Modified ensemble Kalman filter for nuclear accident atmospheric dispersion: prediction improved and source estimated

Area of Science:

• Environmental Science
• Nuclear Engineering
• Atmospheric Science

Background:

• Accurate atmospheric dispersion modeling is crucial for nuclear power plant accident management.
• Estimating radioactive material distribution within 50 km is vital for evacuation and sheltering.
• Meteorological data and source terms are often uncertain in early emergency phases.

Purpose of the Study:

• To develop a method for improving short-range atmospheric dispersion predictions.
• To simultaneously enhance model accuracy and reconstruct source terms using off-site data.
• To address uncertainties in source release rate, plume rise, wind speed, and direction.

Main Methods:

• A modified ensemble Kalman filter data assimilation technique was employed.
• The method was integrated with a Lagrangian puff-model for dispersion simulation.
• Twin experiments were conducted to validate the approach.

Main Results:

• The proposed method significantly improved predicted concentration distributions.
• Temporal profiles of source release rate and plume rise height were successfully reconstructed.
• The response time lag of the ensemble Kalman filter was reduced.

Conclusions:

• The developed method offers a valuable tool for nuclear accident emergency management.
• It enhances the reliability of short-range atmospheric dispersion predictions.
• The approach is applicable to other hazardous material release scenarios.