Extreme event simulation & prediction
Simulation and prediction of extreme hydro-meteorological events using wavelet-based spectral transformation and machine learning.
Hydroclimate Extremes · Modelling · Forecasting
Hydroclimate science for a resilient future.
We advance the prediction of extreme hydrometeorological events — floods, droughts and compound hazards — by developing wavelet-based spectral transformation methods, multi-model ensemble frameworks, and physics-aware machine learning.
Hydroclimate ExtremesFrom flood to drought — across scales
01 / About
Hydroclimate Extremes Modelling & Forecasting.
HydroclimateX Lab develops advanced methods for predicting extreme hydrometeorological events — floods, droughts, and compound hazards — across timescales from seasonal to decadal. Our work integrates wavelet-based spectral transformation, multi-model ensemble frameworks, and physics-aware machine learning to bridge cutting-edge research and operational water management.
We maintain the open-source WASP (WAvelet System Prediction) toolkit, available in R, Python and MATLAB, along with companion tools for quantile mapping (WQM), predictor identification (NPRED), and synthetic data generation.
Explore our open-source tools ↗02 / Research interests
Five connected research directions.
Multi-agent & human–water systems
Multi-agent models capturing feedbacks between reservoir operations, agricultural water use, and climate variability.
Landslide hazard prediction & risk
Rainfall-induced landslide hazard prediction integrating satellite precipitation, soil moisture, and terrain susceptibility.
Hydrological–ecological modelling
Hydrological-hydrodynamic, hydro-ecological, and water quality process modelling for integrated watershed management.
Climate model bias correction
Post-processing and bias correction of climate and weather forecast models using frequency-domain quantile mapping (WQM).
Research approach
From spectral decomposition to operational decisions.
- 01 Observe Multi-source hydro-climate observations · reanalysis · GCM outputs
- 02 Decompose Wavelet spectral transformation to isolate predictive frequency bands
- 03 Predict Models built on spectrally refined predictors (WASP, WQM, NPRED)
- 04 Decide Actionable forecasts for water management, agriculture, and energy
03 / Publications
Peer-reviewed research.
Selected publications from the lab. The full list is synced daily from Google Scholar. Papers are grouped into research directions using title keywords. Showing current records.
04 / Showcase
WASP · WAvelet System Prediction.
WASP refines predictor representation using wavelet theory for improved hydrologic prediction. The interactive demo runs on the FastAPI backend deployed on Alibaba Cloud. Upload your data or use the built-in demo dataset.
05 / Contact
Join the HydroclimateX Lab.
We welcome motivated MSc and PhD students and postdoctoral researchers interested in hydroclimate extremes, AI/ML for environmental prediction, and open-source scientific software development.
- Hydroclimate extremes modelling and forecasting
- Wavelet-based methods and AI for water resources
- Multi-agent systems and human–water feedbacks
- Open-source toolkit development (R / Python / Web)