Impact of Space Weather on the Natural Night Sky

The natural night sky is animated with its own sources of illumination. This fact has been largely overshadowed by the important effort to prevent unwise artificial nighttime lighting from blinding humans to the wonders of the universe.

There is a developing body of literature detailing the effects of artificial-light pollution on humans and ecology. Hänel et al. (2018) provide an extensive review of instruments and methods employed to measure the night sky. Their stated goal is to be a reference for interested parties in the health, ecology, and other research fields who have no experience in measuring and interpreting night sky brightness data. Tables 2 and 3 in their paper provide a summary of both natural and human caused sources of night sky brightness. They also provide an overview of measurements which can be made by single channel, imaging, and spectrophotometric instruments. No attempt is made to describe changes in the natural night sky over time.

Cinzano et al. (2001) produced "The first World Atlas of the artificial night sky brightness" which depicts in graphic detail the shocking way humans have obliterated the possibility to view the natural night sky for many of our planet's residents. Falchi et al. (2016) provide a significant improvement to the Cinzano et al. (2001) Atlas, a graphic visualization tool, and access to measures of light pollution worldwide.

The Sky Quality Meter (SQM) manufactured by Unihedron of Canada plays a prominent role in the light pollution monitoring effort. To date there have been more than 13,700 Unihedron SQM meters sold worldwide. Most of them are used to provide calibrated measures of the brightness of the night sky. These data have proved valuable in understanding the extent of human caused light pollution.

Duriscoe (2013) models sources of natural night sky brightness to characterize the effects of human caused light pollution at national parks and other remote locations. The detailed natural night sky model, described in this reference, indicates the natural night sky ranges from dark (Bortle 1, 21.7 to 22.0 mag arcsec⁻²) to one with moderate light pollution (Bortle 4, 20.4–21.3 mag arcsec⁻²) without human intervention. By subtracting this natural night sky model from calibrated all sky mosaic images Duriscoe (2013) is able to accurately locate and measure sources of artificial-light pollution.

There are additional references to the fact that the natural night sky is not a dark vault of constant brightness as is idealized by Bortle (2001). Pilachowski et al. (1989) obtained data on 12 nights at Kitt Peak National Observatory during and following a solar minimum in 1986, 1987, and 1988. Krisciunas (1997) produced a database of single-channel photometric measurements of the natural night sky on Maunakea over a whole solar cycle. His yearly averages ranged from 21.906 to 21.287 mag arcsec⁻². He indicates that the scatter in the individual data points is real and dominated by short term variations on timescales of tens of minutes.

Benn & Ellison (2007) measured the night sky brightness from 427 scientifically calibrated CCD images taken with the Isaac Newton and Jacobus Kapteyn Telescopes on La Palma on 63 nights, 1987–1996. They find the night sky is brighter at low ecliptic latitude (0.4 mag) as well as at solar maximum (0.4 mag) when compared with high galactic latitudes and solar minimum conditions, respectively.

Patat (2003, 2008) measured the night sky brightness at Cerro Paranal in Chile during 650 nights, 2001–2006, using more than 10,000 VLT-FORS1 archival images. His data reveal significant seasonal as well as solar activity induced night sky brightness variations (21.02 < V < 22.30).

Neugent & Massey (2010) published new absolute spectrophotometry of the Kitt Peak night sky obtained in 2009–2010 during a deep solar minimum. By combining these data with previously published measurements they were able to conclude that the night sky over Kitt Peak had remained relatively constant over a 20 yr period. However, their zenith measurements 2009 October 17.3 UT, at an R.A. = 1^h17^m28^s, was V = 21.51 mag arcsec⁻² while on 2010 June 14.3 UT, at an R.A. = 17^h12^m41^s, was V = 22.36 mag arcsec⁻² document that the natural night sky goes through significant changes in brightness.

An SQM-LU-DL, pointed at the zenith, in a natural night sky location, records the sum total of the flux of photons from celestial sources, as well as emission from terrestrial atmosphere (airglow). The SQM-LU-DL has a lens which establishes its 20° full width at half maximum (FWHM) field of view and independently logs up to a million readings (Tekatch 2019). The rotation of Earth on its axis causes natural night sky sources to pass overhead in a parade of illuminated patterns. The celestial sources include stars, scattered starlight, diffuse light from galaxies and nebulae, as well as the Zodiacal light produced by sunlight scattered off dust in the solar system. Leinert et al. (1998) provides a tabulated catalog of sources of diffuse night sky brightness.

