2.1. Locations of Interest

Several sites were selected to measure NSB in Colombia based on previous SQM records. These locations include the BBG and the TD, both of which have existing SQM data (^{Urrego Guevara 2016} and ^{Góez Therán & Vargas Domínguez 2021}, respectively). The study also incorporates the Astronomical Observatory of the Universidad Tecnológica de Pereira (hereafter AOUTP) and Cerro Guadalupe (hereafter CG) in Bogotá. Each site has distinct characteristics in terms of topography and primary activities. AOUTP is focused on astronomical research, TD is known for astrotourism, BBG is dedicated to biodiversity conservation, and CG serves as an ecological corridor¹². Figure 1 depicts the geographic locations of these sites overlaid on a pseudo-color map adapted from the New World Atlas of Artificial Night Sky Brightness (^{Falchi et al. 2016}). Table 1 presents the geographic coordinates obtained and the acronyms by which each site is identified.

Fig. 1 The geographical distribution of the four sites analyzed in this study is represented by red points with white labels on a background color map, constructed using data from the New World Atlas of Artificial Night Sky Brightness (^{Falchi et al. 2016}). The color figure can be viewed online. 

2.2. Data Acquisition, Reduction and Analysis

The SQC used in this study is a Canon EOS 6D Mark II featuring a 26.2-megapixel full-frame CMOS sensor and an integrated GPS. It comes equipped with a Sigma 8 mm EX DG fisheye lens able to image in a single shot a field of view of 186°. The entire setup is mounted on a Manfrotto MK055XPRO3-3W tripod. After several years of intensive use, we can confirm that the SQC maintains a photometric precision of ±0.01 mag/arcsec², a CCT precision of ± 30 K, and a spatial resolution of ± 0.02°.

We took data only after astronomical twilight, i.e., when the Sun elevation was ≥ -18°, and on those times when the Moon was ≥10° below the horizon. At each site, we identified the optimal location for the SQC installation, ensuring a 30 meter radius free of nearby luminaries and avoiding any direct illumination on the equipment. Additionally, we prioritized sites with the clearest possible view of the sky to minimize (whenever possible) obstructions from elements such as trees, mountains, and buildings. As a matter of fact, due to the natural surroundings, achieving a fully unobstructed view was not always feasible, particularly at BBG and CG. Subsequently, the camera was positioned to point towards the zenith, with its front aligned to geographical north using a compass. A circular bubble level was placed on the cover to ensure precise leveling. In this study, the ISO was set to 1600 at all sites, while the shutter speed was adjusted based on ambient brightness to prevent saturation and ensure a signal to noise ratio (S/N) greater than 20. All this information was recorded in the observation log, shown here as Table 2, which includes acronyms, date, time, exposure duration, and S/N values.

In order to preserve the maximum amount of information, the images were processed and analyzed in the native raw Canon .CR2 format using the commercial SQC software (v.1.9.9). After each image we also took the corresponding dark, that is automatically subtracted before the image file is written to disk. For further details about the data acquisition, reduction and analysis processes, the interested reader can refer to ^{Angeloni et al. (2024)}.

3.1. Astronomical Observatory of the UTP (AOUTP)

The AOUTP¹⁵ is located in the Central West metropolitan area¹⁶. Specifically, the observatory is situated in the southeastern part of Pereira, on the university campus in the La Julita sector. The observatory has received several international recognitions. Notably, it has been granted the code W63 by the Minor Planet Center, identifying it as an observatory authorized to obtain astrometric data of minor bodies in the Solar System (^{Jiménez Villariaga et al. 2017}). Additionally, the observatory holds the UTP 0383 code from the Stanford Solar Center for space weather monitoring (^{Galvis et al. 2019}). These recognitions establish the AOUTP as one of Colombia’s leading scientific centers for astronomical and space science.

In 2025, the observatory will expand its infrastructure and upgrade its astronomical equipment. A new facility will house a 70 cm diameter telescope, which, along with the existing instrumentation, will establish the observatory as the leading center for astronomical research in Colombia. These advancements will not only elevate the university’s academic contributions but also strengthen its research capabilities, community outreach, and engagement in the field of astronomy.

