Research

AI for Earth System Modeling

Numerical weather and climate models simulate the physics governing the Earth system to generate real-time weather forecasts and long-term climate projections. Since the mid-20th century, numerical weather prediction has been the foundation of weather forecasting, providing enormous societal benefits. Climate models have also been instrumental in understanding the impact of anthropogenic emissions on our current and future climate and informed mitigation and adaptation strategies. Although the accuracy of weather and climate models has improved in recent decades, they have become increasingly complex and computationally expensive. Advances in machine learning techniques, the availability of high-performance computing, and the emergence of large high-dimensional meteorological datasets have made deep learning-based weather and climate models increasingly attractive. My research aims to leverage explainable, physics-informed deep learning techniques to improve existing weather and climate models.

Relevant Publications

  • Nathaniel Cresswell-Clay, Bowen Liu, Dale Durran, Andy Liu, Zachary I. Espinosa, Raul Moreno, Matthias Karlbauer "A Deep Learning Earth System Model for Stable and Efficient Simulation of the Current Climate" arXiv physics.ao-ph https://doi.org/10.48550/arXiv.2409.16247

  • Zachary I. Espinosa, Sheshadri, A., Cain, G. R., Gerber, E. P., & DallaSanta, K. J. (2022). "Machine learning gravity wave parameterization generalizes to capture the QBO and response to increased CO2." Geophysical Research Letters, 49, e2022GL098174. https://doi.org/10.1029/2022GL098174 [press highlight]
  • Polar Climate

    The Arctic and Antarctic are a fundamental part of the climate system, and short-term and long-term changes in their climate can influence Earth's global radiative budget, atmospheric and oceanic circulation, meridional heat transport and extreme weather in the midlatitudes. I am broadly interested in sea ice variability and predictability and better understanding the coupled interactions between sea ice the atmosphere and the ocean.

    Relevant Publications

  • Zachary I. Espinosa, Edward Blanchard-Wrigglesworth, and Cecilia Bitz. "Understanding the drivers and predictability of record low Antarctic sea ice in austral winter 2023."Nature Communications Earth & Environment, https://doi.org/10.1038/s43247-024-01772-2
  • Blanchard‐Wrigglesworth, E., Cox, T., Espinosa, Z. I., & Donohoe, A. (2023). "The largest ever recorded heatwave—Characteristics and attribution of the Antarctic heatwave of March 2022." Geophysical Research Letters, 50(17), e2023GL104910. https://doi.org/10.1029/2023GL104910 [Washington Post]
  • David P. Schneider, Ziqi Yin, Gemma K. O'Connor, Edward Blanchard-Wrigglesworth, Zaria I. Cast, Rajashree Datta, Zachary I. Espinosa (2024). "Increasing Antarctic snowfall mitigates sea level rise less than projected due to meltwater influence on sea surface temperatures." Authorea Preprints (2024).https://essopenarchive.org/doi/full/10.22541/essoar.172411232.25724214/v1
  • Zachary I. Espinosa, Lettie Roach, Cecilia M. Bitz, and Dirk Notz. "Sea Ice in Earth system models". Sea Ice 4 (2025): [in prep; chapter 10 of textbook]
  • Climate Dynamics and Climate Change

    Global climate is controlled by complex interactions between the atmosphere, ocean, crysophere, and land. Disentangling the relative contribution of anthropogenic warming and multidecadal internal variability is challenging due to this complex coupling, limited observational data, incorrect climate models. My research aims to better understand the physics controlling recent and projected climate change. My recent work examines the impact of cloud feedbacks on multidecadal internal variability in the tropical Eastern Pacific.

