Generating Real Run Conditions

This section explains how to generate Monte Carlo data with realistic run conditions. When the detector takes data there are many run conditions that can vary with time. These can include the number of tubes online, the detector responce to energy, the PMT noise rate and the date and time of the run. Also if a calibration source is present the type and position of the source will vary. Many of the data that are used to setup the time dependance of the detecotor are stored as titles banks in the database. A good example of this are the DQxx banks which record the status of each PMT. Some required data are not however stored in titles banks, and this makes the monte carlo generation of a given run a more difficult task than it should be.

The required data that is not stored in titles bank is however still stored in the database in the runlogger system (see section 24). This data can be used to setup the conditions of a given run inside SNOMAN. In order to use this code the run number must be specified. This is done via the symbolic command:

• $mcrun 10000

When the command

• $mc_gen_run_cond $on

is issued SNOMAN will set up the Monte Carlo with the initial conditions of the specified run. The specific conditions set are the:

• The start date and time taken from the RLDT bank
• The run duration taken from the RLDT bank
• The PMT noise rate taken from the RLPN bank

The setting of the PMT noise rate and the run duration are carried out by internal set bank commands that take place during the context setting phase of MCO. The use of internal set banks means that each setting can be overridden in the command file, allowing the user control over the value that each takes. This is not the case for the start date and time. These are controlled by the job bank which has already initialized by the time the context setting phase of MCO is carried out. Other than the inability to override the start date and time in the RLDT bank however the start date and time can be treated from the perspective of the user in exactly the same way as the other parameters.

The setup of run conditions is useful when carrying out Monte Carlo generation of neutrino data, but still requires user input for calibration data. However by use of the command:

• $mc_gen_source_run $on

the user is no longer required to specify source inputs either. During generation of the runlogger the RHDR and CAST banks are stored in the database. The RHDR bank specifies the type of source used in the specified run. The CAST banks contain the position of the manipulator pivot. This is corrected for the pivot source offset to setup the source position using the CASL bank. For each supported source the source geometry, particle decays and any other required source settings are setup via internal set banks. As before the use of internal set banks allows the user to override any given setting with values of their choosing. The currently supported calibration sources are

• The $^{16}$N source.
• The $^8$Li source.
• The $^{252}$Cf fission source.
• The pT source.
• The acrylic U and Th sources

One time dependent issue that is not dealt with using data stored in the runlogger system is the drifting gain of the detector. One way to add this to the simulation is to drift the PMT collection efficiency as a function of time. By the use of the symbolic command:

• $mc_drift_pmt_coll $on

drifting of the collection efficiency is enabled. The titles bank MCDR contains the drift as a 10th order polynomial between the start and end of the specified validity range. To include a discontinuity in the efficiency, it's gradient etc one must use several banks with different validity ranges. When the gain is being drifted the collection efficiency is calculated for each event given the collection efficiency specified at run time and the contents of the MCDR bank. As the collection efficiency has become a dynamic quantity it is stored in the MC bank to allow later extraction.

The use of these three switches allows the user to generate Monte Carlo with the same conditions as real data, only by the specification of the run number and avoids the need for independent determination of such conditions.