The spectral response of the SQM-LU-DL is set by the TSL237 photodiode detector, Hoya CM-500 filter, and weather proof housing window combination. Sánchez de Miguel et al. (2017)'s Figure 4 presents both the measured and expected (from the components) spectral response. The SQM peak response is at approximately 527 nanometers with a FWHM of approximately 215 nanometers.

Figure 2, Hänel et al. (2018) provides a comparison of the spectral responses of the SQM, Photopic vision, Johnson V, DSLR G, and a reference light meter.

SQM-LU-DL instruments have been employed in a large number of sky brightness studies (see, Hänel et al. (2018)) to obtain absolute sky brightness photometry in the sense that the measurements made with one instrument are compared with those obtained with similar, calibrated, instruments.

The Physics of the SQM-LU-DL detector suggest that it is linear down to very low light levels. A solid state charge bucket internal to the TSL237 detector accumulates charge as photons enter it. When the bucket is full the device issues a pulse. To measure the incident photon flux, the average of the time between 8 pulses is determined and used to calculate the brightness in mag arcsec⁻². If there is no incident light, dark current fills the well, and eventually (158.889 s for the CCIDSS detector) a pulse is issued. Since the rate at which the dark current fills the well depends on the temperature, it is necessary to correct for this phenomenon.

Cinzano (2005) measured linearity of an SQM by analyzing the residuals of a comparison with the IL1700 Research Radiometer illuminated by a LPLAB Variable Low-light Calibration Standard (Cinzano 2003). Over a range of surface brightness of magnitudes he concludes that the deviations from linearity are less than 2.6%. This limit is set by the stability of the standard source (1%) and the linearity of the reference radiometer (1%) which were employed to conduct the tests.

Pravettoni et al. (2016) used modified testing equipment in an ISO 17025 accredited laboratory to assess the performance of 16 Unihedron SQMs in weather proof housings. These authors's linearity tests covered the urban light pollution range from 6 to 19 mag arcsec⁻², yielding standard deviations from linear regressions ranging from 0.11 to 0.14 mag arcsec⁻². The accompanying graphs indicate the largest deviations they measured occurred between 6 and 13 mag arcsec⁻². These limits may not be at all applicable to differential photometry in the 22–20 mag arcsec⁻² levels of illumination commonly found in the natural night sky. Also, Sánchez de Miguel et al. (2017) found that slight alignment errors in the optics of an SQM effects its angular response compared with other identical appearing instruments.

The Unihedron SQM-LU-DL detector (TAOS TSL237S) has a dark current which is temperature dependent. The SQM-LU-DL's internal software corrects for this phenomenon by using the temperature measured right behind the detector to correct the observed light flux.

Schnitt et al. (2013) tested a pair of SQM-LU meters, one inside and the other outside of a Unihedron weather proof housing. Both units were placed in the same laboratory temperature controlled chamber. This experiment found that the SQM-LU radiance measures remained with 7% of its initial value over a range from −15°C to 35°C with rates of temperature change from −33°C hr⁻¹ to +70°C hr⁻¹. The 7%, they obtained, should be regarded as an upper limit for the accuracy of the temperature correction of the detector since shifts in the optical alignment could not be ruled out as a significant source of error in this laboratory experiment. It is not clear, what if any, significant, potential detector temperature correction errors bias our measurements. Sources of error in the Schnitt et al. (2013)'s laboratory experiments and the fact that the temperatures and temperature rates of change encountered in the field were much smaller than the Schnitt et al. (2013) laboratory setup suggest that any actual temperature compensation errors are likely to be substantially less than the 7% laboratory result.

Barentine et al. (2018) report that in the field three SQM-L meters agreed with each other to on the order of 0.1 mag (10%). Unihedron, the manufacturer, states an accuracy of ±10% for an SQM-LU-DL when it is used to perform absolute photometry but adds that differential photometric measurements made with the same meter are significantly better than this figure (A. Tekatch 2018, private communication).

This paper presents new methods of analyzing SQM data in which the device is used to make single-channel differential photometric measurements of selected locations on the celestial sphere. These techniques are employed to process data from Cosmic Campground International Dark Sky Sanctuary (CCIDSS), under a special use permit from the Gila National Forest, Glenwood Ranger District in New Mexico, and Aotea/Great Barrier Island International Dark Sky Sanctuary (AGBIIDSS) in New Zealand.

