Overview of the model

The image below shows a schematic overview of how the model is built up.

Output of the model

The model outputs the following data:

1. Vector of the time: all moments in time at which the signal is calculated in s.
2. Matrix of the power: matrix of all the power values of the signal in W. It has dimensions [5, #filters, #timesamples]. The first dimension equals 5, as 5 pwv values are taken in each timesample.
3. Matrix of the sky temperature: matrix of the power converted to sky temperature (has the same dimensions as the matrix of the power) in K.
4. Center frequencies of the filters: The center frequencies of the filters in the filterbank in the MKID chip in Hz.

The pwv values are taken in the following order:

Using the model

The model is operated from main.py. In this file, there are different variables that can be set:

Atmosphere

pwv_0 (float): The value of the precipitable water vapor that is added to the dpwv from ARIS in mm. windspeed (float): The windspeed of the atmosphere in m/s. prefix_atm_data (string): The beginning of the name with which the atmosphere data is saved in ./Data/output_ARIS. For example, if the files are called sample-00.dat-000, sample-00.dat-001 etc, then Prefix_atm_data must be 'sample-00.dat-' grid (float): The width of a grid square in the atmosphere map in m max_num_strips (integer): The number of atmosphere strips that is saved in ./Data/output_ARIS x_length_strips (float): The length of one atmosphere strip in the x direction. This is the number of gridpoints, not the distance in meters. Thus, this is an integer number, taken as a float to make sure calculations with this number result in a float.

Galaxy

luminosity (float): Luminosity of the galaxy, in Log(L_fir [L_sol]) redshift (float): The redshift of the galaxy linewidth (float): The linewidth, in km/s

Observation

EL (float): The elevation of the telescope, in degrees obs_time (float): The observation time. This parameter has to be smaller than max_obs_time, which is calculated using the windspeed and the total length of the strips of atmosphere data, in s. draw_filters (list): List of the filters that need to be plotted in the final plot (starts counting at 1, NOT 0). For example, if draw_filters = [1, 2], the first and the second filter are plotted, and if draw_filters = [1, 347], the first and the last filter are plotted. save_name_plot (string): The name with which the produced plot is saved. save_name_data (string): The name with which the produced data is saved to .Data/output_DESHIMA_model/. The center frequencies of the filters are saved as save_name_data_F.npy and the sky temperature matrix is saved as save_name_data_T.npy

Important

1. All atmosphere strips must have the same length in the x direction and a length in the y direction of at least 30 gridpoints. ('length' means number of gridpoints, not distance in meters)
2. If pwv_0 is changed, signal_transmitter.save_filtered_pwv_map() needs to run again to save the right pwv map (See section 'Changing the atmosphere data' to see how').

Changing the atmosphere data

1. Save the new atmosphere data in ./Data/output_ARIS/
2. Change prefix_atm_data, grid, max_num_strips and x_length_strips in main.py and the x_length_strips property of use_aris.py to the right values.
3. From main.py run signal_transmitter.save_filtered_pwv_map() with the signal_transmitter instance in main.py. This function loads in and filters the complete atmosphere map (consisting of all atmosphere strips). This will save computation time when simulations with a large observing time are done.

Now main.py can be used like always again.

Changing the number of filters or the distribution of the center frequencies of the filters

For each filter, an interpolation between the power and the sky temperature is made. This means that these interpolations need to be made and saved again if the center frequencies of the filters are changed. This is done as follows:

1. Go to ./DESHIMA/MKID/filterbank.py
2. Change num_filters for a different number of filters, F_min for a different lowest center frequency or R for a different spectral resolution. For a different distribution of the center frequencies, edit self.filters in the init function of the filterbank class.
3. Run filterbank.py

Now the new interpolation curves should be saved in ./Data/splines_Tb_sky. Now make sure self.filters in signal_transmitter.py is the same vector as self.filters in filterbank.py and the model can be used again.