The observations of the inner corona (up to 3 R_ʘ) in white-light and emission lines have been made during total solar eclipses over the years, yielding a detailed description of the corona (Baumbach, 1937; van de Hulst, 1950; Habbal et al., 2007, 2010; Habbal et al., 2014; Boe et al., 2018). During total solar eclipses, spectroscopic investigations using the emission lines detected the presence of oscillations and fast magnetohydrodynamic (MHD) waves in the solar atmosphere (Singh et al., 1997; Pasachoff et al., 2002; Sakurai et al., 2002; Singh et al., 2011; Samanta et al., 2016). As eclipses last for a couple of minutes, regular observations of the inner corona will help improve our understanding of the solar atmosphere. It may also shed light on the development of small- and large-scale transients that lead to severe space weather. The availability of observations in emission lines such as Hα 6563 Å, Fe IX 4359 Å, Fe X 6374 Å, Fe XI 7892 Å, Fe XIII 10747 Å, Fe XIV 5303 Å, and Ni XV 6702 Å can provide useful thermodynamic diagnostics of the different processes occurring in the solar corona (Habbal et al., 2011). It should be noted that current ground- and space-based instruments lack coronal observations in the majority of these wavelengths. The Coronal Multi-channel Polarimeter (CoMP) instrument (Tomczyk et al., 2008) in the Mauna Loa Solar Observatory (MLSO) provided regular spectropolarimetric observations of the inner corona up to 1.5 R_ʘ at 10,747 and 10,798 Å until 2018. The detection of Alfvénic waves in the solar corona using the CoMP data provided support for wave-based models for coronal heating (Tomczyk et al., 2007; Morton et al., 2015; Morton et al., 2016; Morton et al., 2019). The observations made using this instrument also provided the first global magnetic map of the solar corona (Yang et al., 2020). Another ground-based instrument, the Norikura coronagraph, located at Norikura, Japan, provided spectroscopic observations of the off-limb corona in emission lines corresponding to Fe X, Fe XI, Fe XIII, and Fe XIV. The emission line–intensity ratios give an estimate of the temperature, while the line width could be used to calculate the thermal and nonthermal structure of the emitting plasma. Such high-resolution spectroscopic observations from Norikura and CoMP revealed the thermal and nonthermal variations in coronal structures and velocities, complex variations in line–intensity ratios inferring the distribution of multi-thermal plasma, and turbulence in these structures (Singh et al., 2003; Singh et al., 2004; Singh et al., 2006a; Singh et al., 2006b; McIntosh and De Pontieu, 2012; Prasad et al., 2013; Morton et al., 2016; Fan et al., 2018; Pant et al., 2019; Tiwari et al., 2019).

The variations in the temperature and nonthermal velocity in coronal structures may give an insight into the processes involved in heating the corona and the acceleration of the solar wind. Therefore, continuous spectroscopic monitoring of the solar corona in such emission lines is required. Among the existing space-based spectrograph instruments, the EUV Imaging Spectrometer (EIS) on Hinode observes the solar corona and the upper transition region in the wavelength range of 170–210 Å and 250–290 Å (Culhane et al., 2007). Another spacecraft, the Interface Region Imaging Spectrograph (IRIS), takes simultaneous imaging and spectroscopic observations of the photosphere, chromosphere, transition region, and corona in three passbands of 1,332–1,358 Å, 1,389–1,407 Å, and 2,783–2,834 Å (De Pontieu et al., 2014). The Spectral Imaging of the Coronal Environment (SPICE) instrument (SPICE Consortium et al., 2020) onboard the recently launched Solar Orbiter is an imaging spectrometer capable of observing the corona in extreme ultraviolet (EUV) passbands of 70.4–79.0 nm and 97.3–104.9 nm. It should be noted that EIS, IRIS, and SPICE perform the observations confined to a small field of view (FOV).

