Artificial Neural Net Modeling Of The Radioactive Contamination Of The Techa River

摘要

To analyze the reasons for the high radioactive contamination of the Techa River in August 2004, when the ~(90)Sr content in the section line at the village of Muslyumovo exceeded 50 Bq/liter, the dynamics of the concentration of ~(90)Sr in Techa River water and the level of the reservoir V-11 were modeled using artificial neural nets. It is concluded that there exists a hidden factor, which is activated at definite times, substantial decreasing the water level in the reservoir V-11, and that cannot be explained on the basis of the precipitation-evaporation balance. The action of this factor is strongly associated with the change of the water flow rate in the left-bank channel. It is suggested that the high radioactive contamination of the Techa River in the summer of 2004 coinciding with the decrease of the water level in V-11, which cannot be explained by the precipitation-evaporation balance, are associated with a discharge of the contaminated water from V-11 into the left-bank bypass.