Geospatial and climate data analysis for National Parks in arid ecosystems
Carl Boettiger, Professor
Environmental Science, Policy and Management
Applications for Fall 2025 are closed for this project.
Arid ecosystems such as the Mojave Desert are undergoing rapid shifts in climate and fire activity, with major implications for ecological resilience and species conservation. The Schmidt Center for Data Science and Environment (DSE) is currently collaborating with the National Park Service to develop ecological models that simulate vegetation response to these changing conditions in Joshua Tree National Park and the Mojave National Preserve. A key part of this work involves understanding the structure, variability, and limitations of available climate datasets to determine their suitability for use in these models. This URAP project will focus on exploring and analyzing high-resolution, downscaled climate projection data from CalAdapt for key variables such as temperature, precipitation, and evapotranspiration. Students will develop a workflow to connect to the CalAdapt API, establishing a repeatable method for data retrieval and analysis for future use in other parks. Time permitting, the project may also include exploratory work with satellite imagery, where students would help visually identify vegetation types in high-resolution images.
Role: We are seeking students interested in climate science, ecology, GIS, and/or computer science. Students will primarily work with climate projection data for the Mojave Desert to explore patterns in key variables across different time spans, emissions scenarios, and climate models. For example, they may analyze how projected precipitation trends compare to historical variability, or investigate how large-scale climate patterns like El Niño influence desert rainfall. Tasks will include analyzing trends, creating visualizations, summarizing findings, and establishing a pipeline for climate data retrieval. Through this project, students will gain experience working with climate model data, strengthen their skills in data analysis and visualization, and put open-source research tools such as GitHub and Docker into practice.
Students with experience in GIS and with an interest in remote sensing may also help estimate vegetation type cover from high-resolution aerial imagery (e.g., NAIP or Nearmap) to build a library of validation pixels that will support future remote sensing-based vegetation monitoring.
Training will be provided for all tasks. Students will work as part of a small team and attend weekly group check-ins scheduled to fit the group’s availability. We will be bringing on two interns who will work together on this project.
Qualifications: Applicants should have some prior experience with Python or R, and familiarity with basic statistics and time series analysis is helpful. Experience using Python for geospatial analysis is a plus. Applicants should be self-motivated and have high attention to detail.
Day-to-day supervisor for this project: Lucia S. Layritz, Maya A. Zomer, Post-Doc
Hours: 6-8 hrs
Related website: https://dse.berkeley.edu/programs/improve-wildfire-recovery-national-parks-beyond
Related website: https://cal-adapt.org/blog/climate-data-access/