[Paper Review] Early attempts at atmospheric simulations for the Cherenkov Telescope Array

C. B. Rulten, Sam Nolan|arXiv (Cornell University)|Mar 10, 2014

Astrophysics and Cosmic Phenomena1 references3 citations

TL;DR

This paper proposes using lidar-derived atmospheric transmission profiles to correct for variable atmospheric conditions in ground-based gamma-ray astronomy, specifically for the Cherenkov Telescope Array (CTA). By fitting MODTRAN models to lidar data, the study shows that correcting for increased aerosol density can restore accurate gamma-ray energy spectra, reducing systematic errors and enabling recovery of otherwise unusable data.

ABSTRACT

The Cherenkov Telescope Array (CTA) will be the world's first observatory for detecting gamma-rays from astrophysical phenomena and is now in its prototyping phase with construction expected to begin in 2015/16. In this work we present the results from early attempts at detailed simulation studies performed to assess the need for atmospheric monitoring. This will include discussion of some lidar analysis methods with a view to determining a range resolved atmospheric transmission profile. We find that under increased aerosol density levels, simulated gamma-ray astronomy data is systematically shifted leading to softer spectra. With lidar data we show that it is possible to fit atmospheric transmission models needed for generating lookup tables, which are used to infer the energy of a gamma-ray event, thus making it possible to correct affected data that would otherwise be considered unusable.

Motivation & Objective

• To assess the impact of variable atmospheric conditions on gamma-ray energy reconstruction in the Cherenkov Telescope Array (CTA).
• To evaluate whether lidar measurements can enable active atmospheric calibration to correct for changing aerosol density and improve spectral accuracy.
• To demonstrate that data affected by poor atmospheric conditions can be corrected and reused, increasing the observatory's duty cycle.
• To develop a method for generating accurate lookup tables using lidar-derived transmission profiles that reflect real-time atmospheric quality.

Proposed method

• Lidar measurements were collected using a Leosphere Easy-Lidar ALS450XT at the H.E.S.S. site in Namibia on 15 August 2008, with a 355 nm wavelength and 10 Hz pulse frequency.
• The Klett inversion and multiangle methods were applied to lidar backscatter data to derive range-resolved atmospheric transmission profiles up to 20 km altitude.
• MODTRAN simulations were used to fit aerosol density profiles to the lidar-derived transmission, generating a 'best-fit' aerosol model for high aerosol conditions.
• Atmospheric air shower simulations were performed using CORSIKA, and telescope responses were simulated to generate lookup tables for energy reconstruction.
• Three simulation cases were compared: normal aerosol density, increased aerosol density with normal lookup tables, and increased aerosol density with corrected lookup tables.
• Reconstructed differential spectra were generated and fitted to a power-law model to quantify systematic shifts in spectral index due to atmospheric effects.

Experimental results

Research questions

• RQ1 How do changes in atmospheric aerosol density affect the reconstructed energy spectrum of gamma-ray events in CTA simulations?
• RQ2 Can lidar-derived transmission profiles be used to accurately model atmospheric conditions for use in simulation-based energy reconstruction?
• RQ3 To what extent can correcting for atmospheric transmission improve the accuracy of reconstructed gamma-ray spectra under poor atmospheric conditions?
• RQ4 Can active atmospheric calibration using lidar reduce systematic uncertainties in flux and energy resolution for ground-based gamma-ray telescopes?
• RQ5 What is the impact of using incorrect atmospheric models on the spectral index and flux reconstruction in CTA?

Key findings

• Under increased aerosol density, the reconstructed gamma-ray spectrum is systematically shifted to softer slopes, with a power-law index increasing from 1.93 (normal) to 2.34 (high aerosol).
• Using a lidar-derived 'best-fit' aerosol model to generate lookup tables restores the reconstructed spectrum to a power-law index of 1.91, closely matching the normal atmospheric condition case.
• The differential flux normalization (Io) is significantly reduced in high aerosol conditions (77 ± 2 events TeV⁻¹) compared to normal conditions (198 ± 3 events TeV⁻¹), indicating reduced effective sensitivity.
• Lidar-based transmission measurements allow for the derivation of accurate atmospheric models that can be used to correct simulation lookup tables, enabling data recovery from otherwise discarded observations.
• The study confirms that systematic errors from atmospheric variability can be mitigated using in-situ lidar measurements, improving spectral accuracy and reducing flux uncertainty.
• Single-wavelength lidar systems have ~30% systematic error in transmission, but Raman lidars with ~5% error offer a promising path for future active calibration in CTA.