High-resolution solar imaging for citizen science

Abstract. High-resolution solar imaging, essential for research into the Sun is usually limited to high-end expensive telescopes found in professional observatories using adaptive optics technology. However, a Schmidt-Cassegrain Telescope (SCT) or a refractor telescope equipped with a narrow-band energy rejection filter and a suitable solar spectrum filter, employing appropriate data acquisition and post-processing techniques can provide an amateur astronomer a costeffective means to high-resolution solar imaging.


Introduction
In the last decade, students in Singapore schools have produced solar images in Hydrogen-Alpha, Calcium-K and White-light wavelengths. There was a keen interest to produce high-resolution solar images of the Sun for research work [1].
Students in several schools in Singapore conducted extensive tests from 2016 to 2019 to customise different solar telescope configurations and build a cost-effective observatory for producing high-resolution solar images [2,3]. National Junior College used the findings of the trials and built a High-Resolution Solar High Observatory (HISRO). Following its commissioning in July 2019, students from NJC and other participating schools have used HISRO for their research work (See Figure 1). This project used D-ERFs with diameter ranging from 165mm to 356mm to enable the host telescope to retain its full-aperture resolving power while reducing the thermal energy going through the telescope to the solar spectrum filter.  The D-ERF reduced the heat stress within the telescope and also reduced the thermally-induced seeing effects as well. A single-band H-Alpha 165mm D-ERF was used on a 150mm refractor telescope. A 295mm D-ERF (for 279mm SCT) and 214mm D-ERF (for 203mm SCT) were used for tri-band (H-Alpha, Calcium-k and G-Band) imaging. A 365mm D-ERF was used for H-Alpha imaging with a 356mm SCT. Typical H-Alpha tranmission and Calcium-K transmission were about 90% and 60% respectively (See Figure 3). "Leakage" into the unwanted wavlength was typically under 5%. Typical SCTs, with SCT aperture ranging from 203mm to 356mm and a 150mm refractor were used as host telescopes.

Solar Spectrum Filter [5, 6]
The project teams tested a range of commerically available etalons and integrated them into donor telescopes fitted with full-aperture D-ERF to form different customised solar telescope configurations. Favourite components included a solid mica etalon and air-spaced etalons without central obstruction for both H-Alpha and Calcium-k imaging. Other components included G-Band filter, H-Alpha blocking and Calcium-K module (See Figure 4).

Astro camera with barlow or reducer.
An astro CCD camera with a high acquisition speed was used for high-resolution solar imaging. To achieve the required imaging scale the necessary barlow or reducer was added before the astro camera. To provide smooth tracking of the Sun, a typical German Equatorial Mount with polar alignment was used to carry the solar telescope and imaging train. To have a dark workspace in the field, the team constructed a field imaging instrumentation console. The enclosure shielded the recording equipment from sunlight and enabled easy video acquisition in the field [7].

Open field observatory [8]
NJC High-resolution Solar Research Observatory has four imaging mounts holding three SCTs and one refractor telescope for high-resolution solar imaging. It was built with the following considerations:  Open concept. It was located at a site for easy access for students to conduct solar observation and research work. (See Figure 5)  Imaging environment. Seeing conditions affect the quality of imaging. Heated air from the ground will produce turbulence and affect the quality of the imaging. The team selected a site surrounded by greenery and situated on a relatively high ground.  Setup time. Students attend lessons during daytime and they need to commence imaging as soon as possible when the seeing conditions permit highresolution imaging. HISRO was able to commence imaging within 5 minutes of reaching the observatory.
The team achieved the above-mentioned requirements by building HISRO at a location where students will walk past in between lectures. The telescopes were left permanently on the mounts and protected from the elements using a newly devised 3layer covering method (See Figure 6). Students are able to commence imaging after a 5-minute setup time. They have to remove the protective covers, deploy the field imaging instrumentation console and power up the telescopes before commencing the imaging session.

Workflow
Atmospheric turbulence randomly distorted the wavefronts emanating from the Sun and was the primary obstacle in achieving diffraction-limited resolution in high-resolution solar imaging. The following workflow was employed to reduce the adverse effect of atmospheric turbulence.

