of FEM. The student will be introduced to High Performance Computing (HPC) and will be provided access to a local HPC cluster through membership in the Portland Institute for Computational Science (PICS). 
Keywords: computational mechanics; finite elements; plastic distortion
Faculty Mentor: Jay Gopalakrishnan  http://web.pdx.edu/~gjay/
Department: Mathematics and Statistics
Community Partner(s): Microstructure Engineering, Portland OR (owned by entrepreneur Dr. Saurabh Puri)
Tools to be used: Python 3, SLURM
Involves teamwork: Yes. The student will be expected to work closely with a PSU PhD student, a PSU faculty, and a company expert.

Project 2: Machine Learning for Geothermal Energy Prospecting

Project abstract: This project will involve the application of machine learning tools for geothermal energy resource estimation and discovery. The goal is to apply existing tools from supervised learning to predict the presence or absence of geothermal resources, with further studies incorporating Bayesian methods to determine the uncertainty in these predictions.

Keywords: supervised learning, Bayesian learning, geothermal energy resource estimation
Faculty Mentor: John Lipor  http://ece.pdx.edu/~lipor/
Department: ECE
Community Partner(s): USGS
Desired skills:programming experience (Python preferred but not necessary), strong background in mathematics
Tools to be used: Python libraries including scikit-learn, pandas, numpy
Involves teamwork:  No

Project 3: Reducing Pollution Exposure for Middle School Students through Machine Learning

[This project is not offered in 2020.]

Project abstract: For students attending Harriet Tubman Middle School in North Portland, concerns exist regarding exposure to harmful traffic-related pollutants. For students in the neighborhood, a key challenge is to recommend walking routes to school that limit exposure to these pollutants while minimizing the time required to walk the route. Using existing pollution data, the student will model this problem as a graph-based shortest path problem, which can then be solved using Python libraries such as NetworkX.

Keywords: shortest paths; dynamic programming; pollution exposure; air quality

Faculty Mentor : John Lipor http://ece.pdx.edu/~lipor/

Department: ECE

Community partner(s): Harriet Tubman Middle School

Desired skills: Programming experience (python preferred but not necessary), strong background in mathematics

Tools to be used: Python libraries including scikit-learn, networkx, pandas, numpy

Involves teamwork: No

Project 4: Sustaining and improving water-related ecosystem services (WES) in the urban environment is becoming important in many growing cities

Project abstract: Together with ongoing land intensification, climate change is also expected to shift the magnitude and timing of runoff in urban areas. Such modifications of the hydrologic cycle include increases in peak flow during the wet season and decreases in baseflow during the dry season. Our integrated approach to the urban WES research has focused on monitoring and modeling of the dynamics of provisioning and regulating WES with explicit engagement of local stakeholders. We have been examining how cities can prepare for climate change via smart planning such as the installation of green storm infrastructure (GSI) and sensors. We are interested in testing how such GSI helps mitigate negative consequences of climate change across different spatial and temporal scales. The REU participants will be involved in the installation and monitoring of stream temperature as well as investigating the dynamics of flow and water quality during specific storm events. Students will obtain field sampling as well as analyzing time-series water resource data that have been collected through various community partners.

Keywords: urban water; ecosystem services; spatial analysis; hydrology; global change

Faculty Mentor: Heejun Chang,  http://web.pdx.edu/~changh

Lab or team:  https://www.pdx.edu/geography/hydrology-and-water-resources

Department: Geography

Community partner(s): 

  • City of Portland
  • US Geological Survey
  • Oregon Department of Environmental Quality
  • Clean Water Services
  • Metro

Desired skills (but not required): GIS, statistics, system dynamic modeling

Tools to be used: ArcMap, R

Involves teamwork: Yes

Project 5 : Creating a computational model of the pregnancy complication preeclampsia

Project abstract: Potential undergraduate projects will be closely tied to work being done by my research team to develop computer simulation models for increasing understanding of the behavior of complex systems. Current and future research areas include: (1) How the subjective experience of concussion survivors influences patient recover trajectory (community partner: OHSU); (2) Measuring and identifying toxic levels of stress in children, and developing models to show how such stress is linked to long-term health (community partner: NRF Global Communities); (3) Measuring human resilience and understanding its relationship to human stress response and adaptation (community partner: OHSU); and (4) The diversion and misuse of pain medicines and its adverse consequences (community partners: Oregon Health Authority, OHSU). REU students will learn how the simulation models work, how to do sensitivity testing, and how to use such models to conduct, interpret and document virtual experiments.

