University of Wisconsin-Madison

First Year Performance Paper - Section 3.4

3.4 Time resolution studies using the DOM flashers

The LEDs on the flasher-boards were used to measure the photon transit (or delay) time for the reception of a large light pulse at the closest DOM above the one flashing. Fig. 11 shows the distribution of delay times for DOM 46 when DOM 47 was flashing. The mean time delay is given by the light travel time from the flasher to the receiving DOM, together with a small electronics offset. The RMS variation of the time delay, shown in Fig. 12 for 58 DOMpairs, should reveal any imprecision or drifts associated with the clock calibration procedure (§ 3.3). There is also a small contribution from scattering in the ice; the RMS values are smallest for the DOMs located in clearer ice.

Photon arrival time delay at DOM 46 when DOM 47 is flashing in clear ice. The mean and rms values are indicated.
Photon arrival time delay at DOM 46 when DOM 47 is flashing in clear ice. The mean and rms values are indicated.

A typical calibrated waveform captured by a String-21 DOM is shown in Fig. 13. Different waveform features, such as time, amplitude, width and area of the primary and secondaries pulses can be used for reconstruction. These waveforms can be described well by a decomposition procedure that yields an ensemble of single photon hit times. The procedure is based on iteratively fitting a waveform with a function that is a sum of a constant and a progressively larger number of terms each describing a single photo-electron response with allowed variations in amplitude and width. An alternative method applies a Bayesian unfolding algorithm to the waveform with single photo-electron response as a smearing function. Both methods give photon arrival times within 0.5 ns of each other.

RMS variation of time delay mesaured with flashers for 58 DOM pairs on the IceCube string.
RMS variation of time delay mesaured with flashers for 58 DOM pairs on the IceCube string.

In contrast to the data generated by muons in the ice, the air shower signals in the surface tanks are much bigger and require different procedures. Fig. 14 shows the signals from an air-shower seen by all sixteen DOMs in the array, as a function of time relative to the first DOM hit. Each panel shows traces from two different tanks at the station. The top row shows the waveforms seen by the high-gain DOMs (5 × 106) in each tank while the bottom row shows the low-gain (5×105) data. In the event shown here the DOMs have not yet been adjusted to give the same response for a given number of photo-electrons. The integrated waveforms correspond to several hundreds of photoelectrons from the light produced by many particles in the shower front hitting the tank in quick succession.

Captured hit event waveform.
Captured hit event waveform.