Imaging window [9]
The best quality images were acquired in the imaging window from 8:15 am to 10:30 am local time, throughout the year. This early window of opportunity will avoid jet stream emanating from the heated ground and rooftop. Nevertheless, the observatory was able to produce images of decent quality during the other times of the day due to the general good seeing condition in Singapore. [10] The camera has to be set up for each telescope configuration for optimum performance. To obtain individual captured frames with the best signal/noise ratio, the researcher needs to set the camera at the fastest shutter speeds with the least gain (brightness amplification) for least noise in the final image. Exposure was typically set below 5 msec to freeze the seeing. The capture rate was usually maintained close to 100 fps to prevent a long duration capture leading to changes in the solar features. Gamma setting was set to provide good image contrast to aid the focussing of the image. During capture, gamma moderation was necessary to ensure that subtle solar details within  [11,12] Before image restoration techniques were applied, the frames captured in the most favourable atmospheric conditions (lucky frames) were selected using solar image processing software. During good seeing, up to 1,000 best frames were selected and stacked by the same solar image processing software to form a single resulting image with a high signal-to-noise ratio. There was a need to experiment and try stacking different number of selected frames to find the best resulting image.

Post-processing [9]
As part of image restoration, the final stacked image was further processed using Lucy Richardson deconvolution and unsharp masking techniques to produce a sharp and yet not overly sharpened high-resolution image revealing the fine features in the captured region of the Sun. The post-processed image could be further enhanced with an image editing software.   Calcium-K and G-band) full-aperture D-ERF, a commercial G-Band (filter passband 10 nm) and x5 barlow. (See Figure 10) With good seeing, the imaging rig, with an image scale of 0.23 arc sec/pixel reached the diffraction limit for this wavelength and this telescope size.
Many amateur astronomers were amazed by what a typical 203mm SCT could achieve. They celebrated this achievement and compared it favourably with a 1.6m specialized solar telescope equipped with advanced adaptive optical system at the National Solar Observatory. This breakthrough sparked off a lively discussion amongst amateur astronomers [14]. Figure 11 shows a G-Band image of twin sunspots in a photospheric region. This image was taken later in the day after the Figure 9 image. The seeing condition had deteriorated and thin clouds veiled the Sun. The gain on the camera was stepped up and a more noisy image resulted. These 2 images (i.e. Figure 9 versus Figure 11) showed the vital part seeing condition played in the eventual quality of image. Subsequent trials showed that a bigger image scale should be used to improve the image quality when seeing has gone downhill.

Survey of the upper chromosphere during solar minima.
A 279mm SCT fitted with a tri-band (H-Alpha, Calcium-K and G-band) fullaperture D-ERF, a solid mica H-Alpha etalon and x0.8 reducer were used to survey the Sun's upper chromosphere in H-Alpha wavelength from January 2020 and February 2020 (See Figure 12). Different sized reducers (e.g. x 0.5 to x0.8 reducers) with different length spacers were used to achieve different image scale ranging from 0.15 to 0.2 arcsec per pixel. Higher image scales were used in poorer seeing conditions. Some of the collected images in Figures 13, 14 and 15 revealed fine fibril details, magnetic field topology and the dynamic nature of a quiet solar chromosphere [15]. As can be observed in Figure 14 and the evolution of fibrils could be traced. A boiling chromosphere was observed in an animated gif from a 5 minute recording session [16].

Survey of the lower solar chromosphere
The same 279mm SCT fitted with a tri-band (H-Alpha, Calcium-K and G-band) full-aperture D-ERF was fitted with a calcium-K wedge for surveying the lower solar chromosphere. (See Figure 16). With a x2 barlow, an image scale of 0.22arcsec per pixel was achieved.
The chromospheric network as observed with this configuration are shown in Figure  17 and Figure 18 [17][18][19].

Fibrils
The same 279mm SCT shown in Figure 1 was fitted with a tri-band (H-Alpha, Calcium-K and G-band) full-aperture D-ERF and an air-spaced etalon without central obstruction to image the dark elongated fibril structures in H-Alpha wavelength [20]. Dark structures were observed in the penumbra of a sunspot, radically expanding around the umbra and creating filamentary structure in the active region (See Figure  19).

Conclusion
The different solar telescopes, when used with appropriate data acquisition, postprocessing and image restoration workflow, showed great potential as cost-effective solutions for amateurs to do high-resolution solar research. More work was needed to overcome atmospheric seeing when imaging with a 356mm SCT and improving image restoration work processes to produce better quality images [21].