It is worth noticing the ecological value of the UTP campus, where 58% of its surface is dedicated to forest conservation and consequently classified as Natural Wild by Botanic Gardens Conservation International¹⁷, as plant species thrive in their natural environment without human intervention. Specifically, 13 hectares are allocated to the UTP Botanical Garden, which houses approximately 542 species of flora that are meticulously identified and protected (^{García Sierra & Ramírez Cardona 2011}). This area serves as a valuable resource for research and environmental education, and a space for recreation and engagement with nature for students and visitors. The conservation and sustainability initiatives implemented on campus have garnered national and international recognition, establishing UTP as a model for university environmental management.

Figure 2 presents an all-sky view from the AOUTP observatory. In the RGB map, the observatory’s dome is visible on the horizon to the south, while the eastern horizon reveals the surrounding mountains and the lights from nearby houses. The sky features several cloudy regions illuminated by the sky glow from the city of Pereira, as corroborated by the cloud map. The NSB map highlights a bright region extending approximately from 240° to 60 in azimuth, corresponding to the cities of Pereira and Dosquebradas. Furthermore, a zone of increased surface brightness is detected in the azimuthal range of 270° to 300°, exhibiting a lower CCT than the rest of the sky in the CCT map, indicating the presence of a cloud amplifying the city’s surface brightness. The combination of these images provides valuable insights into how atmospheric conditions and urban lighting influence sky quality and, consequently, the effective ness of astronomical observations.

To identify the areas or sites that contribute most to the NSB and CCT, we constructed rings at different elevations to determine the directions of the most significant contributing sources in each study region. We refer to these segments of the sky as azimuthal profiles, abbreviated as AZ. Figure 3 presents the azimuthal profile for the AOUTP. Using the ALAN sources from Table 3, we determined the azimuth of the highest contribution to NSB and CCT for each region:

The primary contributors to the NSB for the AOUTP are the Arboleda Mall and the airport, which typically operate almost until midnight. The airport’s lighting remains constant to ensure safe landings, while the mall maintains its lights for various events. In contrast, the maximum CCT values correspond to areas near the AOUTP, such as the parking of the Botanical Garden, sports complex, and residential complexes. These places typically use LED lights and, in many cases, have lighting issues such as over-illumination or improper light orientation, which disperses light both toward the ground and the sky.

To determine the altitude affected by the AOUTP, we constructed a ZA profile (see Figure 4), which revealed distinct brightness patterns across different regions as they approach the horizon. In Region I, a sharp increase in brightness is observed at ZA = 83°, attributed to the city glow. Region II exhibits a rise in brightness between ZA = 73° and ZA = 81°, likely caused by a cloud amplifying the NSB. In Region III, a slight increase is evident between ZA = 25° and ZA = 40°, associated with the presence of a cloud. Additionally, a secondary increase occurs at ZA = 83°, linked to urban glow. In Region IV, NSB increases between ZA = 26° and ZA = 43°, due to cloud covered, and again between ZA = 75° and ZA = 85°, potentially caused by urban glow. It is worth noting that some areas do not extend to the horizon, as they are physically obstructed by features like mountains or the dome.

3.2. Tatacoa Desert (TD)

The Tatacoa Desert is a tropical dry forest situated in the Magdalena River valley, within the department of Huila, approximately 10 km from the municipality of Villavieja¹⁸. It is the second driest region in Colombia, renowned for its eroded landscape, forming a labyrinth of canyons, valleys, and rock formations in striking red and gray hues. The Tatacoa Desert uniquely combines geological, climatic, and biogeographical factors, making it an ideal site for studying Neogene biodiversity in South America (^{Flórez et al. 2013}, ^{Montes et al. 2021}) and paleogeography (^{Dill et al. 2020}).

As one of Colombia’s top tourist destinations, the TD attracts visitors with its semi-arid landscapes, outdoor activities, including stargazing (^{Montealegre Vega & Garavito González 2022}). In September 2019, it was certi ed as a Starlight Destination¹⁹, the first site in Colombia to receive this recognition. The Starlight Foundation, with support from UNESCO, the World Tourism Organization (UNWTO), and the International Astronomical Union (IAU), grant this certification to areas that protect and conserve dark skies.