Keywords: Health; Health Policy

Faculty Mentor: Wayne Wakeland, https://www.pdx.edu/sysc/faculty-wayne-wakeland

Lab or team: https://www.pdx.edu/sysc/research-system-dynamics-and-simulation

Department: Systems Science

Community partner(s): Oregon Health and Sciences University (OHSU)

Desired skills (but not required): Data Analysis, Programming

Tools to be used: Vensim, Netlogo

Involves teamwork: Yes

Project 6:  Towards the understanding of turbulent flows

Project abstract: A common thread found between volcanic plumes, wind farms, urban canopies and forests is their interaction with the atmospheric boundary layer. These flows possess a wide range of time and length scales with embedded non-linearities which give rise to the turbulence phenomena. In pursuit of understanding, it is beneficial to recreate the problem in a control environment such as a wind tunnel. Via laser-based non-intrusive techniques, quantitative observations are made on the physical/transport processes primarily driven by the equations of motion. In other instances, numerical simulations are required to represent the problem. Data analytics are continually employed to extract relevant information as to elucidate the flow physics. The work has direct consequences on efficiency increase in wind energy, prediction of ash dispersion from an eruption, building efficiency to name a few. Participants are provided with numerous options for possibilities of topics.
Keywords: turbulence; fluid dynamics; scaled experiments; data mining

Faculty Mentor: Raul Cal, http://web.cecs.pdx.edu/~cal

Department: Mechanical & Materials Engineering

Community partner(s): TBA

Desired skills (but not required): –

Tools to be used: Varies by project/topic

Involves teamwork: Yes

Project 7: Assessing future change in weather patterns associated with heavy rainfall over Bull Run Watershed

Project abstract:  Extreme precipitation is associated with a multitude of impacts on society and the environment. Among these impacts are challenges brought by heavy rainfall to drinking water quality. For example, heavy rainfall can wash sediment into waterways used for drinking water, requiring treatment of the water prior to delivery. This carries management implications for water providers. Anthropogenic climate warming can alter heavy precipitation in two primary ways. First, a warmer atmosphere can hold more water vapor, so all else being equal, rain can be heavier in a warmer climate. Second, storm frequencies and intensities can be altered bringing more or less intense precipitation depending on the direction of change. Understanding how these changes will play out in future decades is critical for anticipating adaptation measures for dealing with future heavy precipitation. In this project, we are focusing on assessing the projected changes in weather patterns currently associated with heavy precipitation over the Bull Run Watershed, the primary source of drinking water for the city of Portland. Using data from climate models provided by numerous modeling centers around the world, we are developing software to recognize patterns in the model simulated atmosphere that are likely to be associated with heavy rainfall over Bull Run. We are then quantifying whether and to what degree these patterns change in coming decades under future warming according to this large suite of climate model output. This work is in partnership with the Portland Water Bureau and currently in the first year of a two year project.
Keywords: climate change; extreme precipitation; climate models
Faculty Mentor: Paul Loikith (http://web.pdx.edu/~ploikith/)

Lab or Team: Climate Science Lab (https://www.pdx.edu/geography/climate-science-lab)

Department: Geography

Community Partner(s):  Portland Water Bureau

Desired Skills (preferred, but not required): Matlab, basic understanding of meteorology, ability to handle large dataset