    Relevant Publications

  • Zachary I. Espinosa and Mark D. Zelinka. "The Shortwave Cloud-SST Feedback Amplifies Multi-Decadal Pacific Sea Surface Temperature Trends: Implications for Observed Cooling." Geophysical Research Letters, 51(18), e2024GL111039, https://doi.org/10.1029/2024GL111039

  • Education

  • PhD, Atmospheric and Climate Sciences | University of Washington | 2021 - 2025 (expected)
  • MS, Applied and Engineering Physics | Stanford University | 2019 - 2021
  • BS, Computer Science (AI specialization) | Stanford University | 2015 - 2020

  • Professional Experience

  • Research Intern | Livermore, CA | Lawrence Livermore National Laboratory | June 2023 – Sep 2023
  •   - Studied the impact of marine boundary layer clouds on historical East Pacific Ocean cooling

  • Research Intern | Richland, WA | Pacific Northwest National Laboratory | June 2021 – Sep 2021
  •   - Studied the impact of climate change on annual precipitation in the Amazon Rainforest

  • Graduate Research Assistant | Stanford, CA | Stanford Earth Systems Science | Sep 2019 – Sep 2021
  •   - Developed a machine learning parameterization of gravity wave in a global climate model (Sheshadri Group)

      - Publication in Geophysical Research Letters - Espinosa, Zachary I., et al (2022)

  • Machine Learning Engineering Intern | Redwood City, CA | UnifyID | Apr 2020 – Jun 2020
  •   - Developed in-house machine learning pipeline for research & development. Introduced pipeline testing

  • Quantum Engineering Intern | Palo Alto, CA | AT&T Foundry | Jun 2019 – Sep 2019
  •   - Built an open-source python framework for quantum networking (QN) simulations called netQuil, designed to support the implementation of canonical QN protocol

  • Software Engineering Intern | Mountain View, CA | Smartcar, Inc. | Jan 2019 – Jun 2019
  •   - Designed, built, and launched electric vehicle endpoints for Smartcar API

      - Maintained python, node.js, and java SDKs. Contributed to OAuth2 pipeline.

  • Mobile Software Engineering Intern | San Francisco, CA | OXO, Inc. | Apr 2018 – Sep 2018
  •   - Built first iteration MVP mobile app for iOS and Android using React Native, Firebase, Heroku, and AWS RDS.

  • Web and Networking Engineering Intern | Ashton, ID | Henry’s Fork Foundation | Jun 2017 – Sep 2017

  • Awards & Fellowships

  • Department of Energy Computational Science Graduate Fellowship (DOE CSGF) | Apr 2022
  • Graduate Student Equity & Excellence Fellowship (GSEE Fellow) | Sep 2021
  • Achievement Rewards for College Scientists Foundation Scholar (ARCS Scholar) | Sep 2021
  • The GEM National Consortium Graduate Fellow (GEM Graduate Fellow) | Jan 2020

  • Teaching Experience

  • Instructor & Mentor | AI Fellowship Program | VeritasAI | Jun 2024 - Present
  • Guest Lecturer | ATMS 220: Exploring the Atmospheric Sciences | University of Washington | Oct 2023
  • Guest Lecturer | ATMS 220: Exploring the Atmospheric Sciences | University of Washington | May 2023
  • Teaching Assistant | ATMS 101: Weather | University of Washington | Jan 2023 - Mar 2023

  • Talks & Presentations

  • Poster | CFMIP 2024 | The Impact of the Shortwave Cloud Feedback on East Pacific Multi-Decadal Variability |Jun 2024
  • Talk | University of Washington Climate Dynamics Seminar | From Record Low Sea Ice to East Pacific Cooling: Unraveling Southern Hemisphere Extremes| Apr 2024
  • Poster | US CLIVAR Blocking and Extreme Weather Workshop | The Physics of Summertime Antarctic Heatwaves| Mar 2024
  • Poster | American Geophysical Union Fall Meeting | The Physics of Summertime Antarctic Heatwaves| Dec 2023
  • Poster | Graduate Climate Conference | Drivers of the Unprecedented Low Antarctic Sea Ice Extent during Austral Winter 2023| Oct 2023
  • Poster | DOE CSGF Annual Review | The Physics of Summertime Antarctic Heatwaves | Jul 2023
  • Talk | Scientific Committee on Antarctic Research | The Physics of Summertime Antarctic Heatwaves|Jun 2023
  • Talk | University of Washington Climate Dynamics Seminar | Machine Learning Gravity Wave Parameterization Generalizes to Capture the QBO and Response to Increased CO2 | Feb 2023
  • Talk | Graduate Climate Conference | Drivers of Interannual Variability of Summer Sea Ice Extent | Oct 2022
  • Poster | BEPSII Arctic Field School | Drivers of Interannual Variability of Summer Sea Ice Extent | May 2022
  • Talk | AGU Fall Meeting | Machine Learning Emulation of Parameterized Gravity Wave Momentum | Dec 2021
  • Talk | EGU General Assembly | Machine Learning Emulation of Parameterized Gravity Wave Momentum | Apr 2021
  • Poster | AGU Fall meeting | A Data-Drive, Single column Gravity Wave Parameterization in an Idealized Model | Dec 2020
  • Talk | MSCAR | A Data-Drive, Single column Gravity Wave Parameterization in an Idealized Model | Sep 2020
  • Talk | CalGFD | A Data-Drive, Single column Gravity Wave Parameterization in an Idealized Model | Aug 2020
  • Poster | APS March Meeting (Canceled) | netQuil: A playground for quantum networking simulations | Mar 2020
  • Poster | Stanford Deep Learning Poster Session | Distracted Driver Detection | Jun 2018
  • Poster | Stanford Artificial Intelligence Post Session | Tracking Schistosomiasis with Computer Vision | Mar 2018