Cosmic Campground is a 3.5 acre site in the Gila National Forest of western New Mexico, USA, managed by the United States Forest Service. In 2016 January it was accredited as the second International Dark Sky Sanctuary by the International Dark Sky Association recognizing the fact that CCIDSS is located in an exceptionally dark area with no permanent artificial lighting. The nearest significant source of light is more than 65 km away.

Aotea/Great Barrier Island is located about 100 km from Auckland, New Zealand. In 2017 June, it was accredited as the third International Dark Sky Sanctuary by the International Dark Sky Association recognizing the fact that AGBIIDSS has retained essentially all of its natural nighttime darkness due to the fact that nearly 60% of its 265 km² is managed by New Zealand Department of Conservation as a nature reserve, the total absence of utility electricity on the island, and minimal development pressure from its small population of 939 persons.

The International Dark Sky Sanctuary status enjoyed by CCIDSS and AGBIIDSS are a testament to the pristine quality of their night skies and nocturnal environment. This status has been certified by making night sky brightness SQM measurements.

The data used for this project gathered from CCIDSS and AGBIIDSS are that of natural night skies free of artificial-light pollution.

The measurements reported in this paper are made with SQM-LU-DL, narrow field of view, solar (CCIDSS) or battery (AGBIIDSS) powered, sky-quality meters, in weather proof housings, manufactured by Unihedron. This equipment was installed at the AGBIIDSS 2017 July 17 at 149 Sanhills Road, Medlands, on Ateoa/Great Barrier Island in Aukland, New Zealand. The instrument package, installed at the CCIDSS on 2018 July 27 was designed and built to be weather proof, varmint/bug resistant, and safe for unattended operation in a public location. A secure port 2 feet above ground allows access via USB to download data. This arrangement has the advantage of preserving the alignment of the meter inside the weather proof housing since it remains in place during the data read out process.

At both locations the instrument is pointed at zenith. As Earth rotates on its axis, each meter scans a strip of constant decl. returning to the same point on the celestial sphere 4 minutes earlier each day. These facts are employed to use the SQM-LU-DL as a single-channel differential photometer to measure changes in brightness, of the same place on the celestial sphere, with the same instrument.

For the AGBIIDSS data set, 2018 March 26–August 31, the average temperature was 11.21°C (standard deviation = 3.14°C, median = 10.30°C). The average nightly delta T (twilight to twilight) was 1.47°C and average slope was −0.32°C hr⁻¹.

For the CCIDSS data set, 2018 September 4–2019 January 4, the average temperature was 3.34°C (standard deviation = 6.82°C, median = 1.90°C). The average nightly delta T was 7.42°C (twilight to twilight) with average slope −0.77°C hr⁻¹.

The SQM-LU-DL Serial Number 4074, installed at the CCIDSS was calibrated at Unihedron on 2018 April 23 and had a light calibration offset of 19.94 mag arcsec⁻². Anthony Tekatch of Unihedron, extensively retested this unit 2018 July. The first calibration light offset was 0.01 mag brighter than in April, however, the subsequent July retesting repeatedly produced exactly the same light calibration offset of 19.94 mag arcsec⁻² previously obtained. The "light calibration offset" is a value that makes the SQM deliver the reference value (8.71 mag arcsec⁻²) when a reference light source (of 8.71 mag arcsec⁻²) is used. Tekatch's tests indicate that the instrument's zero-point is very stable.

To further test this unit, Tekatch placed it in Unihedron's dark room illuminated by a test lamp placed at a low stable setting. This test has 6272 data points, which are plotted in Figure 1. Tekatch reports outlying points 21.45 mag arcsec⁻² (27 points in total of these brightest outliers) and 21.49 mag arcsec⁻² (3 points total of these darkest outliers) (currently unexplained), but it is quite difficult for him to exclude all possible extraneous light sources. For more than 6 hr the readings were stable between 21.47 and 21.48 mag arcsec⁻².

The cloud detection method, introduced later in this paper, requires the SQM-LU-DL data, fitted to straight line segments, (mag arcsec⁻² versus time) have a Chi-Squared of less than 0.009 for a period of at least 1.5 hr. Anything smaller than the 0.009 value rejects the rising Milky Way (See Figures 14(a) and (b)). These results along with those of Cinzano (2005) imply that the differential photometry of an SQM between adjacent sky positions is likely to be significantly better than 0.03 mag arcsec⁻². These results are consistent with the 0.02–0.05 mag arcsec⁻² experienced by other SQM users in the field (J. C. Barentine 2019, private communication).

