Introduction

The phenomenon of twilight sky brightness is attributed to the scattered radiation of the sun at the observer’s ___location. Yallop, in profiling twilight sky brightness, found that the light intensity and its spectrum vary greatly with the sun’s altitude below the horizon. The scattered light causes a reddening of the blue near sdfdfunset, reaching its minimum brightness at approximately 18° of solar depression. This signifies the end of sunlight-scattered radiation, marking the end of twilight and the commencement of night. The element that contribute to the twilight sky brightness consist of thre major elements.

Solar depression and geometrical relation

The twilight sky brightness is factored by the solar depression, \(\:h\), solar azimuth, \(\:\theta\:\) and height of an object, or zenith difference from the sun, under the subject of observed illumination, \(\:{b}_{ST}\). Schaefer state the equation as1,

$$\:\text{log}{b}_{ST}=4.75-\frac{\theta\:Z}{3}+h\left(12+8.21Z\right)+2.86Z.\:\:\:\:$$
(1)

where \(\:{b}_{ST}\:\)is the twilight sky brightness, \(\:\theta\:\) is object azimuth distance, \(\:Z\) is zenith difference from a object to the sun and \(\:h\:\)is solar depression. Zenith Distance (Z), measured in degrees. It represents the angle between the horizon and the object, with a negative value indicating the Sun is below the horizon (e.g., twilight). Object azimuth distance, (θ), measured in degrees. It indicates the angular distance of the object around the horizon, relative to a reference direction (typically north). Sky Brightness (\(\:{b}_{ST})\), typically expressed in magnitudes per square arcsecond (mag/arcsec2), which describes how much light is spread over a unit area of the sky.

Altitude and azimuth are crucial factors that influence the brightness of the sky background from an object. The higher the azimuth distance (θ) and the zenith difference (Z) of the object, the greater the separation between the object and the Sun. This increased distance reduces the amount of sunlight scattered towards the object, which leads to a darker background sky. Consequently, the twilight sky brightness, represented as \(\:{b}_{ST}\)​, decreases. The reduction in brightness is due to the fact that a higher relative distance diminishes the intensity of scattered light, contributing to a darker sky. Thus, the equation models how the relative position of the object to the Sun, through azimuth, zenith, and solar depression, influences the perceived brightness of the twilight sky.

Sunlight scattering radiation, \(\:\:\varphi\:\left(\theta\:\right)\)

The sunlight scattering radiation is the major factor of twilight illuminance, since the beginning of the twilight is marked by the depression of sun altitude. Since there is no direct lumination towards the observer during twilight, the twilight brightness is characterized into two main part, the first part is the single scattering twilight, usually most intsense during sun altitude below to 4–5° from horizonj, and disappears around solar depression of 10°. Then the twilight sky brightness is replaced other source of light, such as light pollution, sky glow or milky way, in which the second part of twilight brightness, The main of characterization is contrast in behavior of sky color and light polarization during twilight, as it has been proven by Hassan that the reddening of the sky is at minimum during solar depth of 12.5°2.

Aerosol optical depth and ozone composition, \(\:{\tau\:}_{z}\left(\lambda\:\right)\)

It is noted that in addition to sunlight scattering, the concentrated aerosol and unique composition in ozone layer in the sky play a role in determining the brightness of twilight, especially at higher altitude3. Aerosol and ozone also play a role in influencing spectrum of sky color during twilight. Under the normal spectrum condition, the zenith of the sky will appears greenish-red in color, however the ozone composition contributes in blueing the zenith sky, while reddish-orange color on the sky within the background of sunset is due to of aerosols4.

Kastner demonstrated that this empirical expression could be successfully applied to observing celestial objects, notably replicating observations of comet Kohoutek in January 1974. His contemporary, Bruin, shared an interest in determining the visibility of celestial objects during twilight, applying the same concept of brightness contrast to find the limiting visibility of the new moon5,6. Kastner and Bruin’s studies are widely regarded by astronomers for pioneering the combination of contrast thresholds and twilight brightness data to determine limiting visibility. However, their studies did not account for factors such as object azimuth during twilight, ground brightness, and human eye sensitivity, which were later addressed by Schaefer.

Bradley Schaefer’s comprehensive work, “Astronomy and the Limits of Vision,” integrates considerations of atmospheric extinction, source and twilight photometry, contrast thresholds, the resolution of optical aids or naked-eye observation, and the physiological factors of the observer. Schaefer’s research has been applied in various fields, notably in historical and biblical studies, enhancing the understanding of celestial visibility during twilight7,8.

