WaterDMD builds digital twins of complex water systems, virtual models that integrate natural and human components. Using AI, drone monitoring, hyperspectral imaging, and satellite Earth observations, we develop models that simulate and predict the behavior of interconnected water, agriculture, and urban systems.
Current Research Themes
All research projects-
Continental Scale Crop Land Cover Land Use Classification
Deep learning and high-resolution satellite data to classify crop types and land use at continental scale, supporting food security and resource management.
-
Global Soil Carbon Models
Integrating satellite and ground measurements to model current and attainable soil carbon stocks globally, improving estimates of sequestration potential.
-
Hyperspectral Imaging for Crop Stress
Detecting early crop stress from drought, pests, and nutrient deficiency using hyperspectral sensors that capture changes invisible to the naked eye.
-
AI Models for Streamflow Predictions
Machine learning models for streamflow prediction using hydrological, meteorological, and geospatial data, supporting flood forecasting and drought preparedness.
-
Assimilation of Aerial and Spaceborne Earth Observations
Assimilating aerial and satellite observations of soil moisture, evapotranspiration and vegetation indices into crop and water models to improve forecast accuracy.
-
Stakeholder Mental Models
Understanding how stakeholders perceive complex water systems, and integrating those perspectives into our models to improve real-world relevance.