confusing atmospheric extinction profile

[jsitarek]

the current code selecting the atmospheric extinction file is really confusing. https://github.com/cta-observatory/lst-sim-config/blob/7a716348530076bd1972f7cbef9e13af81b4f0a9/array.cfg#L26-L27 says that we want both low extinction and NAVY MARITIME, but then when we check the actual profile there is one set of conditions that selects LOW_EXTINCTION: https://github.com/cta-observatory/lst-sim-config/blob/7a716348530076bd1972f7cbef9e13af81b4f0a9/site.cfg#L22-L31 but then it is overwritten by a second set of conditions that selects NAVY MARITIME: https://github.com/cta-observatory/lst-sim-config/blob/7a716348530076bd1972f7cbef9e13af81b4f0a9/site.cfg#L39-L45

which one do we want in the end? I saw that @YoshikiOhtani was running for LOW_EXTINCTION one by specifying the atmosphere transmission file directly as a parameter I suggest to remove all those conditions and set the atmosphere profile directly the the profile that we want @moralejo @rlopezcoto which one should it be?

[moralejo]

We should in principle use the same as in the previous LST- dedicated production, i.e. low_extinction. The fact that we used that one probably means that part of the light reduction (w.r.t. nominal) we see with muon rings may be atmospheric and not telescope- related. Do you know what is the difference in transmission between the two?

[jsitarek]

I made the test and the difference looks like 4.5% (low extiction, in red, has more light as the name says).

I think I understand what happened. The original site file had a long if statement if PARANAL ... elif LA PALMA .... else ...

the PARANAL block was removed, and elif LA PALMA and else condition as well, but what was left was the LP block and the actual else block. So the LP values were always overwritten by the old "else" values.

EDIT: all this is however in the if TELESCOPE == 0 block, so if the setting is set by the command line, whatever happens here gets overwritten so should be fine in the end. Still it is really confusing to me to set it in both places and this can result in problems if some subproductions are done in different way.

I will make a small fix removing the old else block, because this one is for sure not needed, but maybe the whole way of passing the atmosphere should be rethought. Let's discuss in PR

[maxnoe]

We should not mix the pfp precompiler style of setting things with the templating used by the sim runner.

[jsitarek]

I agree @maxnoe I made PR #32 to remove the erroneous part of the precompiler conditions, but the LP settings are still there. I think it would be more clean to have this atmosphere settings all in the .cfg files without setting them by hand, but we can also do opposite, remove it completely from the site.cfg and use only the -C option.

[moralejo]

Hi @jsitarek, thanks for the test. In the standard CTA simulations for CTA-N Navy maritime is used instead. This means that the tuning of the telescope efficiency that we did (with low_extinction) is not applicable to the CTA simulations.

@YoshikiOhtani, @YusukeSuda, was there any reason for adopting the low_extinction option? We should either tell the team preparing the next CTA simulations to use it, or keep the Navy maritime, and then we modify the value we gave them of the telescope efficiency.

[YoshikiOhtani]

Hi, when I run the productions I copied the CTA-PROD5-LaPalma-baseline.cfg and only modified the telescope configurations, and so I just didn't touch and thus kept the option as default.

[moralejo]

Thanks @YoshikiOhtani. You are right, I had checked the CTA-PROD4-LaPalma-Baseline.cfg, and in that one LOW_EXTINCTION is commented out.

In the CTA-PROD5-LaPalma-Baseline.cfg it is again active, so both Prod5 and your LST production used LOW_EXTINCTION. I will find out whether this is also the idea for the next CTA prod.

[jsitarek]

Hi, the log that I got from @YoshikiOhtani had Starting ./bin/sim_telarray with the following arguments: ...'atmospheric_transmission=atm_trans_2158_1_3_2_0_0_0.1_0.1.dat'

this is the file for the low extinction. Same thing could have happened with PROD4, I think this value in command line overwrites the value selected in the site file (because this is done only for telescope nr 0)

[maxnoe]

$ eventio_print_simtel_history /fefs/aswg/data/mc/DL0/20200629_prod5_trans_80/gamma/zenith_20deg/south_pointing/off0.4deg/gamma_20deg_180deg_run101___cta-prod5-lapalma_4LSTs_MAGIC_desert-2158m_mono_off0.4.simtel.gz | grep -i atm
2020-11-29T09:24:40 ./bin/sim_telarray -c cfg/CTA/CTA-PROD5-LaPalma-baseline_4LSTs_MAGIC.cfg -DNECTARCAM -DNO_STEREO_TRIGGER -DNUM_TELESCOPES=6 -Icfg/CTA -C Altitude=2158 -C iobuf_maximum=1000000000 -C maximum_telescopes=6 -C atmospheric_transmission=atm_trans_2158_1_3_2_0_0_0.1_0.1.dat -C altitude=2158 -C telescope_theta=20.4 -C telescope_phi=180 -C power_law=2.50 -C histogram_file=Data/sim_telarray/cta-prod5-lapalma_4LSTs_MAGIC/0.4deg/Histograms/gamma_20deg_180deg_run101___cta-prod5-lapalma_4LSTs_MAGIC_desert-2158m_mono_off0.4.hdata.gz -C output_file=Data/sim_telarray/cta-prod5-lapalma_4LSTs_MAGIC/0.4deg/Data/gamma_20deg_180deg_run101___cta-prod5-lapalma_4LSTs_MAGIC_desert-2158m_mono_off0.4.simtel.gz -C random_state=auto -C show=all -
2020-11-29T09:24:40 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:40 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:40 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:40 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:41 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:41 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:41 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat
2020-11-29T09:24:41 ATMOSPHERIC_TRANSMISSION atm_trans_2158_1_3_2_0_0_0.1_0.1.dat