The Visible Emission Line Coronagraph (VELC) onboard Aditya-L1 (Seetha and Megala, 2017) will take simultaneous imaging and spectroscopic observations of the inner solar corona in three visible and one infrared (IR) passband from 1.05 to 1.5 R_ʘ (Singh et al., 2013; Banerjee et al., 2017; Raghavendra Prasad et al., 2017). VELC is equipped with a multi-slit spectrograph to study the solar corona with high spatial and temporal resolutions in the three emission lines centered at 5,303 Å (Fe XIV), 7,892 Å (Fe XI), and 10,747 Å (Fe XIII), mostly used during the eclipse observations. The continuous observations provided by VELC will be helpful in attaining its scientific objectives, including the diagnostics of the coronal plasma for temperature and velocity, thereby understanding the process of coronal heating and the acceleration of the solar wind. Moreover, the spectropolarimeteric capability of VELC using the 10,747 Å emission line will be helpful in directly estimating the magnetic fields of the solar corona. Therefore, it is necessary to understand the performance of the instrument beforehand to plan the observations after launch.

In this article, we present the characterization of the VELC spectral channels using synthetic spectra generated using CHIANTI 8.0 for different coronal conditions, taking the instrument parameters into account. In Section 2, we present the process involved in synthesizing the spectra and converting it to simulated observations. We present a detailed analysis and the results including the optimization of the slit width of VELC, followed by the instrument performance for different conditions, in Section 3. Section 4 summarizes the analysis, followed by a discussion.

As discussed before, VELC will take spectroscopic observations of the inner corona at three emission wavelengths centered at 5,303, 7,892, and 10,747 Å. We used the CHIANTI 8.0 atomic database (Dere et al., 1997; Del Zanna et al., 2015) to generate emission spectra for these three ionization states of iron. It should be noted that the mechanism responsible for the emission corona is atomic transitions, unlike the scattering of photospheric light in the case of the K and F corona. Also, it has been found that the F corona dominates beyond 2.5–3 R_ʘ (Morgan and Habbal, 2007), which is beyond the VELC field of view (FOV) of the spectroscopic channels. The emission lines under consideration are about 100 times brighter than the background K continuum (Stix, 2002). We have included the continuum while preparing the synthetic spectra with CHIANTI. The contribution of the instrument to the full width half maximum (FWHM) of the spectra is calculated using the following relation: FWHM instr = ( dispersion pixel scale × slit scale ) 2 + ( dispersion ) 2 , (1) where “dispersion” corresponds to 28.4, 31, and 227.3 mÅ/pixel for the three channels, respectively (Singh et al., 2019). A pixel size of 6.5 μm for the visible spectral channels and 25 μm for the IR channel result in pixel scales of 1.25 arcsec pixel^-1 and 4.8 arcsec pixel^-1, respectively. It should be realized from the optical layout of VELC (Figure 1 in Kumar et al., 2018) that the set of four slits in the slit plane placed in the optical path is incident on the spectrograph. We vary the slit width as 20, 40, and 60 μm and investigate their effects on the instrument output (see Section 3). The density and emission measures for the different scenarios are supplied as inputs to CHIANTI. The synthetic spectra are computed using IDL procedures of CHIANTI, ch_synthetic.pro and make_chianti_spec.pro. The instrumental FWHM calculated using Eq. 1 is provided as an input to make_chianti_spec.pro for the instrument-induced FWHM to the synthetic line. The peak of the intensity obtained, which has the physical unit of photons cm^-2 sr^-1 s^-1 Å^-1, is then converted to equivalent photoelectrons incident on each pixel per second using the following relation: ph _ elecs peak = ( peak intensity ) × area × ( solid angle ) × dispersion × efficiency ph . electrons / pixel/s , (2) where the area of the VELC primary is obtained by taking the diameter of the primary mirror as 195 mm (Kumar et al., 2018). The solid angle subtended is calculated using the slit scale and pixel scale along the horizontal and vertical dimensions, respectively. The efficiency of the spectral channels in the visible and IR wavelengths is taken to be ∼ 5 and ∼ 4%, respectively (Singh et al., 2019).

