The source information of the radionuclide release in nuclear accidents is a key issue of the nuclear emergency response. One way to estimate the source information is by inversing the radionuclide transportation process based on environment radiation monitoring data. The advantage of this method is that the required data are easy to obtain in accident. But it is vulnerable to large uncertainties in both data and transport model. To solve the problem, a source term estimation method based on four-dimensional variational (4DVAR) data assimilation technique was proposed for source term estimation in this study. The proposed method couples 4DVAR with the RIMPUFF air dispersion model. It formulates the inverse modelling of source term estimation as an optimization problem that is trying to find an optimal balance between real observation data and the background field. The advantage of this method is that the radionuclide transport in every time step is included in data assimilation and the result is global optimum in the whole assimilation period. The gradient for cost function is calculated by the backward integration of the adjoint model. Practical imperfectness of measurement were considered and integrated into the cost function. The proposed method was verified using numerical simulation for both homogeneous and heterologous atmospheric condition. The performance of the source term estimation method was also investigated with respect to different release profile, wind speed and atmospheric stability class. The simulation results demonstrate that the estimate matches the true release status well for both homogeneous and heterologous wind field. Also, the experimental results show that the proposed method has strong robustness to wind speed and atmospheric stability.
Verification of Source Term Estimation Method against Measured Data for Decommissioning Waste from a CANDU Reactor