Typically, the SQM-LU-DL at the CCIDSS and AGBIIDSS are programmed to record a sky measurement every 5 minutes for the portion of the day/night cycle during which the Sun is more than about 5° below the horizon (sky fainter than 10.0 mag arcsec⁻²). Unihedron provides a complete software package to program the device as well as to download and visualize the data. Unattended, the SQM-LU-DL can record such data for months. Each recorded data set consists of a header followed by individual data points. Each data point consists of UT date and time, local date and time, temperature in Celsius, battery voltage, the recorded mag arcsec⁻², and a reference tag.

Zoom In Zoom Out Reset image size

Examples of sky brightness versus time data are shown in Figures 2 through 6 for the nights of 2018 September 10, 8, 2018 August 12, 10, and 2018 April 11. On clear nights the data trace is smooth and the passage of the Milky Way and Zodiacal Light overhead can be easily seen (Figure 2). Figure 3 shows that under a natural night sky the passage of clouds causes dips in brightness. Figure 4 shows the passage of an intense band of thunder storms followed by a clear sky. Figure 5 shows an initially clear night during which clouds and fog moved in causing dips in brightness. Figure 6 shows a mostly clear night punctuated by a occasional tiny clouds. In the absence of moonlight clouds cause dips in brightness while Moon illuminated clouds cause increases in night sky brightness measurements.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

An SQM-LU-DL, pointed at the zenith, under a natural night sky, records the sum total of the flux of photons from celestial sources as well as emission from the terrestrial atmosphere. The rotation of Earth on its axis cause the celestial sources to pass overhead in a parade of illuminated patterns. Celestial sources include planets, stars, scattered starlight, diffuse light from galaxies and nebulae, as well as the Zodiacal light produced by sunlight scattered by interplanetary dust in the solar system. Terrestrial sources include airglow produced by solar photoionization and recombination, cosmic ray induced light flashes, and other processes.

The purpose of the data analysis software written in FORTRAN is to select those data points from SQM-LU-DL's data stream obtained when the Sun is more than 18° below the horizon (formulae, Schlyter 2019), the Moon is more than 10° below the horizon (formulae Meeus 1991), and the sky is free of clouds.

The 10° Moon limit is necessitated by large changes in lunar brightness as the Moon passes through its phases every month. On photometric nights, at the Cosmic Campground IDSS measurements show that if the Moon is less than 74% illuminated it provides negligible light at zenith when it is 10° or more below the horizon. The situation changes rapidly when the Moon approaches the horizon. For example, a 19% illuminated Moon, 5° below the horizon, increases the zenith sky brightness by 0.02 mag arcsec⁻². The Moon's zenith illumination, when it is still below the horizon, could be effected by moisture and/or dust in the atmosphere so it is better to be conservative in selecting a value for this parameter.

The program calculates the LST of each point on the celestial sphere. These values are used to sort the data into 48 bins of $\tfrac{1}{2}$ hr in R.A. Each of these bins represents approximately 1/3 of the detector's pixel. This binning preserves the achievable resolution of celestial objects as they pass overhead and allows sky brightness statistics to be calculated for each location on the celestial sphere.

The Fredericksburg Solar A and K Indices (NOAA 2019) are appended to each data point. They provide a measure of geomagnetic activity when the data point was obtained.

For each data point the program performs a linear least squares fit to the SQM brightness versus time data for 45 minutes on either side of each measurement. The resulting Chi-Squared, measures the smoothness of the curve. Clouds produce ragged traces and the photometric natural night sky produces a very smooth curve. A Chi-Squared of less than 0.009 rejects clouds but not the rising Milky Way. See Figures 14(a) and (b) for plots of Chi-Squared versus time on clear nights. Episodes of cloudy weather are rejected by analyzing the smoothness and shape of the data traces and by consulting weather records.

The cloud detection algorithm works well to select valid points from clear skies and to reject data from skies with non-homogeneous clouds. It is less effective at rejecting data obtained through thin, extremely smooth layers of clouds, dust, or haze, which could absorb perhaps 10% or less of the incoming light. To address these conditions weather data and inspection of nightly data traces are employed. Nights of fog and/or thick clouds with readings fainter than 22.3 mag arcsec⁻² can be rejected since as we will show, in the next section, to date, the faintest natural night skies, measured under clear photometric conditions, are brighter than 22 mag arcsec⁻². We continue to try to measure the faint natural night sky brightness limit.