Light pollution is the presence of artificial light in the night environment, resulting from urbanization, streetlights, buildings, and other human-made sources. This unwanted light interferes with the natural darkness of the night sky. Light pollution reduces the contrast between celestial objects and the night sky, making it difficult to see faint stars, galaxies, and nebulae. This diminished contrast obscures many of the finer details that astronomers aim to study. Light pollution interferes with both visual and telescopic observations. For amateur astronomers, it means fewer visible stars and planets. For professional observatories, it necessitates the use of more advanced filtering techniques and can limit the effectiveness of astronomical research9,10.

The limiting magnitude, or the faintest star visible to the naked eye, is significantly reduced by light pollution. In heavily polluted areas, the limiting magnitude can drop from around 6 (in a pristine sky) to 3 or lower, meaning many stars that would otherwise be visible are no longer detectable. For astrophotographers, light pollution creates a persistent glow that can interfere with long-exposure images, adding noise and reducing the clarity of the captured celestial scenes. This often requires additional post-processing to mitigate the effects of the artificial light. This make the light pollution monitoring is a continuous concern for astronomers throughout years considering its factor on the both optical and radio observation, and it has been made mandatory to determine the quality of the observation site11.

Most research on light pollution is limited to the study of the night sky, rarely extending to the study of twilight brightness. However, Garstang and Barry found that artificial light brightens the night sky across all zenith distances, particularly at the horizon where twilight brightness is monitored, as depicted in Figs. 1 and 2. Despite these findings, the specific effects of light pollution on twilight brightness have not been thoroughly replicated or studied12.

Fig. 1
figure 1

Effect of light pollution on all zenith angle (Berry 1974).

Full size image
Fig. 2
figure 2

Effect of light pollution on all zenith angle (Garstang 1986).

Full size image

One of the most significant historical examples of twilight astronomical observation is the sighting of the crescent moon (Hilal) to determine the beginning of a new month in the Islamic calendar. This practice dates back to the early days of Islam and continues to be of great importance in the Muslim world. The Islamic calendar is a lunar calendar, and months begin with the sighting of the new crescent moon. One of the most important months in the Islamic calendar is Ramadan, a month of fasting and spiritual reflection. The start of Ramadan is determined by the sighting of the crescent moon during twilight, shortly after sunset. For centuries, Muslim communities have relied on visual observations of the crescent moon to mark the beginning of Ramadan. This tradition is rooted in the Hadith, sayings, and actions of the Prophet Muhammad, who instructed his followers to begin fasting upon sighting the new moon and to break the fast at the sighting of the next new moon13,14,15.

The observation occurs during twilight, after sunset but before the sky becomes fully dark. This is when the thin crescent moon first becomes visible. Observers typically choose locations with clear horizons, away from light pollution, to increase their chances of sighting the crescent moon. The moon must be visible to the naked eye, and its sighting is often confirmed by multiple observers to ensure accuracy. In modern times, light pollution has made it more challenging to observe the crescent moon in urban areas16.

Another example is the the heliacal rising of Sirius, the brightest star in the night sky, was of great importance in ancient Egyptian astronomy. This event occurs when Sirius first becomes visible above the eastern horizon just before sunrise, during twilight. The heliacal rising of Sirius, also known as Sothis to the ancient Egyptians, marked the beginning of the Egyptian New Year and the start of the annual Nile flood, which was crucial for agriculture. This event was celebrated with festivals and held significant religious and practical implications for the society17.

The observation took place during twilight, just before dawn, when Sirius first appeared in the eastern sky after being invisible for about 70 days due to its proximity to the sun. Observers chose locations with clear views of the eastern horizon, often from temple complexes or other elevated sites. The heliacal rising was not only a calendrical marker but also had deep religious significance, associated with the goddess Isis and the resurrection of the god Osiris.

The heliacal rising of Sirius was used to predict the annual flooding of the Nile, which deposited fertile silt on the riverbanks and was essential for successful crop planting. The event was intertwined with Egyptian mythology and was an integral part of their religious calendar, with temples oriented to align with the rising of Sirius18.

Observation of the International Space Station (ISS) Passes could also occur during twilight. The International Space Station (ISS) is a habitable artificial satellite that orbits the Earth. Observing the ISS as it passes overhead has become a popular activity for both amateur and professional astronomers. The best time to observe the ISS is during twilight, shortly after sunset or before sunrise, when the station is illuminated by the sun, but the sky is dark enough for it to be visible.

Observing the ISS has become a way to engage the public with space and astronomy. Many people enjoy watching the station pass overhead and learning about the astronauts on board and their missions. Schools and educational programs often incorporate ISS observation into their curricula to teach students about space, orbits, and the International Space Station’s role in scientific research. For amateur astronomers, observing the ISS is a gateway activity that can spark further interest in astronomy and space science. In urban areas, light pollution can make it more difficult to see the ISS. However, its brightness often makes it visible even in light-polluted skies.

