This workshop is designed for:

• PhD students and early-career researchers
• Academics and research staff
• Environmental engineers and hydrologists
• Hydraulic modellers and water professionals
• Participants interested in applying Python, AI and machine learning to hydrology, hydraulics and environmental data 

No advanced programming experience is required. The workshop will be introductory and hands-on, with examples tailored to hydrology and hydraulics applications.

Artificial intelligence and machine learning are increasingly being used in environmental hydrology and hydraulics to support data analysis, prediction, uncertainty assessment, model emulation, flood-risk assessment, climate-impact analysis, water-quality modelling and decision support. However, many researchers and practitioners need accessible, practical training to understand how to apply these methods responsibly and effectively.

This full-day workshop will introduce participants to the foundations of Python programming and AI/ML modelling using practical examples from hydrology and hydraulics. The session will combine short lectures, guided coding exercises, research-based case studies and discussion of good practice in applying AI to environmental systems.

The workshop will use real-world environmental datasets and examples drawn from hydrology, hydraulics, flood modelling, rainfall-runoff analysis, water-level prediction, climate extremes and environmental monitoring. Participants will gain practical experience in preparing data, visualising hydrological variables, building simple machine-learning models, evaluating model performance and interpreting results.

AI in Environmental Hydrology and Hydraulics is a practical, hands-on workshop that introduces Python programming and machine learning methods for water and environmental applications. Designed for PhD students, early career researchers, academics and professionals, the workshop will use real-world hydrology and hydraulics examples to demonstrate how AI can support data analysis, prediction, model evaluation and decision-making. Participants will gain introductory coding experience, explore relevant research use cases, and develop a practical understanding of how AI can complement traditional hydrological and hydraulic modelling. A certificate of participation will be provided.

By the end of the workshop, participants will be able to:

1. Understand the role of AI and machine learning in environmental hydrology and hydraulics.
2. Use basic Python tools for importing data from various sources, pre-processing, exploratory data analysis, analysing and visualising environmental datasets.
3. Apply introductory machine-learning methods/models to hydrological and hydraulic problems.
4. Evaluate model performance using appropriate metrics.
5. Interpret AI/ML outputs in the context of environmental modelling and decision-making.
6. Understand opportunities and limitations of AI in water, climate and flood-risk applications.
7. Identify how AI methods can complement process-based hydrological and hydraulic models.

Dr. Sandhya Patidar is an associate professor in Data Science at Heriot-Watt University, UK, with expertise in applied mathematics, statistics, machine learning, uncertainty quantification and environmental modelling. Her interdisciplinary research focuses on the development and application of data-driven and AI-based methods for complex environmental, climate, water and energy systems. She has extensive experience in hydrological and hydraulic modelling, flood-risk analysis, climate-impact assessment, water-resource systems, energy demand modelling, and decision-support analytics.

Dr. Patidar has published widely in applied data science, environmental risk, climate resilience and infrastructure systems, and has contributed to a range of UK and European research projects involving AI, digital twins, energy communities and environmental decision support. Her work combines statistical modelling, machine learning, time-series analysis, geospatial data, uncertainty analysis and stakeholder-focused research. She is particularly interested in trustworthy AI, explainable machine learning, hybrid modelling, and the use of AI to support sustainable and resilient environmental infrastructure.

In this workshop, Dr. Patidar will introduce practical Python and AI/ML methods for hydrology and hydraulics, drawing on real-world research examples and hands-on applications.

LinkedIn     Website

Session focus: Welcome, workshop aims, and introductions to participants.

Content:

• Introduction to the workshop structure
• Overview of AI in environmental hydrology and hydraulics
• Participant expectations and background
• Brief introduction to Python-based practical environment

Session focus: Understanding the role of AI in water and environmental systems.

Topics covered:

• What AI and machine learning mean in environmental modelling
• Difference between process-based models, statistical models and data-driven models
• Examples of AI applications in hydrology and hydraulics
• Flood prediction, rainfall-runoff modelling, water-level forecasting, climate extremes and hydraulic model emulation
• Opportunities, limitations and responsible use of AI 

Example discussion: 

How AI can support flood-risk modelling, water-resource planning, climate adaptation and environmental decision support.