The data points for which the Sun is more than 18° below the horizon, the Moon is more than 10° below the horizon, and there are no clouds as indicated by the smoothness of the SQM data versus time are written to a csv file.

Figures 7 and 8 present data obtained on partly cloudy nights. Short segments of these data traces were flagged as "clear" by the software. The nights data viewed in total and weather records called these data into question and the whole night was rejected. This procedure was repeated on other problematic nights.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Unavoidable gaps in the data caused by the Sun, Moon, and clouds complicate the analysis.

It should be emphasized that in this research project, the SQM-LU-DL is used as a single-channel differential photometer to determine temporal brightness changes in the same place on the celestial sphere with the same instrument. The SQM-LU-DL is much more accurate in this mode compared with the one in which it is used to make absolute calibrated photometric measurements of the night sky.

5.1. Night Sky Light Curves

The 1,231 AGBIIDSS data points obtained on portions of 40 nights from March 26 through 2018 August 31 are shown in Figure 9. Unfortunately, bad weather during 2018 September, October, and November rendered useless the data obtained during this time.

Zoom In Zoom Out Reset image size

The CCIDSS 3,555 data points obtained on all or portions of 63 nights from 2018 September 4 to 2019 January 4 are shown in Figure 10. Unfortunately, bad weather during 2018 July and August rendered useless the data obtained during this time.

Zoom In Zoom Out Reset image size

A striking feature of Figures 9 and 10 is the smooth, well defined, curve of the lowest night sky brightness points versus R.A. The data points along these curves of minimum brightness versus R.A. are likely to represent the sum of the photon flux from celestial sources and airglow, when the Sun is in a quiescent state. Figures 11 and 12 present these data for each binned location on the celestial sphere. These data provide a baseline for measuring changes in the natural night sky.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

During 2018 the natural night sky was measured to range in brightness from 21.99 to 21.13 mag arcsec⁻² at Cosmic Campground International Dark Sky Sanctuary and from 21.99 to 21.14 mag arcsec⁻² at Aotea/Great Barrier Island International Dark Sky Sanctuary. The accuracy of brightest measurements at both locations is set by each instrument's calibration while the faintest measurements should be regarded as upper limits. The darkest readings reported were selected using weather forecast information and the smoothness of the nightly light curves to verify photometric conditions.

The data in Figures 11 and 12 permit one to measure changes in airglow above its quiescent value.

On Figure 13 the circles labeled 1, 2, and 3 are meant to depict the detector's smoothly advancing direction of view on the celestial sphere as Earth rotates on its axis. At both the CCIDSS and AGBIIDSS at position (and time) 1 the signal is increasing with contributions from both the Milky Way and the Zodiacal light. Position (and time) 2 represents the detector's field of view crossing the galactic plane. Position 3 (and time) represents the point at which the combined signal from the Milky Way and the Zodiacal light is declining. The shapes of light curves in Figures 11 and 12 are produced by contributions from both the Milky Way and the Zodiacal Light. This result illustrates the sensitivity of our detectors and analysis techniques.

Zoom In Zoom Out Reset image size

A second feature of Figures 9 and 10 is the large range of brightnesses measured at each sky position on the Celestial Sphere. These data were taken while the Sun was approaching a deep minimum in the solar cycle. Published data suggests during a solar minimum night sky brightness should be between 0.40 and 0.65 mag arcsec⁻² fainter than at solar maximum (Benn & Ellison 2007) (Krisciunas 1997). This phenomenon is likely the result of changes in solar daytime photoionization (followed by nighttime recombination) over the course of a solar cycle. The variation in night sky brightness over a complete solar cycle is documented by Krisciunas (1997) in his Figure 1, which plots yearly averages versus time. However, in the same paper, Figure 5, shows rapid changes in night sky brightness on a timescale of hours.

To begin to investigate the sources of sky brightness changes revealed in Figure 10 at Cosmic Campground IDSS, two consecutive exceptionally clear nights 2018 September 8–10 were identified. Figures 14(a), and (b) are plots of the Chi-Squared for the two nights from the cloud detection algorithm. These two plots show that on clear nights the SQM-LU-DL delivers an exceptionally low noise signal. Figures 15(a) and (b) are the light curves obtained on the two nights. The shape of the light curves on both nights is very similar. The sky brightness on the two nights begin to diverge due to a steady increase in the airglow which is contemporaneous with an approaching geomagnetic disturbance.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

5.2. Night Sky Brightness And Geomagnetic Events

The purpose of this section is to investigate night sky brightness increases which are contemporaneous with geomagnetic events.

