Focus areas
Four fronts of flood resilience
VindaFlow bridges computational hydraulics and modern machine learning — turning research into operational tools for the people who need them before the water arrives.
AI Flood Forecasting
Machine-learning and hybrid physics–AI models for river flood prediction — rainfall–runoff surrogates, LSTM/transformer forecasting, and uncertainty quantification.
Dam-Break Early Warning
Real-time early-warning pipelines for dam failure scenarios: breach modelling, surge propagation, inundation mapping, and alert dissemination.
Cyclone Floods in MENA→
Flash-flood response of arid wadis to tropical cyclones and extreme rainfall across the MENA region — from Shaheen to Mediterranean lows.
River Hydraulics & Engineering→
1D/2D hydrodynamic modelling, sediment transport, river training works, and numerical methods for open-channel flow.
Projects & tools
Working tools, not just papers
Dam-Break Standards Hub
A reference website bringing together international standards, guidelines and best practices for dam-break analysis — breach parameters, flood routing, and hazard classification — in one navigable place.
River Morphology Atlas
An interactive multi-method web app for river morphology prediction and classification — the first of its kind to run several state-of-the-art methods side by side, with validity screening, an agreement dashboard, and full provenance for every number.
Oman Cyclone Forecast — Local Interpretation
An interactive WebGIS replaying Cyclone Shaheen (2021) over Oman — AI forecast scenarios against what actually happened, sub-basin impact scores, and per-wilayat rainfall convergence.
Flood a Real River — 3D
Raise the water over real photogrammetry terrain and watch the inundation spread — a wetland river reach and the Colorado River in the Grand Canyon, with an adjustable flood-water surface, right in the browser.
Reach-Scale Erosion Depth Calculator
A planned tool for computing erosion (scour) depth across different river reaches — comparing established scour equations per reach and flagging where their validity limits are exceeded.
Research & publications
Recent work
Sample: Deep-learning rainfall–runoff forecasting for arid catchments
Placeholder entry — replace with your real publication. Example of an LSTM-based flood forecasting study for wadi systems with sparse gauging.
working-paperSample: Neural surrogate models for real-time dam-break inundation mapping
Placeholder entry — replace with your real publication. Example of a surrogate model that emulates 2D shallow-water simulations in milliseconds.
working-paperSample: Hydrological response of Omani wadis to tropical cyclones
Placeholder entry — replace with your real publication. Example of a study on cyclone-driven flash floods in the Arabian Peninsula.
Open proposals
Demo-stage projects seeking sponsors
These systems already exist as working demos. With the right partner — a government agency, NGO, or research fund — they can become operational services protecting real communities.
Sample: AI early-warning system for wadi flash floods
Placeholder proposal — replace with your real project. A working demo of an end-to-end flash-flood early-warning pipeline for arid catchments: nowcasting rainfall, AI runoff forecasting, and automated public alerts.
Sample: Dam-break digital twin for reservoir safety
Placeholder proposal — replace with your real project. A demo-stage digital twin that couples live reservoir telemetry with pre-computed breach scenarios to give operators instant inundation forecasts.
Notes
Latest from the blog
Welcome to VindaFlow
Why this site exists: sharing research and tools at the intersection of AI, river hydraulics, and flood early warning.
Sample: The Ritter solution — the "hello world" of dam-break hydraulics
Placeholder post — replace with your own writing. A walkthrough of the classic analytical solution for instantaneous dam failure on a frictionless bed.