Session focus: Introductory Python skills for handling hydrological and hydraulic datasets.

Topics covered:

• Working with Jupyter Notebook or Google Colab
• Introduction to Python libraries: pandas, numpy, matplotlib, scikit-learn
• Importing CSV and time-series data
• Basic data inspection and cleaning
• Handling missing values and outliers 

Hands-on exercise:
Loading and exploring a hydrological dataset, such as rainfall, river flow, water level or climate variables.

Session focus: Understanding environmental data before modelling.

Topics covered:

• Time-series visualisation
• Rainfall, river flow and water-level plots
• Scatter plots, histograms and correlation analysis
• Seasonal and event-based patterns
• Identifying trends, extremes and anomalies 

Hands-on exercise:
Participants will visualise environmental time-series data and identify patterns relevant to hydrological or hydraulic interpretation.

Expected skills gained:
Participants will learn how to use Python to create simple yet informative plots from environmental datasets.

Session focus: Building simple predictive models.

Topics covered:

• Supervised learning concepts
• Training and testing datasets
• Input features and target variables
• Regression models for environmental prediction
• Model evaluation using RMSE, MAE, R² and error plots
• Overfitting and generalisation 

Hands-on exercise:
Building a basic machine-learning model to predict river flow, water level or runoff using rainfall and/or climate variables.

Session focus: Understanding whether an AI model is useful and trustworthy.

Topics covered:

• Model performance metrics
• Residual analysis
• Comparing observed and predicted values
• Interpreting model errors during extreme events
• Basic explainability ideas
• Limitations of black-box models in environmental decision-making 

Hands-on exercise:
Participants will evaluate a trained model and discuss whether the result is scientifically meaningful.

Session focus: Research-informed applications and case studies.

Indicative use cases:

• Rainfall-runoff modelling
• River flow and water-level forecasting
• Flood susceptibility and flood-risk mapping
• Hydraulic model emulation
• Climate extremes and future risk assessment
• AI for environmental monitoring and anomaly detection
• Hybrid modelling: combining physical models and machine learning 

Discussion:
Where AI adds value, where physical modelling remains essential, and how hybrid approaches can improve reliability.

Session focus: Opportunities, challenges and research frontiers.

Topics covered:

• Data quality and uncertainty
• Bias in environmental datasets
• Transferability across catchments and regions
• AI and climate-change impact modelling
• Physics-informed machine learning
• Data assimilation and digital twins
• AI emulators for fast hydraulic and hydrological simulations
• Reproducible and open research workflows 

Discussion:
How AI can support future environmental research, professional practice and decision-making.

Session focus: Summary of key learning points and discussion on applying the learning to participants’ own research or professional interests.

Activity:
Participants will reflect on a hydrology or hydraulics problem from their own work and identify:

• What data would be needed
• What AI/ML method could be suitable
• What model outputs would be useful
• What risks or limitations should be considered
• How the results could support decision-making
• Questions and answers
• Certificate of participation 

Participants will receive a certificate of participation, which may be positioned as CPD-equivalent training, subject to the conference organiser’s approval.

• A laptop
• Access to Google Colab or Anaconda/Jupyter Notebook
• Basic familiarity with Excel spreadsheets or data files (*.txt)
• No advanced coding experience required.

• Python notebooks
• Sample environmental datasets
• Step-by-step practical exercises
• Reading list and further resources
• Slides and example code 

The short course/workshop will be delivered using a practical, accessible digital infrastructure that supports hands-on learning and active participation. Participants will be guided through Python-based exercises using platforms such as Google Colab or Jupyter Notebook, allowing them to run code, explore datasets, visualise results and build introductory AI/ML models without needing advanced local software installation. The training materials will include prepared notebooks, sample hydrology and hydraulics datasets, step-by-step coding examples, slides and supporting reading resources. Where possible, open-source tools and reproducible workflows will be used to encourage participants to continue applying the methods after the workshop. A suitable teaching room with reliable internet access, projection facilities, and space for laptop-based work will be provided to the participants, along with basic technical support during the session. This infrastructure will ensure that the workshop is interactive, facilitates online delivery, is inclusive, and is suitable for participants with varied levels of programming experience.