During solar minimum photoionization in Earth's atmosphere may be reduced, however, holes in the solar corona produce intense bursts of charged particles. When this enhanced solar wind interacts with Earth's magnetic field it produces geomagnetic disturbances, storms and auroras. The Space Weather Prediction Center (NOAA 2019) archives geomagnetic and solar data. These records include Fredericksburg daily mid-latitude and planetary A and K indices going back to 1994. Additionally, the the Potsdam Indices of Global Geomagnetic Activity (GFZ 2019) is a valuable source of geomagnetic data.

The Ap Index (scale 0–400) is a linear scale while the Kp Index (scale 0–9) is a quasi-logarithmic scale. "Ap is defined as the earliest occurring maximum 24 hr value obtained by computing an 8-point running average of successive 3 hr ap indices during a geomagnetic storm event without regard to the starting and ending times of the UT-day. It is uniquely associated with the storm event." (Allen 2004).

For 2018 the average daily difference between the Fredericksburg Ap and the Potsdam Ap is 1.02 (stdev = 1.22) indicating good agreement between these two sources. Alvestad (2019a, 2019b) maintains the "Solar Terrestrial Activity Report" which contains a compilation of current and archival solar and geomagnetic data in a very valuable format.

In the plots to follow the GEOS data (NOAA 2019), Solar and Heliospheric Observatory (SOHO)/DSCOVR data (Alvestad 2019b), Potsdam Ap data (GFZ 2019) are compared with SQM-LU-DL data from Cosmic Campground IDSS and Aotea/Great Barrier Island IDSS.

Data recorded at Cosmic Campground IDSS, in the fall of 2018, revealed that increases in night sky brightness are contemporaneous with increases in geomagnetic activity. Figures 16(a) and (b) were created to explore this observation. Episodes of increased night sky brightness are contemporaneous with disturbances in Earth's magnetic field.

Zoom In Zoom Out Reset image size

Figures 17(a)–(c) show individual night sky brightening and the contemporaneous geomagnetic disturbances at Cosmic Campground IDSS.

Zoom In Zoom Out Reset image size

It is interesting to note the maximum increase in night sky brightness occurs after the peak of the geomagnetic event.

Figures 18(a) and (b) show two of the individual night sky brightening events and the contemporaneous geomagnetic disturbances observed at Aotea/Great Barrier Island IDSS. A third such event at JD 2458361.3 days had too little data to plot as an event (see Figures 21(c) and (d)). The data do not reveal a simple cause and effect relationship between increases in night sky brightness and space weather conditions. See Section 6.

Zoom In Zoom Out Reset image size

5.3. Night Sky Brightness And Solar Activity

Figures 19(a)–(d) are presented to explore the relationship between elements of space weather and night sky brightness.

Zoom In Zoom Out Reset image size

Figure 19(a) is a plot of the 10.7 cm (2.8 GHz) solar flux and night sky brightness measurements at Cosmic Campground IDSS relative to quiescent solar values of Figure 12. The 2018 natural night sky had brightening events even though the 10.7 cm solar flux was at a low steady level.

Figure 19(b) is a plot of the SOHO/DSCOVR mean solar wind speed and night sky brightness measurements at the Cosmic Campground IDSS. The brightness values plotted are relative to quiescent solar values of Figure 12. Brightening of the natural night sky is contemporaneous with Earth entering a stream in the solar wind.

Figure 19(c) is a plot of solar electron fluence and night sky brightness measurements at Cosmic Campground IDSS. Brightness increases plotted are relative to quiescent solar values of Figure 12. The largest increases in night sky brightness are contemporaneous with bursts of solar wind electrons at 1 au.

Figure 19(d) is a plot of the solar proton fluence and night sky brightness measurements at Cosmic Campground IDSS. Brightness increases plotted are relative to quiescent solar values of Figure 12. The largest increases in night sky brightness are contemporaneous with bursts of solar wind protons at 1 au.

Figures 20(a)–(f) explore individual contemporaneous increases in night sky brightness and GEOS solar proton fluence, and SOHO/DSCOVR mean solar wind speed at both Cosmic Campground IDSS and Aotea/Great Barrier Island IDSS.

Zoom In Zoom Out Reset image size

The data for Figures 20(a) and (b) were obtained at Aotea/Great Barrier Island IDSS in 2018 late March. The night sky brightening event was contemporaneous with only minor changes in SOHO/DSCOVR mean solar wind speed, GEOS solar proton fluence, and geomagnetic disturbances (see also Figures 21(c) and (d).