Today, the Tatacoa Desert is a hub for astro-tourism and a premier destination in Colombia for astronomy enthusiasts. The desert hosts several observatories and astronomical complexes. For our measurements, we were at the Astrosur Observatory²⁰, located approximately in the center of the desert; it provided the necessary safety conditions to conduct our research.

TD is one of the locations that presented a NSB measurement with the SQM instrument during the night of April 14-15, 2014. These measurements were made during a total lunar eclipse, and therefore with a full Moon, allowing for a record of approximately 9 hours of measurement. Each measurement was taken at 5-minute intervals; for more details, refer to ^{Góez Therán & Vargas Domínguez (2021)}. As this is the only existing record of sky quality for this location, we will use the darkest measurement, which is 21.26 mag/arcsec², as the reference value. This reference will be fundamental for evaluating the evolution of sky quality in future studies.

The all-sky view from TD is depicted in Figure 5. The upper left image presents an RGB format, where the Milky Way can be clearly seen arcing from north to south, with Jupiter standing out as the brightest point in the sky. TD is the only location that offers an ALAN-free direction. According to the light pollution map, this direction is in the azimuthal range of 140° to 145° and corresponds to the Cordillera de los Picachos National Park. This protected area is known for its great diversity of fauna, including endemic and endangered species, and therefore experiences minimal human intervention.

On the horizon, a white glow stands out in the southern and eastern parts of the image, corresponding to the cities of Neiva, located ~ 37 km away, and Villavieja-Aipe, situated roughly 7 and 10 km, respectively. To the north, another yellow/orange glow originates from the cities of Melgar (121 km), Girardot (125 km), and Bogotá (193 km). This glow is intensified by the presence of clouds, which can be identified in the cloud map. The NSB map shows the areas with the highest surface brightness, highlighting that the Milky Way is a natural contribution of light. However, the most intense contributions come from the cities, with Neiva as the main source of brightness. On the other hand, when observing the CCT maps, it can be seen how the CCT of the night sky is confused with the CCT of the cities, with a white glow coming mainly from LED lights.

Figure 6 presents the azimuthal profile for TD. In this plot, NSB maxima can be identified for the three regions of interest at AZ = 203°, AZ = 263°, and AZ = 35°, respectively. Similarly, CCT maxima are observed for regions I and II at AZ = 194° and AZ = 261°, with a minimum for region III at AZ = 34°. The NSB analysis highlights the need to implement controls on the type of illumination in Neiva, Villavieja, and Aipe to prevent the TD sky from losing its dark sky quality or Starlight certification.

Figure 7 presents the zenith profiles of NSB/CCT. In Region I, there is an increase in brightness at ZA = 64° due to the passage of the Milky Way, with an additional increase attributed to the city of Neiva. Conversely, in the CCT, no significant variation is observed up to ZA = 69°, also corresponding to Neiva. In Region II, which has the second highest NSB value, an increase is observed as the horizon is approached. Within the range of ZA = 29° to ZA = 53°, the NSB is primarily influenced by the natural contribution of the Milky Way. However, beyond this ZA, the NSB rises rapidly due to the light pollution from the cities of Villavieja and Aipe. Additionally, the CCT analysis indicates that LED lights predominate in this direction, resulting in the highest CCT values observed in this region.

In Region III, the NSB contribution shows a considerable increase at ZA = 71° due to the ALAN originating from the cities. It is not possible to determine the NSB of the Milky Way in this region since its passage occurs very close to the cities. The CCT shows a noticeable decrease from ZA = 60°, corresponding to the area affected by the city’s contributions. Finally, Region IV, the ALAN-free direction is distinguished by its stable NSB profile, identifying it as the darkest area within the studied region. It is important to note, however, that stability in NSB variation does not necessarily indicate low light pollution; an area consistently illuminated by ALAN would also exhibit minimal variation, but with higher brightness levels. The NSB map remains relatively constant up to ZA = 38°. Beyond this angle, an increase in NSB is observed, primarily due to light scattering caused by the city of Neiva. This effect becomes more pronounced at ZA = 48°, where the presence of clouds significantly amplifies the sky glow. Conversely, the CCT profile exhibits a gradual decrease starting around ZA = 14°, reaching its minimum value near ZA = 78°. Beyond this zenith angle, the CCT begins to rise again. This increase is attributed to the absence of clouds over the mountainous region, allowing for clearer visibility of this portion of the sky.

