Understanding radioactive aerosol behavior is important during severe accidents in nuclear power plants. In particular, pool scrubbing at a suppression pool is a key factor for estimating the leakage of radioactive aerosol into the environment. Therefore, we have to elucidate the mechanism of the aerosol transfer from the gas phase to the liquid phase during bubbly flow. Currently, models to evaluate the amount of aerosol removal by pool scrubbing have been incorporated into integrated severe accident analysis codes such as MELCOR. However, the models for bubbles and aerosols behavior have not been validated enough by experiment. Therefore, the measurement of bubbles and aerosols behavior is required for validation and reconstruction pool scrubbing model. However, the quantitative measurement of bubbly flow is very difficult by complicated bubble shape and overlapping of bubbles in visualization. The wire-mesh sensor (WMS) is one of the most effective measuring methods for complicated bubbly flow. The reconstruction method of bubbly flow by WMS measurement is suggested by Prasser et al. This method can measure the bubble surface, meanwhile, individual bubble velocity and volume are not one-to-one correspondence because bubble volume is calculated by using radial average velocity. Kanai et al. proposed a method for acquisition bubble three-dimensional velocity method by bubble tracking. In this method, bubble velocity and volume are one-to-one correspondence. Thus, we suggested an algorithm of bubble surface measurement by applying a bubble tracking method. The bubble three-dimensional velocity and surface area is obtained by this algorithm. As a result, bubble rising velocity in bubbly flow is higher than bubble terminal velocity according to causing liquid phase velocity. In addition, the gas-liquid total interfacial area is overestimated in MELCOR predict by overestimating bubble breakup. This result suggests that the overestimating of DF at downstream is caused by overestimating bubble breakup in MELCOR model.