Gabriele Bertoli Flood risk • AI & data assimilation • EO
Research Portfolio

Hydrological forecasting, flood risk, and AI-enabled data assimilation.

Researcher at the University of Florence and Visiting Researcher at ESA’s Φ-lab (SWOT). I work on uncertainty-aware flood forecasting, hybrid ML + variational DA (3D-Var/4D-Var), and emerging social-sensing workflows for rapid impact mapping.

Flood forecasting (discharge/peaks) 3D-Var / 4D-Var in latent space LSTM • XGBoost • Random Forest Convolutional Autoencoders Ecosystem services & exposure LLM-based social sensing
Selected papers Featured work Get in touch

About

PhD researcher (thesis submitted; viva pending) in Civil and Environmental Engineering (joint programme: University of Florence and TU Braunschweig) with a focus on flood risk, hydrological forecasting, and AI-enabled data assimilation. My work integrates machine learning with variational DA and representation learning to improve discharge forecasts across diverse catchments (e.g., EFAS and LamaH-CE).

Research profile (expanded)

I developed latent-space 3D-Var/4D-Var approaches using convolutional autoencoders and LSTM surrogates, and worked on environmental flood exposure (EnvXflood) combining ecosystem-service thinking with participatory valuation. I am also developing an LLM-based “social sensing” workflow to extract flood location and extent from social media, with multimodal extensions (image/video) for rapid impact mapping.

Research interests

Keywords

  • Flood forecasting; hydro-meteorological extremes; early warning systems
  • Flood and multi-hazard risk; uncertainty-aware modelling
  • Environmental exposure; ecosystem services; participatory valuation
  • Social sensing; multimodal flood mapping; LLM-based information extraction

Methods

  • Variational DA: 3D-Var / 4D-Var
  • Latent-space modelling: convolutional autoencoders
  • Deep learning: LSTM; hybrid ML (XGBoost/RF)
  • Probabilistic evaluation for extremes; explainable ML

Applications

  • EFAS; LamaH-CE; multi-scale discharge forecasting
  • Peak-flow forecasting and extremes evaluation
  • Flood exposure assessment (incl. environmental assets)
  • Rapid mapping from text/image/video observations

Selected publications

Loading latest publications…

Fetching the most recent public works from ORCID.

View full publication list on ORCID

Projects & research outputs

EnvXflood

Environmental flood exposure framework
  • Ecosystem-services framing + participatory valuation for environmental assets.
  • Case studies and qualitative exposure attributes for environmental typologies.
  • Open GIS dataset (Tuscany): doi:10.17632/55phb4xw5k.1

Latent-space DA for discharge forecasting

CAE + LSTM surrogates • 3D-Var/4D-Var
  • Representation learning to reduce DA computational cost.
  • Multi-catchment evaluation (e.g., EFAS / LamaH-CE style settings).
  • Focus on accuracy gains and scalable workflows.

LLM-based social sensing for floods

Information extraction • rapid mapping
  • Extract flood location/extent from social media text.
  • Planned multimodal extension (image/video) for impact mapping.
  • Designed for rapid, uncertainty-aware situational awareness.

CASTLE – WebGIS & open data (in preparation)

Project manager (WebGIS & Open Data)
  • Platform requirements, metadata specification, QA/QC and delivery validation.
  • Supports reservoir mapping and water-management decision workflows.
  • Project site: CASTLE

Experience

Visiting Researcher

ESA Φ-lab
  • Earth observation, SAR-based hydrometric measurements.
  • Generative AI for Earth System Science; SWOT-related work.

Researcher

University of Florence
  • AI methods + ecosystem services for sustainable water-resources management.
  • Hydraulic risk mitigation and decision-support workflows.

PhD (joint programme + visits)

University of Florence & TU Braunschweig & Imperial
  • Thesis: Flood Risk: An Interdisciplinary Approach Integrating Hydrology and Data Science.
  • ML + variational DA for discharge forecasting; EnvXflood framework.

Research Fellow

University of Florence
  • Flood risk of critical facilities (health, education).
Collaborations & service

Collaborations

  • PRIN CASTLE: small agricultural reservoirs mapping/quantification.
  • PRIMA AG-WaMED: non-conventional waters and participatory governance.
  • PRIMA NEXUS-NESS: WEFE Nexus decision-support and knowledge integration.

Academic service

  • Peer review: Nature Water; LNCS (Springer).
  • Organising: BIP “Green and Carbon-Neutral Cities” (EUniWell Arena 3), University of Florence.

Teaching & supervision

Workshops (University of Florence)

  • HEC-RAS modelling of Nature-Based Solutions (hands-on).
  • HEC-RAS & HEC-HMS hydraulic/hydrological modelling (hands-on).
  • Materials preparation, dataset setup, live demos, in-class support.

Supervision & outreach

  • Supervised 1 M.Sc. thesis and 1 B.Sc. thesis.
  • Public engagement on geo-hydrological risks (talks/workshops).
  • Seminar lecturer: flood risk of critical facilities (Politecnico di Milano).

Contact

For collaborations, speaking invitations, or project inquiries, email is the fastest channel.

✉️ gabriele.bertoli@unifi.it ORCID Google Scholar GitHub
Technical skills (high level)
  • AI/Data Science: Python, ML, explainable AI, data assimilation.
  • LLMs (local inference): Ollama; Docker (containerised deployments).
  • DevOps/Compute: Linux, server management, Git/GitHub, Colab.
  • Geospatial: GIS, WebGIS. Web/Markup: HTML, website administration.