data analysis // revision

 Our testing methods went through sensor accuracy, prediction accuracy, simulated testing, and on site testing. So, we analyzed the data that we gathered and used this for corrections and then to compare against SRR and analyze.  The first phase of our testing included comparing the sensor readings coming from AvAlert and putting them up against measured values. From this, we could account for the difference by code correcting it to make sure the numbers AvAlert was analyzing were as close to real values as possible.  First, we tested depth by digging real snowpits to compare the depth that AvAlert is reading with the actual pit depth. We performed on-site calibration in order to get reliable values for depth readings, testing at a baseline of 3ft. Within 6 recalibration attempts, we got it so close that our sonar reading was only 0.02 ft away, or 0.6% different than actualized values.   We also found similar success with the accuracy of the other sensors. Temperature was off at most by 0.1 degrees, with an average difference of less than 0.25%. Angle was off by at most 1 degree, which was later code corrected down close to zero. In addition, altitude was off by around 40 feet, which at an altitude around 10-11,000 feet is only around 0.36% off.