Concerning the essential of this study, the brightness behavior profile of evening twilight is monitored, using the human population as the main variable. The various human population density is examined in the terms ___location profile to understand general effect of light pollution on the twilight brightness.

Methodology

In this study, the Sky Quality Meter (SQM) is utilized to measure the sky brightness during twilight. The SQM is a portable device developed by Unihedron, designed to assess sky brightness in units of mag/arcsec². It employs a light intensity-to-frequency detector, which converts the output into a digital LED reading. To ensure that the measurements are within the visible wavelength, the sensor is equipped with a near-infrared blocking filter. The SQM will be pointed towards the position of the horizon sunset or sunrise, as such shown in Fig. 3.

Fig. 3
figure 3

SQM setup during acquisition of twilight magnitude brightness.

Full size image

To find evidence of the alteration of light pollution on twilight sky brightness, the main variable, human population density is manipulated and represented with site observation profiles. Our observation sites are divided into three profiles, Urban, Rural and Pristine. Urban is the ___location that has dense population and extreme volume of artificial light propagation towards is sky, the rural or suburban is the ___location has medium level of population and artificial light, while the pristine ___location is the remote ___location far from any human activity.

Urban sites are characterized by high population densities and significant artificial light pollution. These areas are typically within or close to major cities where light pollution is most intense. Typically greater than 1,000 people per square kilometer. These locations usually fall into Class 7 to 9 on the Bortle Scale and larger than 18.00 mag/arcsec2 reading of Sky Quality Meter, indicating highly light-polluted skies where the Milky Way is not visible and only the brightest stars are seen. With examples of City centers, heavily populated metropolitan areas. These sites were selected based on their high population density and significant artificial lighting. Examples Kuala Lumpur, with a population density exceeding 10,000 people per square kilometer.

Rural or suburban sites have moderate population densities and intermediate levels of artificial light pollution. These areas are usually on the outskirts of cities or in small towns where light pollution is present but less intense than in urban areas. The population density typically ranges from 100 to 1,000 people per square kilometer. These locations generally correspond to Class 4 to 6 on the Bortle Scale and 20.3 to 18.00 mag/arcsec2 reading of Sky Quality Meter, indicating skies where the Milky Way is visible under good conditions but with some impact from light pollution. Examples such as Suburban neighborhoods, small towns, and peri-urban areas. Selected for their intermediate population density and moderate artificial light levels. An example is a Port Dickson, with a population density of around 500 people per square kilometer.

Pristine sites are remote areas with very low population densities and minimal artificial light pollution. These areas are far from urban centers and human activities, providing the best conditions for observing natural twilight brightness. The Population Density typically less than 100 people per square kilometer. These locations fall into Class 1 to 3 on the Bortle Scale and 22–20.3 mag/arcsec2 reading of Sky Quality Meter, indicating dark skies where the Milky Way is prominent, and faint celestial objects are visible. National parks, remote rural locations, and protected natural reserves. Pristine site are chosen for their low population density and minimal light pollution. Examples include Mersing in Johor, dan Kundasang in Sabah, where population density is less than 10 people per square kilometer. The observation sites are listed in Table 1.

Table 1 Observation sites.
Full size table

We utilize ArcGIS maps that depict light pollution levels across different regions. These maps are generated using satellite data and provide a comprehensive overview of light pollution distribution. The SQM readings from each observation site are compared with the corresponding light pollution levels indicated on the ArcGIS maps. This step helps validate whether the SQM readings accurately reflect the expected sky brightness based on the mapped light pollution levels.

Before beginning data collection, the SQM devices are calibrated using standard procedures recommended by the manufacturer (Unihedron). This ensures the instruments provide accurate readings.Regular calibration checks are performed throughout the data collection period to maintain the accuracy of the SQM devices. Any discrepancies are noted and corrected.Consideration of environmental factors such as humidity, temperature, and altitude that might affect the SQM readings. Calibration adjustments are made if necessary.

If the SQM data indicate that a ___location is pristine (i.e., very low light pollution), we cross-check this with the ArcGIS light pollution map. A pristine classification on the map should correspond with low light pollution levels recorded by the SQM. Any anomalies between the SQM readings and ArcGIS data are investigated. Potential sources of discrepancies, such as temporary local light sources or atmospheric conditions, are identified and documented.