3.3. Jose Celestino Mutis Bogotá Botanical Garden (BBG)

The “Jose Celestino Mutis” Botanical Garden in Bogotá is a research center dedicated to conserving high Andean and paramo forest ecosystems. Located in the metropolitan area of Bogotá²¹, in the Engativá locality, the Garden spans approximately 20 hectares. It boasts an impressive collection of flora, featuring 55,000 individual plants representing 1,187 species distributed across 34 ecological zones. The garden also includes various botanical collections, such as ethnobiological, herbarium, carpotheque and tissue collections (^{Cadena-Vargas et al. 2020}).

The BBG has become an environmental reference point and a must-visit destination for those interested in exploring and learning about the city’s and country’s rich biodiversity. In recent years, it has emerged as one of Bogotá’s main tourist attractions and serves as a vital green lung in the heart of the metropolis. Additionally, it offers educational programs and interactive activities for visitors of all ages, fostering environmental awareness and respect for nature. The garden also plays a crucial role in scientific research and conservation efforts, collaborating with various institutions to preserve Colombia’s unique plant species (^{Cadena-Vargas et al. 2021}).

As previously mentioned, BBG had existing ALAN measurements recorded using an SQM. In 2014, a study was conducted using 13 measurement points distributed throughout the garden, starting near the southern part of the lake and following a circular route that ended at the main entrance. For our study, equipped with an SQC instrument, we chose a single measurement point to capture the most comprehensive information of the entire sky. To maximize sky coverage, we strategically selected a location approximately 80 meters from the Tropicarium²² and the greenhouse circuit. This location aligns with an intermediate position between Points 4 and 5 from the study by ^{Urrego Guevara (2016)}.

Furthermore, considering the differences in equipment and the SQM’s limitation of providing only a single value, we used the average NSB value recorded at the 13 points in the 2014 study during September as a reference. This average value was 16.31 mag/arcsec².

Figure 8 shows the all-sky view in RGB format from the BBG, highlighting the distribution and presence of clouds. The silhouettes of trees and part of the tropicarium structure are visible, with clouds distinctly appearing both amidst and above the tree-tops. Since the BBG is located in an urban area like Bogotá, where elevated levels of NSB are anticipated, the scales for the NSB and CCT maps were adjusted accordingly. The NSB map utilizes the Las Vegas scale, ranging from 11 to 21 V mag/arcsec², while the CCT map spans from 1600 to 7800 K. Both maps indicate an increased contribution from a ZA = 60° towards the horizon.

To more accurately assess each region and identify the azimuths of significant contribution in NSB and CCT, Figure 9 illustrates the azimuthal profiles for the BBG. Selected altitudes of 30°, 40°, and 50° were chosen to mitigate interference from trees. However, Region IV is excluded from the profile due to the tropicarium and trees appearing prominently in that direction. Region I shows a peak NSB and a minimum CCT at AZ = 188°, corresponding to the Compensar Stadium. Region II reveals its maximum NSB at AZ = 170°, associated with the bowling alley and auxiliary courts of Compensar, while the minimum CCT at AZ = 122° aligns with the Velodrome. Region III displays the maximum NSB and minimum CCT at AZ = 296°, corresponding to the Normand a fields.