Tools to be used:  Matlab and/or Python languages
Involves teamwork: Yes

Project 8: Biomolecular Structure, Dynamics and Function

Project abstract:  The Reichow Lab is focused on understanding how biomolecular nano-machines (e.g., proteins) assemble, interact and function at the molecular level. To achieve this aim, we apply high-resolution 3D-imaging by electron cryo-microscopy (CryoEM)—a revolutionary method that allows us directly visualize the structures of biological macromolecules, with near atomic-level detail. Modern CryoEM methods apply intensive computational image processing routines to classify and geometrically correlate 100’s of thousands of individual 2D-images (terabytes of data) in order to calculate a protein’s 3D-structure. Once a 3D-structure has been determined, we apply in silico methods to simulate the atomic-level molecular dynamics that allow these proteins to carry out their animated functions within the cell. The 3D structures and atomistic-movies provided by our research are used to advance the knowledge of how biological phenomena are performed at the molecular level, and to provide an architectural framework for developing novel pharmaceutical and genetic therapies to correct aberrant mechanisms associated with disease. REU students in the Reichow Lab will be immersed into these experimental and theoretical computational methods applied to describing the fascinating and dynamic world of molecular biology.

Keywords: biochemistry, molecular biomedicine

Faculty Mentor: Steve Reichow  https://www.pdx.edu/profile/dr-steve-reichow

Lab or team: The Reichow Lab, https://www.pdx.edu/reichowlab

Department: Chemistry and OHSU’s Department of Biochemistry and Molecular Biology

Community partner(s):  Oregon Health and Sciences University (OHSU) 

Desired skills (but not required): Biophysics, Bioinformatics, Computational Science or Molecular Biology; however, students with other relevant interests are welcome to apply!

Tools to be used: Computational molecular modeling,Molecular Dynamics Simulation, Statistical Analysis

Involves teamwork: Yes

Project 9: Urban Heat and Vulnerable Populations

Project abstract: With changes in our global climate, city managers are trying to find avenues for mitigating local impacts of urban heat waves on vulnerable populations (e.g. poor, older adults, children, isolated, etc.). However, without knowledge about where, who, and what heat waves impact, developing mitigation approach may be ineffective. An REU student will work with a interdisciplinary research team to evaluate options to describe and mitigate the distribution of urban heat in U.S. cities. Using a combination of qualitative quantitative, and/or spatial methods, the student will assemble empirically derived datasets on urban heat, and integrate demographic datasets to assess the impacts of future heat events on human health and well being.

Project 10: Management of Urban Forests

Project abstract: Many urban planning organizations are advancing urban forestry campaigns with the goal of addressing challenges of climate change. At the same time, we have limited knowledge about the extent to which the rate of urban development is affecting the presence of tree canopy.  Using quantitative data from tree removals and development permits, the REU student will quantify and futurecast change to the urban canopy based on past/current trends, exploring alternative scenarios where relevant. Computational modeling will help to understand the factors that affect tree survival rates, while also contributing to decision making about urban natural resource management.

Project 11: Air Conditioning and Residential Energy Use

Project abstract: A warming climate will require communities to increase the amount of energy, most likely to keep a building cool during the hottest days of the year. While the number of households with air conditioning units is increasing, very little is known about future needs for additional energy. By assessing the extent to which residential buildings currently use energy, the REU student will explore patterns in the extent to which current and future climates will affect the extent to which communities will demand additional energy. Working with actual residential energy use data, and future projections of the ‘number of cooling days,’ we will develop models for predicting energy use across the city of Portland and other northwest cities as relevant. The results will be of relevant to examining whether the current energy grid infrastructure is capable of increasing demand.

The REU student can choose and co-develop from a couple of relevant project, which are described above. The broader title of these three projects is: Seeing C3PO: Cities, Communities, and Climate for Planning Organizations, which aims to prepare society for a changing climate. 

Faculty Mentor: Vivek Shandas https://www.pdx.edu/profile/vivek-shandas

Department: Urban Planning

Project 12: Numerical Algorithms for Solving Nonsmooth Optimization Problems and Applications to Image Reconstructions

[This project is not offered in 2020.]

