Climate AI Student Researcher

Job Summary And Description

Join a team working at the intersection of climate science, data science, and energy system resilience. The Energy Systems and Climate Analysis Group at EPRI is a multidisciplinary team of leading earth scientists, climate economists and energy system analysts. We conduct research to inform real‑world decision‑making, helping electric companies and government partners understand how climate hazards are evolving and what they mean for the future of energy.

This internship offers you the opportunity to apply AI/ML methods to an established research portfolio on climate risk and resilience, including efforts to characterize current and projected trends in relevant weather and climate variables, identify potential climate impacts, and conduct quantitative climate risk assessments.

Climate AI Student Researcher

Remote/Home Based

Throughout The Internship, you will work closely with project leads to translate data‑driven findings into research outcomes and communications for end‑users. The successful candidate may work on 2–3 specific projects, including examples such as:

• Evaluation of Extreme Weather like lightning, wind gusts, and hail. Assess the ability of modern climate products and down scaled datasets to represent these extremes to help identify systematic biases and improve the reliability of risk assessments for critical utility infrastructure.
• Compound Hazard Emulation:
  Develop AI/ML surrogate models to simulate the interaction of multiple concurrent climate drivers, such as extreme heat and humidity. These emulators provide a computationally efficient alternative to traditional physics‑based simulations for long‑term planning.
• Advanced Return Level Analysis:
  Research robust statistical and machine learning methods to estimate the frequency and intensity of rare, high‑impact weather events. This work is critical for updating engineering standards and hardening the grid against low probability occurrences.
• Renewable Energy Forecast Validation:
  Assess the skill of AI‑driven wind and solar predictions against traditional numerical weather prediction (NWP) benchmarks. You will quantify the added value of these models in reducing uncertainty for energy resource adequacy studies.
• Natural Language Data Interface:
  Restructure Python modules to integrate Large Language Models (LLMs), allowing non‑technical end‑users of weather/climate data to query complex datasets using intuitive natural language. This project bridges the gap between high‑level climate science and actionable operational insights.

You will join a collaborative research environment and work side‑by‑side with established experts to tackle some of today’s most pressing societal challenges – energy and climate change – in collaboration with electric companies and government entities. We prioritize scientific rigor, data‑driven insights, and real‑world relevance. Our interns gain hands‑on experience contributing to major research initiatives and often publish journal articles.

The role is fully remote and there is some flexibility in defining the internship timeframe and work schedule. Preference will be given to candidates able to commit to a minimum target of a mostly full‑time schedule for 3–4 months duration. Potential for internship to begin before or extend beyond the typical summer period; in such cases, hours can be arranged around coursework or teaching loads, ranging from 15 hours (part‑time) to 40 hours (full‑time) per week.

• PhD candidate in meteorology, climate sciences, geography, earth sciences, environmental engineering, energy resources, computer science, or a related field; preference for current PhD students.
• Excellent Python skills
• Technical aptitude and experience with quantitative methods and data analyses.
• Familiarity with AI/ML methods and models, including deep learning, probabilistic modeling, and generative architectures.
• Experience collecting and analyzing weather and climate data (weather and climate model outputs, surface station observations, reanalysis products, satellite observations,…