The zenith profile is presented in Figure 10. Regions I and II exhibit an increase in NSB at ZA = 48°. Region III shows a rise in NSB from ZA = 14° to ZA = 31°, attributable to the presence of a cloud, with a more pronounced increase from ZA = 55°. In Region IV, there is a gradual increase in NSB; however, data beyond ZA = 52° are unavailable due to obstacles. Conversely, the zenith profile of CCT demonstrates a decrease as it approaches the horizon. In Region I, this decrease initiates at ZA = 39°. In Region II, the change occurs at ZA = 45°. Region III shows an initial decrease from ZA = 9° to ZA = 34°, followed by a rapid decline beginning at ZA = 40°. Region IV exhibits a decrease from ZA = 12° to ZA = 23°, which is attributed to the reflection of light on the clouds. This decrease continues gradually until ZA = 52°.

3.4. Cerro Guadalupe (CG)

The Cerro de Guadalupe is part of the Cerros Orientales, a group of mountains located to the east of Bogotá and the Cruz Verde pramo in the eastern mountain range. These hills are designated as a national forest reserve and their strategic location makes them a crucial ecological connector for the city, facilitating the interaction of diverse mountain ecosystems (^{Garzón Díaz 2014}). The Cerros Orientales also play a vital role in regulating the climate and providing water for Bogotá (^{Bohorquez-Alfonso 2008}). Additionally, they boast rich and varied biodiversity, harboring numerous species of endemic flora and fauna that contribute to the conservation of the region’s natural heritage (^{Jiménez-Alvarado et al. 2017}).

With an elevation of 3,273 meters above sea level, the Cerro de Guadalupe is the highest viewpoint accessible in Bogotá, offering a spectacular panoramic view of the Colombian capital. Since pre-Hispanic times, this site has held great significance, as rulers used it for surveillance and control over the city. Today, the hill is a subject of research in various fields such as history (^{Mejía 2006}), environment (^{Rodríguez Ramos 2015}), architecture (^{Meza 2008}) and archaeoastronomy (^{Bonilla Romero 2011}, ^{Bonilla Romero et al. 2017}). It is also a prominent site of religious pilgrimage, housing the statue of the Virgin of Guadalupe, a place of worship for many faithful. The hill features hiking trails that attract tourists and mountaineering enthusiasts, who appreciate its natural beauty and unparalleled views (^{Rueda Esteban 2017}).

Figure 11 presents all-sky view from the parking lot of the Sanctuary of the Virgin of Guadalupe on Cerro de Guadalupe. The RGB format in the top left panel showcases numerous trees and, to the southeast, the sanctuary’s advertisement sign along with clouds in the same direction. Similar to the BBG, there is a noticeable scale change in NSB map and CCT map. Although this location is not within the city limits, its elevation above the city results in contributions from the urban environment. The NSB map (top right) indicates an increase in surface brightness between AZ = 240° and AZ = 15°, attributed to the city of Bogotá. Additionally, there is a brightness increase from AZ = 90° to AZ = 135°, corresponding to light scattered by clouds above the advertisement sign. This is corroborated by the cloud map and the CCT map (bottom right), which show a decrease in CCT in these regions.

This decrease is more precisely detailed in the azimuthal profile. Figure 12 displays rings centered at altitudes of 20°, 30°, and 50°, accounting for the presence of trees along the horizon. The 50° altitude ring is the only one that provides full azimuthal coverage, as vegetation is unevenly distributed. In contrast, the 20° and 30° altitude rings capture variations in NSB and CCT only in specific directions, limiting the ability to detect changes in these parameters across all azimuths. The most significant NSB peak is at AZ = 232°, corresponding to the city of Bogotá. Additionally, there are notable increases in surface brightness at AZ = 132° and AZ = 117°. For CCT, two peaks indicate a decrease, corresponding to the city’s glow on the cloud, with another decrease peak at AZ = 62°, corresponding to the direction of the church on the hill.

The zenith profile is illustrated in Figure 13. Region I shows a sharp increase in NSB at ZA = 72°, which decreases at ZA = 80° due to light dispersion by the cloud and further decreases from ZA = 87 due to the city’s glow. The CCT decreases from ZA = 58°. Region II exhibits an increase in NSB and a decrease in CCT from the zenith to ZA = 60°, limited by tree coverage. Region III presents the most significant increase in NSB and decrease in CCT starting at ZA = 57°, corresponding to a cloud in the sector. Region IV shows a gradual increase from the zenith to ZA = 64°, reaching maximum brightness, then decreasing and increasing again due to clouds among the trees, with the strongest CCT decrease at ZA = 65°.

