Transfer Operator Framework for Earth System Predictability and Water Cycle Extremes
For chaotic dynamical systems, nonlinear instabilities lead to exponentially divergent trajectories in the evolution of system states. Unless a simulation is initialized with an infinite-precision snapshot of the state of the true system and all known physical effects that go into its evolution are directly computed, the future state predicted by the simulation will quickly diverge from that of the true system. Moreover, the Earth system is highly structured and contains localized coherent structures that are particularly important to predict. Predicting extreme events associated with coherent structures, like hurricanes and blocking events, is crucial for understanding the effects of global warming on the water cycle.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE National Nuclear Security Administration (NNSA)
- DOE Contract Number:
- 89233218CNA000001
- OSTI ID:
- 1772387
- Report Number(s):
- LA-UR-21-22681
- Country of Publication:
- United States
- Language:
- English
Similar Records
Autonomous reinforcement learning agents for improving predictions and observations of extreme climate events
The inherent limitations of population modeling in environmental risk assessment and an alternative: Community conditioning