In this paper, we focus specifically on the impact of light pollution on twilight sky brightness. Therefore, the variations caused by these external factors are not the primary focus, as they have been extensively discussed in existing literature (Faid et al., 2016, 2018; Herdiwijaya, 2014; Studi et al., 2013). This allows us to concentrate on the core issue of how artificial lighting affects twilight sky brightness. Most existing publications have followed this approach by addressing the variation in twilight sky brightness data due to clouds, haze, moonlight, sudden headlights, and uncalibrated SQMs. Since these variations have already been thoroughly discussed, this paper directly focuses on examining the influence of light pollution on twilight sky brightness without reexamining the well-documented sources of data variation.

Result and discussion

Table 2 presents the results of zenith light pollution measurements at the observation sites. These results were compared with ArcGIS light pollution maps to ensure the validity of the instrument calibration during data acquisition. It was found that the data for sites number 9 and 12 were not valid. The light pollution map indicates zenith light pollution values of 21.39 and 20.99 for sites 9 and 12, respectively, while the SQM data showed values of 18.94 and 19.05. This discrepancy suggests that the instruments used at sites 9 and 12 were not properly calibrated, and therefore, the data from these sites must be rejected.

Table 2 Zenith light pollution on observation sites.
Full size table

After eliminating the data from Pantai Nenasi and Pulau Bunting, a total of 72 data points on twilight brightness were collected. This dataset includes 42 data points from pristine sites, 29 from rural sites, and one from an urban site. The data were categorized based on their respective site locations and illustrated in graphs to highlight the effect of light pollution on twilight sky brightness. Each site is represented by an average value, and trendlines with an R value greater than 0.98 were used to ensure reliability. The results are depicted in Figs. 4, 5, and 6. All the graphs are then combined into one to examine the overall effect of light pollution on twilight brightness. The average readings for each site profile are measured and expressed in Fig. 7, providing a comprehensive view of how light pollution impacts twilight brightness across different environments.

Fig. 4
figure 4

Alteration of light pollution on Pristine ___location.

Full size image
Fig. 5
figure 5

Alteration of light pollution on rural ___location.

Full size image
Fig. 6
figure 6

Alteration of light pollution on urban ___location.

Full size image
Fig. 7
figure 7

Twilight brightness on different ___location profiles.

Full size image
Fig. 8
figure 8

Twilight brightness on different site profiles.

Full size image

The graph in Fig. 7 illustrates the impact of light pollution on the twilight sky brightness. The variations in twilight brightness magnitude across different sites reveal a recognizable pattern, clearly distinguishing between two phenomena: atmospheric scattering and light pollution. This pattern helps to understand the contributions of natural and artificial factors to the observed twilight sky brightness.

The contrariety is illustrated in Fig. 8. Part I, denoted as the natural phenomenon of twilight, represents the scattering of sunlight. This explains why the twilight brightness across the three site profiles (urban, rural, and pristine) is essentially the same, indicating that light pollution has not yet affected this part of the twilight period. Part II, however, depicts the role of light pollution in altering twilight brightness. Here, the twilight brightness in urban areas is higher than in rural locations. Additionally, the degree of solar altitude at which twilight brightness reaches its stability, marking the commencement of night, is lower in urban areas compared to rural ones. This demonstrates the significant impact of light pollution on the natural progression of twilight.

Table 3 Parameters of twilight when brightness stability is reached.
Full size table

This behavior as shown by Table 3 is important as the degree of twilight brightness upon reaching its stability at suburban and urban site defy the major work found by the twilight study, which is around 16°-18°. This result shows a major impact of light pollution on the twilight brightness profile that. The primary data used for this study is from Teluk Kemang, as it has records from 2008. Teluk Kemang was once a premier site for lunar crescent observation and achieved several world records for lunar crescent sightings. However, Teluk Kemang has recently been unable to produce new records, with light pollution being the primary culprit.

From 2008 to 2022, Teluk Kemang experienced a spillover effect from the tourism economy of Port Dickson. This spillover led to increased light pollution due to the construction of nearby hotels and other tourist facilities. As a result, the quality of the night sky has deteriorated, impacting the ability to observe faint celestial objects such as the lunar crescent. To demonstrate the impact of light pollution on Teluk Kemang, process of smoothing the data through trend lining was generated using Tablecurve 2D, resulting in Fig. 9. This approach allows us to clearly visualize the variations in twilight brightness over different years. By applying trend lines, we can identify underlying patterns and anomalies in the data.

Fig. 9
figure 9

Twilight sky brightness reading over the years.

Full size image
Fig. 10
figure 10

Twilight sky brightness over the years at 17 degree of solar depression.

Full size image
Fig. 11
figure 11

Twilight sky brightness stability variation over the years.