4.1. Photometric Indicators for Cross-Site Comparisons

For comparative studies between different locations, a large number of photometric indicators of the visual quality of the night sky have been suggested in the literature (^{Deverchère et al. 2022}; ^{Duriscoe 2016}). These indicators enable us to evaluate the impact of light pollution from various perspectives, such as astronomical (^{Falchi et al. 2023}) and biological (^{Falchi et al. 2021}). In Table 4, we present the alt-az coordinates and the corresponding NSB/CCT values for the darkest and brightest points of the sky, as well as for the zenith (averaged over a 5° radius circle), within a 1° wide ring centered at 30° elevation, and over the entire celestial hemisphere (i.e., all-sky) for each site.

To better understand the types of light affecting the sky at each site, Figure 14 presents a comparative histogram illustrating the percentage distribution of the sky based on CCT. The histogram for TD, shown in blue, reflects a distribution with temperatures typical of a natural sky, but it also reveals a significant percentage of LED light, with a peak around 4100 K, indicating the presence of neutral-tone LEDs.

Fig. 14 The histogram represents the distribution of CCT values for four measurement sites: BBG (green), CG (yellow), AOUTP (red), and TD (blue). Each color corresponds to a different site, showing the proportion of area (%) covered by different CCT values across the range of 1600 to 6600 K. The bin width is set to 100 K. The color figure can be viewed online. 

In contrast, the histogram for AOUTP, represented in red, shows a bimodal distribution with peaks at 4500 K and 5000 K, suggesting a sky influenced by urban light sources. The 4500 K peak is associated with neutral light LEDs, commonly used in commercial areas, while the 5000 K peak corresponds to cool light LEDs, typical of large installations such as sports arenas, parking lots, or industrial facilities. Lastly, the histograms for BBG and CG reveal significant concentrations in the higher CCT ranges, between 5000 K and 6500 K. The peak near 6100 K suggests the presence of cool or high-intensity LEDs, characteristic of densely illuminated urban areas like Bogotá.

4.2. Comparison with Other Sites

The NSB and CCT maps of the Colombian skies presented in this study, derived from data collected by the SQC, constitute the first recorded measurements of light pollution near the equator as documented in the Light Pollution Map database. This platform compiles all les generated by the SQC instrument, allowing users to upload and share their observations, thereby ensuring public access to the data. Despite significant efforts to facilitate access to SQC data, measurements are still lacking for many cities.

For this reason, when comparing our data with those from the database, specific selection criteria were applied for each location. The AOUTP was compared with the Astronomskogeo-zikalni Observatorij Golovec (AOG) in Slovenia, as both are observatories located less than 10 km from the center of their respective cities. For TD, the Eifel National Park²³ (ENP) was chosen, as both sites are designated for dark sky preservation. In the case of BBG and CG, a comparison was made with Mont Royal Park²⁴ (MRP) and the Montreal Olympic Stadium (MOS), both located in Montreal, Canada. This comparison was based on the search for a city that had two measurement points no more than 12 km apart, where one of the sites was a park or an environmentally protected area.

First, the AOUTP was compared with the AOG in Slovenia. The AOG is located in Ljubljana, a city of approximately 285,604 inhabitants (2021 data²⁵), while the AOUTP is situated in Pereira, which has around 477,068 inhabitants (2018 data²⁶). They exhibit significant differences in terrain and climate, which may influence light pollution measurements. Ljubljana lies in a basin surrounded by mountains, potentially trapping light pollution, whereas Pereira is located in the Andean region, with distinct atmospheric conditions. These differences highlight the importance of contextual factors when interpreting light pollution data.