Zoom In Zoom Out Reset image size

Data presented Figures 20(c) and (d) represent a relatively simple situation in which Earth enters and exits a stream in the solar wind. Data were obtained in 2018 October at Cosmic Campground IDSS.

2018 November Earth entered a series of solar wind streams of varying intensities. Figure 17(c) shows a very unstable period in Earth's magnetic field. During this period Earth entered a minor stream in the solar wind; had a lull in geomagnetic activity; experienced G1 and G2 Geomagnetic Storms; and experienced a lull in geomagnetic activity followed by an explosion in polar auroras. The data ends with Earth's exit from the series of streams in the solar wind. This sequence of events is described by the spaceweather.com archives Phillips (2019). These data show that Earth's interaction with the solar wind can be very complicated (see also Figures 17(c), 20(e) and (f)).

5.4. Summary Plots

Despite our instrument's dusk to dawn coverage every night, the observations were, unavoidably, interrupted by the Sun, Moon, and clouds. In fact, clear astronomical dark conditions accounted for only 10.1% of the total number of possible 5 minute time intervals at Cosmic Campground IDSS (2018 September 4–2019 January 4) and for only 2.7% of the possible 5 minute time intervals at Aotea/Great Barrier Island IDSS (2018 March 26–August 31). Thus, our data must be regarded as a small sample of the Space Weather's interaction with near-Earth environment.

Borovsky (2018) states in his abstract "It is well known that the time variations of the diverse solar wind variables at 1 au (e.g., solar wind speed, density, proton temperature, electron temperature, magnetic field strength, specific entropy, heavy-ion charge-state densities, and electron strahl intensity) are highly intercorrelated with each other. In correlation studies of the driving of the Earth's magnetosphere-ionosphere-thermosphere system by the solar wind, these solar wind intercorrelations make determining cause and effect very difficult." Kutiev et al. (2013) analyzes the complexities of the interactions between solar wind and Earth's upper atmosphere. The work of these and many other authors make it no surprise that in this paper we are confined to categorizing night sky brightness events as contemporaneous with, rather than being in a simple cause and effect relationship with geomagnetic events and elements of the solar wind.

Section 5.1 contains SQM-LU-DL light curves of the night sky, their accuracy on photometric nights, and how they are used to measure changes in the night sky brightness.

In Section 5.2 we present observations of individual geomagnetic and contemporaneous night sky brightness events. In every case the maximum in night sky brightness occurs approximately 3 or more days after the peak of the geomagnetic event.

Section 5.3 explores the night sky brightness events which are contemporaneous with changes in the solar activity.

Figure 19(a) shows the 10.7 cm Solar Flux is low and at a relatively constant level during a period when there were significant night sky brightness variations. Figure 4 of Walker (1988) and Figure 1 of Pilachowski et al. (1989) show a significant correlation between 10.7 cm solar flux and night sky brightness measurements when the 10.7 cm flux is above the level we encountered. Figure 3 of Krisciunas et al. (2007) shows during a solar minimum the 10.7 cm solar flux is at a relatively low value compared with near solar maximum with night sky brightness changes near the solar minimum. Figure 19(a) is in agreement with these observations. It would appear different physical processes are at work in creating night sky brightness variations during solar maximums and solar minimums, respectively.

Figures 19(b)–(d) show night sky brightness increases are contemporaneous with increases in solar wind speed, as well as, some solar wind proton and/or electron fluence events.

Section 5.4 contains summary Figures 21(a)–(d). These plots compare SOHO/DSCOVR mean solar wind speeds, GEOS Solar Proton Fluence, and Potsdam Ap Indexes with the night sky brightness increases of more than 0.2 mag arcsec⁻² above quiescent values of Figures 11 and 12.

Figures 21(a) and (c) show streams of increased mean speed in the solar wind as observed by the SOHO/DSCOVR satellite. Earth's entry into these enhanced flows appears to be contemporaneous with increases in the Ap index and with brightness increase night sky events.

The greatest increases in night sky brightness are contemporaneous with large increases in solar proton and electron fluence as shown in Figures 19(c) and (d). However, geomagnetic disturbances and night sky brightness events are not always contemporaneous with increases in solar proton fluence (see Figures 21(b) and (d)). These observations indicate several different physical processes may be involved in night sky brightness events.

