Position extension: Data Scientist

Summary

For interested applicants with the necessary additional qualifications, this position serves as an extension to the Drone Pilot & Field Crew Leader position or the Field Crew Member position. The position will assist with data science and software development for multiple forest remote sensing projects. Key responsibilities include developing methods for annotating understory objects in drone and ground-based imagery for computer vision model training, developing software tools and analysis workflows for spatially co-registering multi-source data (drone imagery, ground imagery, and inventory data), building 3D understory models from 360-degree ground imagery, training and deploying computer vision models for tree detection and forest regeneration mapping, developing cloud-native workflow orchestration systems, and conducting data science experiments to advance automated forest inventory methods. The position may also include standardization of forest inventory data for inclusion in the Open Forest Observatory database and limited travel for short field trials.

Compensation

Approximately $27-30/hour or $53,100-69,073/year depending on degree, qualifications, and UC Davis HR assessment

Employment period

This position would extend the Drone Pilot & Field Crew Leader position or the Field Crew Member position by 5 or more months, potentially constituting a full 1-year position, with potential for further extension depending on funding and performance. Part-time work (commitment of 50-100% time) is possible. Start date between February and April 2026.

Location

100%: UC Davis office (with a potential exception for some limited travel for short field trials). Remote work outside of field season is permitted.

Purpose

This position will perform software development and/or data science experiments to support two projects:

An expansion of the Open Forest Observatory project (“Open Forest Observatory 2.0”) that is developing methods for automating forest inventory by combining over-canopy drone-based reconstructions with under-canopy 360-degree (e.g. GoPro) imagery to better capture understory trees, shrubs, and logs. This project involves development of computer vision models, 3D reconstruction methods, and data integration workflows.
A study of forest recovery 10-40 years post-wildfire (“Delayed Forest Reestablishment”) that will use drone and NAIP imagery to train and validate a computer vision model for mapping trees establishing following decades-old fires across southern, central, and northern California. This project involves computer vision model development, multi-source imagery co-registration, and accuracy assessment workflows.

Job description

Responsibilities will be tailored to the candidate’s skills and interests and may include, but are not limited to, the following.

Open Forest Observatory 2.0 Project

Develop methods for annotating key understory objects (e.g., trees, logs, shrubs) in drone and ground-based imagery for computer vision model training, train and supervise others in performing this task, and perform some of this work directly
Develop software tools and analysis workflows to spatially co-register drone imagery, ground imagery, and ground inventory data
Develop software tools and analysis workflows to build 3D understory models from 360-degree ground imagery, including assigning object classes (e.g., tree, log, shrub) to the 3D understory model by projecting labels from 2D images by employing (and potentially extending) Geograypher
Aid in development and expansion of a workflow orchestration system (using the Kubernetes-based Argo platform) for massively parallel, cloud-native, multi-step processing of drone imagery, including photogrammetry, post-processing of photogrammetry products (e.g. CHM computation, cropping), and tree detection
Develop software tools, analysis workflows, and data science experiments for using computer vision for geospatial tree detection in drone imagery, potentially employing multi-view raw drone imagery (as opposed to orthomosaics), and comparing the results to existing geometric (i.e. CHM-based) methods
Create and expand user-friendly documentation of OFO tools and data, including example workflows and tutorials

Delayed Forest Reestablishment Project

Develop methods for, supervise, and/or perform manual annotation of regenerating trees in co-registered drone and NAIP imagery
Develop software tools and/or reproducible analysis workflows to co-register drone and NAIP imagery, train computer vision models to detect forest regeneration, and perform inference on imagery across California
Evaluate the accuracy of forest regeneration inferences (human and ML) made from NAIP and drone imagery against ground-based validation data

Additional Responsibilities

Write and run scripts to transform ground-based forest inventory (stem map) data from external contributors into the standardized format required for the OFO ground reference data catalog
As time and interest permit, assist with limited travel for short field trials to test data collection protocols and workflows

Work location & camping

The duty station will be Davis, CA (UC Davis).

Work schedule

The work schedule will generally consist of a normal Monday-Friday workweek, with a potential for flexible work schedules and/or part-time work if preferred.

Minimum qualifications

Bachelor’s degree, Master’s degree, or Ph.D. in computer science, data science, ecology, environmental science, forestry, geographic information science, robotics, or a related field
Experience using Python for software development and/or data science, including collaboration using version control tools such as Git and GitHub
Experience collecting and/or analyzing drone-derived imagery of vegetation
Experience with scripted geospatial data processing and analysis
Excellent organizational skills for data management and project coordination
Attention to detail
Ability to work independently and make appropriate executive decisions without a supervisor present
Experience successfully collaborating with individuals from diverse backgrounds

Desired qualifications

Experience training and performing inference using machine artificial neural networks, including computer vision models
Experience developing geospatial data processing or data science workflows/software professionally, outside academia
Understanding of and experience training and/or deploying computer vision models such as convolutional neural networks, vision transformers, and neural radiance fields
Experience processing and analyzing 3D geospatial data, including photogrammetry workflows, point cloud processing, and/or multi-source data integration (e.g., drone imagery, satellite imagery, and ground-based measurements)
Experience with workflow orchestration systems and cloud computing platforms

Application due date

Review of applications will begin on December 22, 2026 and continue until the position is filled.

To apply

Please submit a cover letter (including your interest in the position, relevant experience, and availability dates), CV/resume, unofficial transcripts from any completed (within last 3 years) and in-progress degree programs, and contact information for three references (including name, organization, email, and relationship to you) using this Google form. The form includes more detailed instructions. For questions about the position, contact Derek Young, djyoung@udavis.edu.

Please note: for applicants with a Bachelor’s or Master’s degree and the relevant qualifications who are interested in BOTH the data science and field logistics position extensions, combined into a full one-year position, there is an alternative application portal that includes drone piloting, crew leadership, field logistics and management, and data science work. Given the long HR timeline for this alternative position, interested applicants should also apply at the alternative position posting by December 22. This alternative application is not required for candidates with a Ph.D., as the HR timeline for Ph.D. hiring is generally faster.