Tiantian Yang
University of Michigan, MI·Engineering·Early-career
Floods, droughts and sudden weather extremes make it harder to run dams and reservoirs safely. Operators need reliable forecasts weeks to months ahead, along with decision tools that turn those forecasts into flexible plans for releasing water.
- Fields
Engineering
Earth & Environmental Science
Computer Science
- Topics
reservoir operation
water resources management
hydrology
- Grants
- 1 known grant, through Aug 2028
What they work on
- Correct errors in space and time in subseasonal-to-seasonal precipitation forecasts from the North American Multi-Model Ensemble using deep learning models.
- Develop an Ensemble Boosting Tree-based Predictive Control Model that uses improved ensemble forecasts to simulate and plan adaptive reservoir water releases.
- Quantify how weather uncertainty spreads into hydrological forecasts and reservoir decisions through hydrological modeling across 671 U.S. watersheds and 316 dams, tested with USBR and USACE.
More about the research
Current rule-based and optimization methods struggle to use uncertain forecast ensembles, which are sets of many possible future scenarios.
Why it matters
Dam operators at agencies such as the Bureau of Reclamation and the Army Corps of Engineers could use better long-range forecasts and machine-learning decision tools. These could help them plan water storage and releases that handle sudden floods or shortages while still meeting water supply needs. Communities downstream of reservoirs could face less risk from extreme weather and dam failures.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Zhuo Wang
University of Illinois Urbana-Champaign, IL·Earth & Environmental Science
Investigate how predictable weather extremes and their environmental impacts on Midwest agriculture are at subseasonal to seasonal (2-week to seasonal) forecast lead times.
Steven Greybush
Pennsylvania State University, PA·Earth & Environmental Science
Design and optimize a deep learning nowcasting system for convection initiation that combines forecasting, diagnosis and generation of storm scenarios, trained on high-resolution satellite, radar and surface weather data.
Andrew Gronewold
University of Michigan, MI·Earth & Environmental Science
Produce reliable projections of how often and how intensely climate-driven floods, droughts and related hazards will strike U.S.-Canada transboundary waters, starting with the Great Lakes, using statistical and process-based models.
The school
University of Michigan: 124 labs on PI Pages, $2.1B a year on research, 963 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.