Lectures

[Jan 25]  Lecture 1:  Course Intro

Lecture Slides:  Course Intro: Tackling Climate Change with Machine Learning  [Brown login required]

Readings (Optional)

• McGovern & Allen (2021). Training the Next Generation of Physical Data Scientists. Eos. doi:10.1029/2021EO210536. 

• Karpatne, et al. (2018). Machine learning for the geosciences: Challenges and opportunities. IEEE Transactions on Knowledge and Data Engineering.  https://arxiv.org/abs/1711.04708.

• Reichstein, et al. (2019). Deep learning and process understanding for data-driven Earth system science. Nature. https://doi.org/10.1038/s41586-019-0912-1.  [available online via Brown library].

• Bortnik & Camporeale (2021). Ten ways to apply machine learning in Earth and space sciences. Eos. https://doi.org/10.1029/2021EO160257.  

• McGovern, et al. (2022). Why we need to focus on developing ethical, responsible, and trustworthy artificial intelligence approaches for environmental science. Environmental Data Science. doi:10.1017/eds.2022.5.

Resources

• Sign up for the Climate Change AI newsletter here.

[Jan 30]  Lecture 2: Intro to Climate (Part 1) 

Lecture Slides

• Climate 101: The Climate System and Evidence for Climate Change

Resources for Climate Information

[Feb 1]  Lecture 3: Intro to Climate (Part 2) 

Lecture Slides - lecture given by grad TA John Nicklas 

• Climate 101: Greenhouse Gases and Climate Models

Resources

• Global Climate Change: Vital Signs of the Planet (NASA)

• [Animation] Molecules Vibrate (UCAR Center for Science Education)

• Greenhouse Gas (GHG) Emissions and Removals (EPA)

• Explainer: Climate Models (MIT Climate Portal)

• Q&A: Why clouds are still 'one of the biggest uncertainties' in climate change (Horizon: The EU Research & Innovation Magazine, Nov 2020)

• Climate Grand Challenges (World Climate Research Programme)

[March 5]  Lecture 11:  Natural Hazards & eXplainable AI (XAI) Part 1

Lecture Slides: Natural Hazards and Extreme Weather

Reading

• Kuglitsch et al. (2022). AI for Disaster Risk Reduction: Opportunities, challenges and prospects. World Meteorological Organization (WMO) Bulletin. 

Resources

• The human cost of disasters: an overview of the last 20 years (2000-2019). UN Office for Disaster Risk Reduction (2020).

• WMO Atlas of Mortality and Economic Losses from Weather, Climate and Water Extremes (1970-2019). World Meteorological Organization (2021). 

• U.S. Billion-Dollar Weather and Climate Disasters. NOAA National Centers for Environmental Information (2023). [dataset] 

• More dangerous heat waves are on the way: See the Impact by Zip code. Washington Post. August 15, 2022. 

• The National Risk Index. FEMA. 

• EM-DAT International Disaster Database

• Crowdsourced Earthquake Early Warning & Damage Detection examples:

  1. 1. • Kong et al. (2016). MyShake: A Smartphone Seismic Network for Earthquake Early Warning and Beyond. Science Advances.

        • Chachra et al. (2022). Detecting damaged buildings using real-time crowdsourced images and transfer learning. Scientific Reports.

[March 7]  Lecture 12:  Student-led discussion #2

Slides:  FourCastNet & Family (ML Weather Emulators) by Caleb & John  

Resources

• Coding Activity: Explore and plot ECMWF open dataset 

• Video explainer & article: Can AI Help Us Predict Extreme Weather? Vox. (Feb 2024). [YouTube short]

• Guibas et al. (2022). Adaptive Fourier Neural Operators. ICLR. 

• Fractal Hyperparameters: [arXiv:2402.06184] [video]

Readings  (see Canvas for details)

• Overview (blog post): The Rise of ML in Weather Forecasting. ECMWF Science Blog. 2023.

• FourCastNet (NVIDIA)

• • • Blog post: Modeling Earth's Atmosphere with Spherical Fourier Neural Operators. NVIDIA Developers Technical Blog. 2023. 

    • Paper: Pathak et al. (2022). FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators. https://arxiv.org/abs/2202.11214. 

  • GraphCast (Deep Mind) 

    • Blog post:  GraphCast: AI model for faster and more accurate global weather forecasting. Deep Mind Research Blog. 2023.