Full size image
Table 4 Twilight sky brightness increment with estimated population function at Teluk Kemang.
Full size table

The “Estimated Population” quantity shown in Table 4 is derived based on data collected from the Malaysia Department of Statistics. This data is typically gathered through national surveys, censuses, and population projections conducted by the department. These estimates take into account various demographic factors such as birth rates, death rates, immigration, and emigration trends, providing a detailed and accurate reflection of the population within specific regions or across the country. For clarity, the population estimates in Table 4 are not calculated directly within the study but are sourced from authoritative governmental records, ensuring reliability and consistency in the data used for analysis. This external data provides a foundation for comparing other relevant variables and making informed assessments within the context of the study.

While the data in Fig. 9 appears similar in profile, a significant increase in twilight sky brightness is discovered upon zooming in on solar depression greater than 17°, as portrayed in Fig. 10. This closer examination reveals critical insights that are not immediately apparent in the broader view. The raw data of twilight sky brightness exhibits minor irregularities unique to each month, influenced by factors such as solar activity, atmospheric clarity, and cloud density. Despite these variations, a clear pattern of increasing twilight sky brightness over the years is visible, particularly as the brightness stabilizes at the end of twilight. This trend is illustrated in the box and whisker graph in Fig. 11. The increase in twilight sky brightness highlights the impact of human population growth, specifically light pollution, on the twilight sky brightness profile. In Teluk Kemang, the human population increased drastically from 2008 to 2016, corresponding with the rise in twilight sky brightness as shown in Table 4. Another intriguing influence of light pollution is its effect on the ending of twilight. Typically, the end of twilight is indicated by the stabilization of twilight brightness readings and is influenced by the degree of solar depression. However, it has been observed that the conclusion of twilight, which marks the onset of night, behaves differently depending on the zenith light pollution levels. This suggests that light pollution not only affects the brightness during twilight but also alters the duration and characteristics of twilight itself, delaying the transition to complete night (Fig. 12).

Fig. 12
figure 12

Twilight Zenith and horizon brightness at different sites.

Full size image

This behavior is significant because the degree of twilight brightness upon reaching stability at suburban and urban sites deviates from the established findings in twilight studies, which typically indicate stabilization around 18° of solar depression. This highlights another impact of light pollution on the twilight brightness profile. Both Berry (28) and Garstang (20) produced graphs showing the zenith distance dependency of light pollution, indicating that sky brightness readings can differ by up to 0.5 mag/sec² from zenith to horizon when the sensor is pointed towards the city. Our study, which explores the relationship between horizontal and zenith sky brightness, combined with the data from Berry and Garstang, resulted in the graph generated by Sigmaplot, as shown in Fig. 13. In this graph, a positive zenith distance indicates that the sensor is pointing towards the source of light pollution (i.e., the city), while a negative zenith distance shows the opposite direction. Berry and Garstang argued that sky brightness could differ by up to 0.5 mag/sec² from zenith to horizon. However, our study found that the maximum difference from zenith to horizon is 0.2 mag/sec² when the sensor is pointed away from the city center. This graph is crucial for demonstrating the stability of twilight sky brightness across various site profiles, highlighting the impact of light pollution on these readings.

Fig. 13
figure 13

End point of sky brightness reading.

Full size image

Conclusion

The study of twilight sky brightness reveals significant insights into how various factors such as solar depression, sunlight scattering, aerosol optical depth, and light pollution impact the visibility of astronomical objects. Light pollution significantly alters twilight brightness, increasing it particularly at greater solar depressions (> 17°). The study found that sky brightness can differ by up to 0.2 mag/sec² from zenith to horizon when sensors are pointed towards urban light sources, contrasting with Berry and Garstang’s findings of a 0.5 mag/sec² difference. Teluk Kemang, once a premier site for lunar crescent observation, has experienced increased light pollution from 2008 to 2022 due to tourism development in nearby Port Dickson. This has deteriorated the quality of the night sky, impacting the ability to observe faint celestial objects such as the lunar crescent.

This research is signifinat to new Hijri month determination and Subh and Isya’ prayer time. The accurate observation of the crescent moon (Hilal) is crucial for determining the beginning of a new Hijri month, including Ramadan. Understanding the impact of light pollution on twilight brightness can improve the accuracy of these observations, ensuring the timely and precise determination of important dates in the Islamic calendar. The study’s findings on twilight behavior are significant for determining the precise timings for Subh (dawn) and Isya’ (night) prayers. As these prayer times are closely linked to the conditions of twilight, understanding the effects of light pollution can help in setting accurate prayer schedules, especially in urban areas where artificial lighting is prevalent.