The instrument contribution to the scattered intensity and noise arising due to the detector is also included in the simulated spectra. Scatter studies for the continuum channel of VELC (observing at 5,000 Å with the 10 Å passband) were carried out using Advanced System Analysis Program (ASAP) by Venkata et al. (2017). As VELC has narrowband filters for the three channels, we scaled these scattered intensity values, taking into account the passband of individual channels. The scatter of the continuum was scaled to the spectral channels using the following equation: scatter spec = dispersion 10 × slit scale × ( pixel scale ) spectral ( pixel scale ) continuum 2 × ( Scatter continuum )  ph . electrons / pixel/s , (3) which gives the number of scattered photoelectrons in the spectral channel generated in a pixel every second. The pixel scale for the continuum channel is 2.5 arcsec pixel^-1. The constant factor in Eq. 3 is obtained using the procedure for conversion followed by Patel et al. (2018). The scatter is assumed to be circularly symmetric and is added to the synthetic spectra obtained using CHIANTI in Eq. 2. This could be considered as the worst case scenario as scatter is inversely related to the square of the incident wavelength. The final spectra obtained are then used for further analysis.

The synthetic spectra for each channel are added with the dark noise associated with individual detectors. For the visible spectral channels using the CMOS detector, the dark noise (D) is up to 15 electrons, whereas in the high-gain mode of the IR CCD, it is 42 electrons, with a readout noise (R) of 2 and 80 electrons, respectively (Singh et al., 2019, VELC team). The photon noise (p) is calculated as the square root of the total photoelectrons generated, including the coronal signal photons obtained in the spectra and the scattered photons. The resulting signal-to-noise ratio (SNR) is then calculated as follows: SNR = S p 2 + D 2 + R 2 , (4) where S is the number of incident signal photoelectrons which in the cases analyzed will be the photons determined from synthetic spectral signal. The SNR is calculated for the peak intensity of the simulated spectra and also at ±0.5 Å from the peak intensity wavelength. It is required to have an idea about the signal strength at ±0.5 Å as it will be helpful to determine sufficient signal requirements near the wings of the spectral line for obtaining a better fit.

VELC onboard Aditya-L1 will provide a unique opportunity to simultaneously image and perform spectroscopic observations of the inner solar corona in three visible and one IR passband. It will be used for spectroscopic diagnostics of the corona up to 1.5 R_ʘ using three emission lines, 5,303, 7,892, and 10,747 Å. It is necessary to simulate the performance of the instrument that will be useful for designing the observation plan after the launch. In this work, we used synthetic spectral data to characterize the spectral channels of VELC for different solar conditions. We synthesized the spectra for the three channels using the CHIANTI atomic database, taking into account the instrument characteristics. We also added the contribution of the instrument, including the scattered intensity and detector noise, to the synthesized spectra. The scattered intensity available for the continuum channel of the instrument was scaled for the spectral channel parameters. The final spectra were then analyzed using the signal-to-noise ratio calculated at the line center wavelength and at ±0.5 Å from the line center.

We simulated the synthetic spectra taking the isothermal condition with average coronal temperature as 10^6.25 K for the three channels and estimated the SNR for the emission lines at coronal heights from 1.1 to 1.5 R_ʘ. For the simulation, the slit width was varied as 20, 40, and 60 μm. We found that on increasing the slit width, the SNR at the peak intensity and at the wings increased for all the three channels. It was identified that for the slit width of 60 μm, the detector for IR channels fills up to 77% of its capacity with 500 ms exposure in the high-gain mode to be used for the spectropolarimetery mode. In order to keep a modest margin for this particular mode, the analysis was done taking the slit width as 50 μm. It was found that a slit width of 50 μm leads to 65% filling of the IR detector in the spectropolarimeter mode, providing a sufficient SNR at the same time. Thus, based on the requirement for this particular mode, we believe that a slit width of 50 μm for the VELC spectral channels will be sufficient to study different regions of the solar atmosphere.