Project abstract: Many optimization algorithms rely on the gradients or the Hessians of the objective functions, while solving optimization problems with nondifferentiable/nonsmooth objective functions is required in many recent applications.  The main goal of this project is to study mathematical foundation and develop numerical algorithms for solving optimization problems without requiring the differentiability of the data. Our first approach is to study generalized differentiation for nondifferentiable functions and develop numerical algorithms based on generalized derivatives of the objective functions. The second approach involves building smoothing techniques to approximate nondifferentiable functions by differentiable functions and then apply gradient-based optimization methods to these approximations. We will apply our methods to the image recovery problem aiming at building numerical algorithms accompanied with MATLAB codes for reconstructing images with missing or distorted pixels. We will also explore further applications to other areas such as medical imaging, missing data recovery, and building movie recommendation systems.

Keywords: convex optimization; generalized differentiation; smoothing techniques; image recovery

Faculty Mentor: Mau Nam Nguyen http://web.pdx.edu/~mnn3/

https://sites.google.com/pdx.edu/convex-analysis-optimization

Department: Mathematics and Statistics

Community partner(s): Thermo Fisher Scientific

Desired skills (but not required): Programming with MATLAB

Tools to be used: A laptop/desktop with MATLAB software installed

Involves teamwork: No

Project 13: Computational linear elasticity, with applications to geological phenomena

[This project is not offered in 2020.]

Project abstract:  A mathematical model for the deformation of a structure due to external and internal forces is provided by the equations of linear elasticity.  This system of partial differential equations can generally only be solved approximately, by transforming them into a large linear system of equations.  This project involves developing and testing code for computing and visualizing these approximate solutions.  The eventual goal is to produce software that can handle realistic geological models related to phenomena such as landslides, lahars and floods.  Depending on the background and interests of the student, the project could develop in different ways, and we will determine the most appropriate goals for the REU after we have been able to meet a few times.
Keywords: linear elasticity; numerical methods for differential equations; computational methods in geology
Faculty Mentor: Jeff Ovall (http://web.pdx.edu/~jovall/index.html)
Lab or team: There is no lab, but a desk will be provided.  If necessary, a laptop will be provided as well.
Department: Mathematics and Statistics
Community partner(s): David George, US Geological Survey
Desired skills: Ability to program in some compiled language—modern Fortran or C/C++ preferred.  The code you will be working with is written in modern Fortran, but you do not need to be familiar with that language in advance.
Tools to be used: Desk, computer, paper and writing implements, ingenuity
Involves teamwork: No

Project 14: Analysis and visualization of bridge condition data

Project abstract: Maintaining our aging and deteriorating infrastructure consisting of bridges, roads, and tunnels in a safe and functional condition poses a significant challenge. Transportation agencies charged with this task are always looking for new ways to better use the limited monies they receive. This project aims to help agencies make better decisions regarding maintenance and repair and how funding should be used, using publicly available bridge condition data. We are working on several aspects related to this: statistical analysis of bridge condition data, modeling (e.g. regression) of data, and data visualization. We use tools such as MATLAB, R, and Python for our work.

Keywords: infrastructure; bridges; data analysis; data visualization; modeling; asset management

Faculty Mentors:

Department: Civil and Environmental Engineering

Community partner(s): na

Desired skills: fundamental background in either statistics/probability, data analysis/modeling, or data visualization

Tools to be used: MATLAB, R, and Python

Involves teamwork: Yes, will work with graduate student on project

Project 15: Agent-based activity generation for city infrastructure and emergency planning

Project abstract: The goal of this project is to develop agent-based models to generate daily activities for a population of people that are characterized by a set of demographic attributes based on US census data. The daily schedules for each agent will consist of a sequence of typical geo-located activities, such as sleeping, shopping, working, etc., with their respecting starting and ending times. The socio-technical modeling approach can be used for city infrastructure planning and emergency scenarios.

Keywords: activity; people; simulation
Faculty mentor: Christof Teuscher, https://www.teuscher-lab.com
Lab or team: https://www.teuscher-lab.com
Department: Electrical and Computer Engineering
Community partner(s): Metro
Desired skills (but not required): Python, C, C++, Java, or Matlab
Tools to be used: custom made
Involves teamwork: No