New light-based method predicts complex fluid movements
Researchers used a photonic system to model atmospheric circulation and thermal convection.
Scientists have developed a way to use light to predict how fluids move. This method uses a specialized photonic system to model complex flows. It can simulate systems like global atmospheric circulation.
Previously, boson sampling was used to connect physics with computer science. It was often used for tasks like image recognition or classifying data. However, earlier versions had scaling issues that limited their use.
The researchers created a new scheme called bright scattershot boson sampling, or BSBS. This method uses higher photon number components to solve previous scaling problems. It uses all randomly generated input photon number distributions to work more efficiently.
The team used a digital twin to test the BSBS architecture. This is a simulation based on multiphoton statistics. They tested it on an eight-dimensional Lorenz-type model of thermal convection. They also tested a 16-dimensional model of global atmospheric circulation.
The photonic approach accurately reproduced the timing of these systems. It remained effective even with photon losses of up to 50 per cent. The BSBS method performed better than using coherent light. It also had a longer prediction horizon than a classical reservoir-computing model.
A reservoir computing model, or RCM, is a system with an input layer, a reservoir with memory, and an output layer. In this study, the researchers used a feedback-driven protocol. This means a measurement at time step helps determine the next step at .
The researchers mapped the RCM onto a reconfigurable bright scattershot boson sampler. This setup exploits high-dimensional correlations from measurement outcomes. This approach helps model nonlinear dynamical systems without solving complex equations.
These results show that near-term photonic quantum computing can address real-world tasks. This is possible even on noisy intermediate-scale devices. The work moves toward modeling fully developed turbulence on a photonic platform.
Key facts
- The BSBS scheme can model 16-dimensional atmospheric circulation.
- The system works even when 50 per cent of photons are lost.
- The photonic model predicts further into the future than classical reservoir computing.