DATA ANALYSIS ON NON-FILTERED DATA OF HYDROPHONE USING SEDIMENT-CHANNEL EXPERIMENTAL MODEL

Abstract

To clarify the characteristics of non-filtered data of hydrophone, hydraulic experiment was carried out using actual gravel of the Sho River and the Oyabe River in sediment-channel experimental model. Sound pressure data from the hydrophone was analyzed using non-filtered data which sampling frequency was 100 kHz.

As a result of hydraulic experiment, the relationship between maximum sound pressure data and mas and momentum of gravels had large variations. It was too difficult to estimate mass and momentum of gravel using only non-filtered data of the hydrophone. As a result of multiple regression analysis, multiple regression function using maximum sound pressure and dominant frequency of NF data, water depth and water velocity setting as explanatory variables could be most satisfactorily reproduced experiment data.