    • Paper:  Lam et al. (2023). Learning skillful medium-range global weather forecasting. Science. https://doi.org/10.1126/science.adi2336

  • Pangu-Weather (Huawei Cloud) 

    • Perspective: Ebert-Uphoff & Hilburn (2023). The outlook for AI weather prediction. Nature. https://doi.org/10.1038/d41586-023-02084-9.

    • Paper: Bi et al. (2023). Accurate medium-range global weather forecasting with 3D neural networks. Nature. https://doi.org/10.1038/s41586-023-06185-3.

[March 12]  Lecture 13:  Natural Hazards & eXplainable AI (XAI) Part 2

Lecture Slides: Explainable AI (XAI)

Reading

• McGovern et al. (2019).  Making the Black Box More Transparent: Understanding the Physical Implications of Machine Learning.  Bulletin of the American Meteorological Society. 

Resources

• Introductory book (available free online):  Interpretable Machine Learning: A Guide for Making Black Box Models Explainable by Christoph Molnar

• Flora et al. (2024). A Machine Learning Explainability Tutorial for Atmospheric Sciences. AI for Earth Systems. 

• A critical perspective on XAI:  Rudin (2019).  Stop Explaining Black Box ML Models for High Stakes Decisions and Use Interpretable Models Instead. 

• Interpreting linear models: Common pitfalls in the interpretation of coefficients of linear models.  Scikit-learn documentation.

• Symbolic Regression: Udrescu & Tegmark (2020). AI Feynman: A physics-inspired method for symbolic regression. Science Advances.

• LIME paper: Ribeiro et al. (2016). "Why Should I Trust You?": Explaining the Predictions of Any Classifier. KDD. 

• SHAP paper: Lundberg & Lee (2017). A Unified Approach to Interpreting Model Predictions. NeurIPS. 

[March 14]  Lecture 14:  Natural Hazards & eXplainable AI (XAI) Part 3

Lecture Slides:  XAI for Weather and Climate

Reading

• Toms et. al. (2020). Physically Interpretable Neural Networks for the Geosciences: Applications to Earth System Variability.  JAMES. 

• [optional]  Hilburn et al. (2021). Development and Interpretation of a Neural Network-Based Synthetic Radar Reflectivity Estimator Using GOES-R Satellite Observations. JAMC. 

Resources

• Adebayo et al. (2018). Sanity Checks for Saliency Maps. NeurIPS.

• Lapuschkin et al. (2019). Unmasking Clever Hans predictors and assessing what machines really learn. Nature Communications. 

• Mamalakis et al. (2022). Neural network attribution methods for problems in geoscience: A novel synthetic benchmark dataset. Environmental Data Science. 

• Olah et al. (2017). Feature Visualization: How neural networks build up their understanding of images.

• Montavon et al. (2019). Layerwise Relevance Propagation: An Overview.  Lecture Notes in Computer Science. 

• Interactive Demo on Layer-wise Relevance Propagation (LRP)

• Barnes et al. (2020). Indicator Patterns of Forced Change Learned by an Artificial Neural Network. JAMES.

[March 19]  Lecture 15:  Student-led paper discussion #3

Slides: Flood Forecasting by Brad & Michael

Readings

• Nevo et al. (2022). Flood forecasting with machine learning methods in operational framework. Hydrology and Earth Systems Sciences. 

• Nearing et al. (2023). AI Increases Global Access to Reliable Flood Forecasts. https://arxiv.org/abs/2307.16104

Resources

•  Flood Forecasting Project @ Google Research

  • Flood Hub

  • Bloomberg article (sponsored):  How AI is helping communities anticipate floods.

  • Google Blog: Our support for early warning systems. (June 2023)

  • Google Blog: Helping more people stay safe with flood forecasting. (May 2023)

  • Google Flood Hub Presentation. [YouTube, 12 minutes]

[April 2]  Lecture 17: Farms & Forests and Earth Observation 

Lecture Slides:  Agriculture & Forest Ecosystems

Readings

• Rolnick et al. (2022). Tackling Climate Change with ML - Section 6: Farms & Forests.  ACM Computing Surveys. [pdf - Section 6 only]

• [optional]  Nakalembe & Kerner (2023). Considerations for AI-EO for agriculture in Sub-Saharan Africa. Environmental Research Letters.

Resources

• IPCC, 2022. Climate Change 2022: Mitigation of Climate Change. Chapter 7: Agriculture, Forestry, and Other Land Uses. 

• IPCC, 2019. Climate Change and Land. 