  • Service & Leadership

  • Graduate Student Representative | Program on Climate Change Graduate Student Steering Committee | Sep 2023 - Sep 2024
  • Graduate President | UW American Meterological Society Chapter | Sep 2021 - Sep 2023

  • Projects

    Drivers of the Seasonal Delay of Rainfall in the Amazon Rainforest

    The findings of this work were presented at the Pacific Northwest National Laboratory (PNNL) August 2021 Intern Symposium

    Tropical precipitation has a distinct annual cycle characterized by an amplitude, the range between wet and dry seasons, and phase, their onset timing. Previous studies have reported a seasonal delay in the onset of precipitation in observations over the northern tropical land driven by changes in greenhouse gases (GHG) and anthropogenic aerosols (AER). Though it is well known that the phase of the annual precipitation cycle is projected to be delayed over land under global warming, it is unclear whether changes in GHG and AERs are the dominate external forcings for different regions. Due to its outsized importance in the global Earth system, changes in the precipitation cycle in the Amazon are of particular interest. Here, we use multi-model output of historical and individual forcing simulations to show that the seasonal delay of precipitation in the Amazon rainforest cannot be fully explained by changes in GHG and AER. We examine the impact of land use and land cover change (LULCC; e.g. deforestation) on the seasonal delay of rainfall by examining the historical and future simulations without changes in LULCC, and we perform an atmospheric energy budget analysis for the Amazon Basin.

    This work was done at PNNL in collaboration with Dr. L. Ruby Leung and Dr. Fengfei Song.


    NetQuil: A Quantum Playground for Distributed Quantum Computing Simulations

  • White Paper
  • Pypi Installation
  • Doc Center
  • GitHub Source Code
  • NetQuil is an open-source Python framework for quantum networking simulations built on the quantum computing framework pyQuil, by Rigetti Computing. NetQuil is built for testing ideas in quantum network topology and distributed quantum protocol. It allows users to create multi-agent networks, connect parties through classical and quantum channels, and introduce noise. NetQuil also makes running multiple trials for non-deterministic experiments, reviewing traffic in real-time, and synchronizing agents based on local and master clocks simple and easy. We provide an overview of the state of distributed quantum protocol and a basic introduction to netQuil's framework. Finally, we present several demonstrations of canonical quantum information protocols built using netQuil's distributed quantum gates and pyQuil.

    This project was completed with Matthew Radzihovsky and Yewon Gim at AT&T Foundry, Palo Alto.


    Henry's Fork Foundation Water Quality Monitoring Network

    Website

    While at the Henry's Fork Foundation (HFF), I helped build an automated data-collection network of sensors in the Henry’s Fork Water Shed near Yellowstone National Park. To accompany this data-collection network, I then led an effort to build and published a real-time water quality monitoring website. Since 2018, the website has been frequently used by local scientists to study the health of the Henry's Fork hydrological system and has served as the cornerstone of HFF’s scientific communication mission.

    This project was completed with Melissa Muradian and Justin Appleby.