Cosmic Campground IDSS and Aotea/Great Barrier Island IDSS each experienced three night sky brightness events during which the night sky brightened by more than 0.2 mag arcsec⁻² above quiescent values of Figures 11 and 12. For all six of these, the maximum night sky brightness occurred 3 to 8 days after the time of greatest of geomagnetic activity.

(See Figures 21(a)–(d)).

The natural night sky is never without light. It is a constantly changing kaleidoscope of celestial and atmospheric sources of illumination. The unique result of this paper is these statements are true even in deep solar minimum conditions. In particular, in 2018, as Solar Cycle 24 entered a deep solar minimum, the natural night sky was measured to range in brightness from 21.99 to 21.13 mag arcsec⁻² at Cosmic Campground IDSS and from 21.99 to 21.14 mag arcsec⁻² at Aotea/Great Barrier Island IDSS. The measurements show the natural night sky goes through Bortle Scale 1–4 (Bortle 2001) without human intervention. The brightest readings at both locations were contemporaneous with Earth being in an energetic stream in the solar wind blowing from a coronal hole in the Sun's atmosphere and with Earth's magnetic field being in a disturbed state (i.e., minor G1-G2 Geomagnetic Storms predicted to be possible).

The 2018 natural night sky had brightening events even though the 10.7 cm solar flux was at a low steady level.

It is likely both changes in the brightness of natural night sky and disturbances in Earth's magnetic field are driven by variations in the solar wind. In particular, there appears to be a relationship between the mean speed of the solar wind and the brightness of the natural night sky. The largest increases in night sky brightness are contemporaneous with pulses in solar proton and electron fluence. Some lesser night sky brightness events seem uncorrelated with these phenomena. These observations suggest there may be several different physical processes producing night sky brightness events. With more research, it may be possible to identify them as the temporal length of our database increases.

Charged particles in the solar wind take days to reach near-Earth environment after a coronal hole is observed to be facing in our direction. Use of this information could make it possible to predict increases in Earth's natural night sky brightness several days in advance. Scientists could benefit by incorporating these predictions into observing schedules.

Some rapid changes in natural night sky brightness could be a result of the Russell & McPherron (1973) effect which comes about when the solar wind opens cracks in Earth's magnetic field allowing charged particles to enter the upper atmosphere.

What we have learned during this solar minimum leads us to search for other solar driven changes in night sky brightness as the Sun begins to move into solar maximum conditions.

Additionally, it should be emphasized there will be unexpected changes in the natural night sky when Earth is impacted by a significant coronal mass ejection.

The raw data produced by an SQM-LU-DL demonstrates a stable instrument capable of making accurate differential photometric measurements within several hundredths of a magnitude.

The techniques introduced in this paper provide a powerful tool to relate simple, single pixel, photometric measurements of the natural night sky to solar wind changes and geometric activity. The results provide insight into the correlation between changes in the Sun and Earth's responses to them.

Appreciation is expressed to the referee whose suggestions have substantially improved this paper. Anthony Tekatch, Eric Christensen, and John Barentine provided information, encouragement, editing suggestions, and advice. In addition to content contributions, Patricia A. Grauer is responsible for editing changes which make the document coherent. The program DataGraph for macOS by Visual Data Tools, Inc. was used to make the figures in this paper.

Two friends, businessmen, took their after dinner wine out on the patio while the wives cleaned up. They sat in darkness looking at the natural night sky. Robert Robinson gestured toward the natural night sky of Catron County and remarked to Bill McCabe, "Too bad we cannot use that to help the economy." Bill McCabe's online search for "dark skies," later that night, became the genesis of what is today the Cosmic Campground International Dark Sky Sanctuary.

Retired Glenwood District Ranger Pat Morrison identified the natural night sky at the Cosmic Campground IDSS and started the process to make it a part of the Gila National Forest. Bill McCabe was a relentless advocate for an astronomy campground. Craig Wentz was a member of the crew who installed the SQM-LU-DL at the Cosmic Campground. This project would not be possible without Erick Stemmerman, Adam Mendonca, Bob Shanks, John Baumberger, and other Gila National Forest employees who facilitated the data collection process.

Special thanks to Justin LeGrice and Mark Durling for assistance with the setup and operation of the data collection equipment. The natural night sky of Aotea Great Barrier Island is located within Auckland, New Zealand's largest city hosting over a third of the countries population, and remains protected thanks to the continued commitment of the Aotea Great Barrier Island Local Board, Auckland Council and the residents of Aotea Great Barrier Island.