SKA configuration changes

[garrelt]

The design of SKA-Low has been changing over the years and the design that's used in tools21cm is probably no longer valid. Specifically, what will be built is called AA* and the baseline design, AA4, may be built in the future, if funding appears.
The most up to date information about SKA design (station positions, etc.) has been made available by SKAO in a python package, see https://www.skao.int/en/science-users/ska-tools/494/ska-staged-delivery-array-assemblies-and-subarrays
It would be good to make use of this in tools21cm.

[micbia]

I need to double-check if the new configuration they mention is the same that we used during the SKA Data challenge 3a (which was provided by SKAO). If that is the case, I have already prepared the new configuration in a format that can be implemented in tools21cm

[sambit-giri]

If they match, please add those files to the src/tools21cm/input_data folder.

[sambit-giri]

I have now added the files provided by @micbia to the above folder, which can be access through the function get_SKA_Low_layout.
These changes are related to commit 0d63a9a.

[sambit-giri]

As the new files contain the antennae layout in cartesian corrdinates, contrary to the previous configuration in tools21cm that was in geographic coordinates. However, the calculation of uv tracks get_uv_daily_observation converted the geographic coordinates to cartesian. To have the flexibility to switch between the corrdinates, I have introduced two new functions that are geographic_to_cartesian_coordinate_system and cartesian_to_geographic_coordinate_system.

[sambit-giri]

I have a working version (post this change) where functions relying on telescope layout, such as get_uv_map and noise_cube_coeval, can use different SKA antenna configurations corresponding to different construction stages. These functions take a string (AA0.5, AA1, AA2, AAstar or AA4) via an input variable subarray_type. An example to work with these layouts is provided in this Jupyter notebook.

[sambit-giri]

Additionally, I have renamed some scripts (noise_model.py->radio_telescope_noise.py, telescope_functions.py->radio_telescope_sensitivity.py) to group together all the modules that model the instrumental effects (radio_telescope_*.py files).

[sambit-giri]

@micbia has added a SKA antenna sensitivity table, which is given as the System Equivalent Flux Density (SEFD):

$\mathrm{SEFD} = \frac{2k_\mathrm{B}T_\mathrm{sys}}{A_\mathrm{eff}}$ ,

where $k_\mathrm{B}$, $T_\mathrm{sys}$ and $A_\mathrm{eff}$ are the Boltzmann constant, system temperature and effective collecting area of the antenna. We will add more tables as the SEFD depends on the observation strategy, such as the direction of observtion and time of the day.

We used the aproximate parametric forms for $T_\mathrm{sys}$ and $A_\mathrm{eff}$ (see e.g., Giri et al. 2018; Bianco et al. 2023). Currently, we are expanding the features to use these SEFD tables.