It can be seen from Table 2 that the SNR for 7,892 Å is relatively poor compared to the other two. This is because different lines have different formation temperatures. Therefore, we also studied the effect of different coronal plasma temperatures on the performance of the VELC spectral channels for the optimized slit width. We synthesized the spectra for the three channels at 1.1 R_ʘ, varying the temperature from log(T) = 6.0 to log(T) = 6.5 in steps of 0.1. We found that the SNR at the line peak varies with the change in temperature such that the maximum SNR observed for 5,303, 7,892, and 10,747 Å was at log(T) of 6.3, 6.1, and 6.2, respectively. These temperatures are close to their line formation temperatures. We could observe a similar increase in the SNR for the wings of the synthesized spectral lines. We analyzed the instrument’s performance with a slit width of 50 μm to synthesize the spectra with their line formation temperatures. The SNR was then estimated at the line peak and the wings for coronal heights from 1.1 to 1.5 R_ʘ. From the results of this analysis presented in Table 4, it could be observed that there is a sufficient SNR in all the channels for the line peak and the wings, which decreases at larger heights. The IR channel has a very good SNR even at larger heights due to its bigger pixel size compared to the other two.

As these analyses were based on isothermal coronal conditions, we also used an MCMC simulation taking a Gaussian distribution of temperature peaking at 10^6.1 K having a width of 10^0.3 K to estimate the expected signal at different heights in the VELC FOV. We found that there is a spread in the estimated counts based on the temperature distribution, with a cluster obtained corresponding to the temperatures close to the line formation temperature for individual channels (Figure 3). Overall analysis shows that for the optimized slit width of 50 μm and considering the peak line formation temperatures of each line, a reliable signal could be obtained for all the channels which could be used for analysis. For the larger heights, when the SNR becomes ≤ 5, pixel binning could be considered to enhance the signal with respect to the background. There is also an option to increase the exposure time without saturating the detector, which could also boost the signal. For the study of very fast transients requiring short exposure times, the observations could be taken at a high cadence. Such short exposure frames could be further binned to increase the SNR.

We also performed a study of a CME case approximating the enhanced electron density due to a CME of 2016-01-01 from KCor images, which has a similar FOV as VELC. We found that the visible spectral channels of VELC could be operated for CME observations with an exposure time more than 1 s, whereas the IR channel will be operated in the low-gain mode following the trigger provided by the onboard CME detection algorithm with a reduced exposure time. More CME cases could be analyzed in the future with different enhancements that could help in planning for spectral diagnostics of CMEs using VELC. We also considered two specific scientific cases, magnetometry and Doppler mapping, with regards to the SNR requirement from a VELC point of view. For the measurement of the magnetic field using the IR channel, we calculated the Stokes-V intensity, thereby estimating the minimum integration time required. For the Doppler mapping, we used the green line at a height of 1.4 R_ʘ, where SNR ≈ 5. We also considered a case where the intensity of the line is 20% of the case mentioned above. For the two cases, we added the noise and a Doppler speed, which were measured later by fitting a Gaussian profile. We found that a minimum SNR of 5 will be required for Doppler mapping using VELC.

It should be noted that increasing the slit width results in the broadening of the spectral lines and hence a reduction in the spectral resolution. Singh et al. (2019) studied the effect of increasing the slit width for the three spectral channels of VELC. Their analysis also indicates that the slit width should be increased to enhance the SNR at the same time, optimizing the spectral resolution to meet the scientific requirements. These estimates will be helpful in identifying the expected observed spectra including the instrument contribution, which could be deconvolved during the processing to get the true line profile. The effect of spacecraft drift and jitter as studied by Ranganathan and Sankarasubramanian (2019) has not been considered for the studies presented here but will be included in our future work. Such complete spectral information will be helpful for the data pipeline development and extracting the signal from the background.

In this work, we have presented a few of the solar conditions. A similar study could be extended to other solar features such as loops, plumes, and coronal holes which result in different ambient coronal conditions. Such extensive studies covering different cases could help in preparing the optimized plan for maximizing the science output from the instrument. It should be noted that this study could also be extended to future missions, which include spectrographs for preparing the target science cases that could be addressed with the instrument capabilities.