Page 417 - AWSAR 2.0

numerous others.
Upon distributing the
amplitude in the form of
power law and applying some
order statistical tools such as
maximum likelihood estimate),
we get the power (exponent)
value for the data set used.
This exponent, also known as
the b-value, gives us pretty
usable information about
the mechanical health of the
snow sample being used. Mathematically, the formula
states that the exponent will
become smaller when the
size of events (AE amplitudes) increases. A variation of b-value with time was observed for all the experimental sets, and it was found that its value reduced consistently prior to a fracture. This was confirmed on all samples with the help of the load cell and video- recorded data.
The same procedure was applied to the AE data recorded from an experimental avalanche slope in the upper Himachal Pradesh region. During the complete winter season, the AE data contained multiple snowstorms and a single avalanche event in it. As the data contained the AE from an actual
Mr. Rahul Sheoran || 393
avalanche slope, it had a lot of external noise. The noisy events were removed using a larger sample size as well as using the data from an automatic weather station (AWS) installed close to the site. The extreme weather events were identified from the AWS data and removed for the final processing to be done. The reduction in b-value was observed 8 h prior to avalanche, and it reached global minima half an hour before the event.
This was a single field trial with an actual avalanche and hence could not be described as a dependable proof of successful forecasting; but, what it proves is that this method of AE data processing holds a good amount of potential in the field of avalanche prediction and warning. With additional field data and improved statistical algorithms (procedures), the methodology can mature into a dependable predictor of an avalanche. As more data start to pour in, the robustness of the system will increase and the risk factors for our jawans will be reduced during their patrolling and transit in
the snowbound areas of the Great Himalayas.
Depending on different compression rates, some samples failed (fracture in the weak layer) while others did not. The AE amplitude (maximum voltage peak generated for each micro-fracture) parameter was used to study the internal micro-fractural response of each sample to the applied load.