[April 4]  Lecture 18: Learning with Limited Labels

Lecture Slides:  Learning with Limited Labels

Readings

• Rolf et al. (2024).  Satellite Data is a Distinct Modality in Machine Learning. arXiv preprint.

• Xie et al. (2015). Transfer Learning from Deep Features for Remote Sensing and Poverty Mapping. AAAI. 

Resources

• Safonova et al. (2023). Ten deep learning techniques to address small data problems with remote sensing. International Journal of Applied Earth Observation and Geoinformation.

• Ma et al. (2024). Transfer learning in environmental remote sensing. Remote Sensing of the Environment. 

• Pan & Yang (2009).  A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering. [link to author pdf]

• Finn et al. (2017). Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks. ICML. 

• Wang et al. (2019). Generalizing from a Few Examples: A Survey on Few-Shot Learning. ACM Computing Surveys.

• Snell et al. (2017). Prototypical Networks for Few-Shot Learning. NeurIPS.

• Van Engelen & Hoos (2020). A survey on semi-supervised learning. Machine Learning. 

[April 9]  Lecture 19:  Farms & Forests Case Studies

Slides

Readings (see Canvas for assigned case study)

• Case Study A: Early et al. (2022). Scene-to-Patch Earth Observation: Multiple Instance Learning for Land Cover Classifivation. NeurIPS Tackling Climate Change with ML Workshop.  

• Case Study B:  Narayanan et al. (2022). Curator: Creating Large-Scale Curated Labeling Data Sets Using Self-Supervised Learning. AAAI Fall Symposium.

• Case Study C: Tseng et al. (2023). Lightweight, Pretrained Transformers for Remote Sensing Time Series. NeurIPS Tackling Climate Change with ML Workshop.  

[April 11]  Lecture 20: Student-led paper discussion #4

Slides: Foundation Models for Geospatial AI by Ayushman & Julian

Reading (see Canvas for reading assignment)

• Jakubik et al. (2023). Foundation Models for Generalist Geospatial Artificial Intelligence. https://arxiv.org/abs/2310.18660.

• Blog post: NASA and IBM Openly Release Geospatial AI Foundation Model for NASA Earth Observation Data. NASA Earth Data News. (August 2023). 

• Essay: R. Ramachandran (2023). From Petabytes to Insights: Tackling Earth Science's Scaling Problem.  AGU Leptouhk Lecture Essay. [Sections: The Unchanging Problem of Scale and AI - A New Inflection Point]

Resources

• Prithvi model page on Hugging Face.

[April 18]  Lecture 22:  Ocean Ecosystems 

Lecture Slides - lecture given by grad TA John Nicklas 

• Oceans, Marine Ecosystems & Biodiversity

Resources

• UNESCO Intergovernmental Oceanographic Commission. State of the Ocean Report 2022. 

• IPCC Sixth Assessment Report. Chapter 3: Oceans and Coastal Ecosystems and their Services.

• Ocean Visions (Prof. Emanuele Di Lorenzo, Chairman and Co-Founder)

• NOAA Coral Reef Watch.

• The Widest-Ever Global Coral Crisis Will Hit Within Weeks, Scientists Say. NYTimes (April, 2024).

• Ocean Heat Has Shattered Records for More Than a Year. What's Happening? NYTimes (April, 2024).

• J. Swiezewski (2022). Counting Nests of Shags with YOLO to Assess the Wellbeing of the Antarctic Ecosystem. Appsilon Blog. 

[April 23]  Lecture 23: Guest Speaker - Dr. Kellenberger (Yale)

Slides: Deep Learning with Few Labels for Environmental Applications by Benjamin Kellenberger

Readings (see Canvas for details)

• Kellenberger et al. (2019). Half a Percent of Labels is Enough: Efficient Animal Detection in UAV Imagery Using Deep CNNs and Active Learning. IEEE Transactions on Geosciences and Remote Sensing. 

• Rußwurm et al. (2023). Meta-learning to address diverse Earth Observation problems across resolutions. Nature Communications Earth & Environment. 

• Tollenaar et al. (2024). Where the Continent is Blue: Deep Learning Locates Bare Ice in Antarctica. Geophysical Research Letters.

• Zbinden et al. (2024). On the selection and effectiveness of pseudo-absences for species distribution modeling with deep learning. https://arxiv.org/abs/2401.02989

Resources

• Benjamin Kellenberger's website and Google Scholar

• Annotation Interface for Data-Driven Ecology (AIDE): https://github.com/microsoft/aerial_wildlife_detection/tree/v3.0