Calculation of GLE parameters in real time

The NM-BANGLE-model stand-alone application

Since the beginning of the FP7 NMDB project up to now, the following main tasks for producing the model stand-alone application have been realized:

• Model adjustment to the NMDB situation
  The NM-BANGLE model was initially functioning offline using an extended number of
  neutron monitor stations widely distributed around the globe. To allow its real-time function, in the frame of the NMDB, a crucial adjustment of the initial model code has been realized. Consequently, in its current version the NM-BANGLE model takes into consideration exclusively data from the 23 NMDB-stations
• Code structural adjustment and preparation of the inputs
  The NM-BANGLE model code was altered in order to allow the inserting of its various input parameters in dynamic mode. Moreover, the digital form of the output files was defined, decided and implemented.
• Code compilation
  The initial code of the NM-BANGLE model was created in Matlab programming
  language. In order to produce the model stand-alone application a code compilation in C language has been realized.

Input parameter description
Cosmic ray neutron monitor data from twenty three (23) NMDB stations in real time are planned to be used as the main input of the model.The asymptotic directions of viewing for each one of the NMDB stations are an input of the
NM-BANGLE model too. These input files, however, in the current on-line version of the model, are not calculated inside the stand-alone application frame, for reasons of extreme complexity. Specifically, for every station the asymptotic directions of viewing were estimated once (e.g. for medium magnetic field conditions, i.e. Kp index was a priori set equal to 3) and they are going to be used in this form for all upcoming GLEs. Of course this decision has an uncertainty factor that can influence all future produced real-time GLE-results. For this reason, all preliminary future results have to be interpreted on the
basis of this fact. In a future model version the use of asymptotic directions for different levels of magnetic activity is intended.


NMDB data usage

• NM data preparation for the model
  The NM-BANGLE model data represent the following quantity, calculated for each
  station, for each moment that the application is set to running mode:
  x= (I-Im)/Im
  where, I is the cosmic ray intensity registered by the station at a moment and Im is the medium cosmic ray intensity, calculated for a time period of one hour before the alert for the actual GLE.
• Neutron monitor data downloading
  A new interface to the NMDB data base has been created in order to be able to collect every minute the minutely cosmic ray intensity for each station. Also the same program ask from the database to auto calculate and export the hourly running average of the cosmic ray intensity for the last hour in every duty cycle of the NM-BANGLE model.

NM-BANGLE model function and its collaboration with the ALERT
The start of the NM-BANGLE application execution is given by the ALERT program,
when a GLE is detected. We have created a program that monitors the behavior of the NM-BANGLE application. When the exports of the model are ready, this program can start a new cycle of NM-BANGLE run in order to produce new data. This procedure can be periodically and we have the functionality of controlling the total time in which the NM-BANGLE application continuously will produce new data. The duty cycle of the NM-BANGLE application can vary by the mean of duty time depended from the parameters of the model and the processing power of the server. An example of time sets is 3 min NM-BANGLE duty cycle and one hour continuously running after a GLE Alert.


Description of the output files
A text file containing the following information is produced after each application run:

• Spectral index of the solar cosmic ray spectrum and related errors
• Anisotropy parameter and related errors
• Position of the maximum anisotropy (longitude and latitude) and related errors
• Model parameter C, characterizing the ratio between solar cosmic ray amplitude
  and galactic cosmic ray background, and related errors.
• Linear regression coefficient of determination, characterizing the goodness of the model results and standard deviation.

In its current version, the application produces results in digital form.

For an analytical description of the physical parameters calculated by the model see in Plainaki et al., JGR, 2007 here.