Abstract
The spatial filtering velocimetry is useful in the remote sensing of surface velocity of natural turbulent flows. The essential point of this velocity-sensing is to extract the central frequency of the output signal, which is proportional to the mean velocity of the flow. For extracting the central frequency, the spectrum analysis by means of the maximum entropy method (MEM) was applied.
The performance of this sensing associated with the MEM processing was examined with a computer simulation. The random moving image with a spatio-temporal correlation function was generated by a computer software to simulate various turbulent flows. The spatial filtering of the artifitial random moving images generated the temporal output signal by using the parallel slit reticle. Followings were shown by this simulation: (1) for an artificial random flow having a known component of constant velocity, the estimated error of the velocity in the case that the order number (m) of the MEM processing was 10 was smaller than other cases of m=5, 15, 20 or 25, and (2) for a flow having a time-varing velocity component, the accuracy of the measurement was about 5% in the case of m=10.
This sensing method was successfully used for measuring the surface velocity of debris flows spontaneously occured at Kamikamihori Rabin of Kamikochi.
