Research Professional II (School of Geography, Environment, and Development) (Part Time) (Temporary)

Updated: over 1 year ago
Location: Tucson, ARIZONA
Job Type: PartTime
Deadline: Yes




Sign In
Create Profile



Research Professional II (School of Geography, Environment, and Development) (Part Time) (Temporary)
Tucson, AZ, United States
|
req10719
Apply Now
Share
Save Job

Back to Search
Research Professional II (School of Geography, Environment, and Development) (Part Time) (Temporary)
Posting Number req10719
Department Sch Geography, Dev & Environ
Department Website Link https://geography.arizona.edu/
Location Main Campus
Address Tucson, AZ USA
Position Highlights We have a NASA funded project through the Terrestrial Hydrology Program to fuse different satellite data products for urban flood mapping using deep learning. We will be processing different sensors and modalities of satellite image data (optical, radar) to extract flood probabilities in urban settings based on past events. To train our deep learning models, we have a hand-labeled flood image dataset using high resolution Planetscope data. The hand-labeled dataset consists of 200 annotated images of 1024*1024 pixels from 8 urban regions at a resolution of 3m. To augment our training dataset, we plan to generate and train our networks on weakly-labeled training samples. Our objective is to estimate inundation probability at 30m resolution every 3 days in urban areas with the help of our unique hand-labeled dataset. Our team includes University of Arizona (Beth Tellman- PI, Rohit Mukherjee and Hannah Friedrich at the Social[Pixel] Lab: https://beth-tellman.github.io/), Columbia University (Upmanu Lall, Pierre Gentine, Andrew Kruczkiewicz), and University of Virginia (Venkat Lakshmi). While we have deep learning experience on our team using CNNs to segment imagery and GANs for super resolution, we need to build custom architectures to incorporate time series data from high frequency temporal but low spatial resolution satellite sensors (AMSR2, SMAP) into our model (e.g. via RNNs or LSTMs). Work hours may vary, from 20 - 30 hours each week.
We are looking for someone proficient in building custom deep learning solutions for our specific task.
The deep learning architecture ideally should:
● Ingest any combination of satellite datasets for a specific urban flood event
● Combine high temporal resolution/low spatial resolution data with higher spatial resolution/lower temporal resolution data
● Extract useful information from each available satellite data
● Predict flood extent probability with reasonable accuracy
This position is expected to begin early July and ends September 30th, with the possibility of being extended based on availability of funding.

The University of Arizona has been recognized for our innovative work-life programs. For more information about working at the University of Arizona and relocations services, please click here .

Duties & Responsibilities
  • Understand urban flood mapping objectives from our team.
  • Work with the team to leverage the hand-labeled urban flood image dataset to train deep learning models.
  • Build deep learning models to incorporate time series data into our flood extent mapping model.
  • Build custom deep learning networks to fuse multiple sources of satellite image products, including:
       -Brightness Temperature: AMSR2, SMAP
       -Precipitation, Wetness: GPM
       -Digital Elevation Models: MERIT
       -Impervious Surface: GIMS
       -Optical Imagery: MODIS, Landsat, Sentinel-2, Planetscope
       -Synthetic Aperture Radar: Sentinel-1
  • Train deep learning models for different combinations of satellite image products on the hand-labeled urban flood image dataset.
  • Prepare well documented code for reproduction.
Minimum Qualifications
  • Bachelor's Degree in Computer Science or Related Field.
  • Minimum of 3 years of relevant work experience is required.
Preferred Qualifications
  • Graduate Degree in Computer Science or Related Field.
  • Experience with either Pytorch (preferred) or Tensorflow.
  • Experience training computer vision deep learning models (Experience with satellite imagery is a plus) for both spatial and temporal resolution.
  • Experience working with time series data and training RNNs or LSTMs.
  • Experience using Git.
  • Experience working with High Performance Computing clusters.
FLSA Non-Exempt
Full Time/Part Time Part Time
Number of Hours Worked per Week 20 - 30
Job FTE 0.50 - 0.75 FTE
Work Calendar Fiscal
Job Category Research
Benefits Eligible No Benefits
Rate of Pay DOE
Compensation Type hourly rate
Grade 7
Career Stream and Level PC2
Job Family Researchers & Scientists
Job Function Research
Type of criminal background check required: Name-based criminal background check (non-security sensitive)
Number of Vacancies 1
Target Hire Date 7/11/2022
Expected End Date 9/30/2022
Contact Information for Candidates Elizabeth Tellman Sullivan
[email protected]
Open Date 6/28/2022
Open Until Filled Yes
Documents Needed to Apply Curriculum Vitae (CV)
Special Instructions to Applicant
Diversity Statement At the University of Arizona, we value our inclusive climate because we know that diversity in experiences and perspectives is vital to advancing innovation, critical thinking, solving complex problems, and creating an inclusive academic community. As an Hispanic-serving institution, we translate these values into action by seeking individuals who have experience and expertise working with diverse students, colleagues, and constituencies. Because we seek a workforce with a wide range of perspectives and experiences, we provide equal employment opportunities to applicants and employees without regard to race, color, religion, sex, national origin, age, disability, veteran status, sexual orientation, gender identity, or genetic information. As an Employer of National Service, we also welcome alumni of AmeriCorps, Peace Corps, and other national service programs and others who will help us advance our Inclusive Excellence initiative aimed at creating a university that values student, staff and faculty engagement in addressing issues of diversity and inclusiveness.

Apply Now
Share
Save Job

 
  © 2019 The Arizona Board of Regents on behalf of The University of Arizona
Equal Opportunity Employer Minorities/Women/Vets/Disabled



Similar Positions