Figure 15 compares the NSB maps of AOUTP and AOG. From the maps, the darkest point values are highlighted, which are 18.12 mag/arcsec² and 19.58 mag/arcsec², respectively. The photometric indicator for the 30° elevation ring shows values of 17.11 mag/arcsec² and 18.89 mag/arcsec², respectively. This indicates that the night sky over AOG is 1.78 mag/arcsec² darker compared to Pereira. This difference can be attributed to several factors, including the higher population density of Pereira and the presence of an airport within its urban area, both of which contribute to increased NSB. Moreover, the frequent partial cloud cover over Pereira further elevates the recorded NSB by scattering and amplifying artificial light sources.

Fig. 15 Comparison of all-sky NSB images from the AOG (left) obtained from the Light Pollution database and from the AOUTP (right). The date, time, and geographic coordinates are displayed at the top left of each map, while NSB values in [V mag/arcsec²] are represented in the color bar on the right. The dark point, along with its coordinates, is marked above the table containing the NSB data at the bottom left. The color figure can be viewed online. 

In the case of TD, we compared our data with ENP, located in western Germany within the state of North Rhine-Westphalia, near the Belgian border in the Eifel mountain region. The park is situated close to the towns of Schleiden and Monschau, as well as the cities of Aachen and Cologne. Covering approximately 110 km², ENP is home to thousands of endangered animal and plant species. In 2014, it was designated as a protected area by the International Dark Sky Association²⁷, becoming one of the few regions in Germany where the Milky Way is visible to the naked eye (^{Gabriel et al. 2017}).

Figure 16 shows the NSB maps of TD and ENP. From the maps, the values of the darkest points stand out, which are 21.36 mag/arcsec² and 21.22 mag/arcsec², respectively. The photometric indicator for the 30° elevation ring shows values of 20.77 mag/arcsec² for TD and 20.38 mag/arcsec² for ENP, indicating that the night sky over TD is 0.39 mag/arcsec² darker than that over ENP. This difference is primarily due to the proximity of large cities near ENP, such as Cologne (Germany, 50 km away) and Liege (Belgium, 60 km away), compared to Neiva, which is 37 km away from TD. Furthermore, the populations of Cologne and Liege are 2.7 and 1.7 times larger than that of Neiva, respectively, contributing to the higher NSB observed in ENP. Despite these differences, it is important to note that both sites have implemented measures to protect their night skies, supported by their respective certifications. However, special attention should be given to the NSB levels at TD, as ENP is located in one of the most densely populated regions in Europe, making it particularly vulnerable to the effects of surrounding light pollution.

Fig. 16 Comparison of ENP and TD data, with analogous information to Figure 15. The color figure can be viewed online. 

As outlined earlier, we selected two locations in each city, situated no more than 12 km apart, for analysis. The first point of interest focuses on conservation parks that blend natural and urban environments: BBG in Bogot and MRP in Montreal. These spaces are distinguished by their commitment to environmental preservation within an urban setting.

The second comparison point includes CG in Bogot and MOS in Montreal. Both locations are within 12 km of their respective parks and serve as prominent city viewpoints. In Bogot, CG is a religious sanctuary that, beyond its spiritual significance, provides a breathtaking panoramic view of the city. Meanwhile, MOS features the iconic inclined tower of the Montreal Olympic Stadium, the tallest of its kind in the world. Known as the Montreal Tower, it offers a stunning 360-degree vantage point, allowing visitors to admire the city and its surroundings from an unparalleled perspective.

These cities exhibit notable differences in their geography and demographics. Bogot, situated high in the Andes Mountains at 2,565 meters above sea level, is defined by its rugged, mountainous terrain. In contrast, Montreal is located on an island in the St. Lawrence River, featuring a relatively at and gently rolling landscape. Demographically, Bogot is significantly larger, with approximately 8.034.649 inhabitants as of 2018 according to DANE, compared to Montreal’s 1,762,949²⁸ residents in 2021. However, Montreal is notable for having multiple SQC measurements recorded in the Light Pollution Database.

Figure 17 compares the NSB maps of MRP and BBG, highlighting the darkest points, which are 17.78 mag/arcsec² and 17.25 mag/arcsec², respectively. The photometric indicator (30° elevation ring) shows values of 17.08 mag/arcsec² and 16.01 mag/arcsec², respectively. According to the photometric indicator, MRP has a sky approximately 1.07 mag/arcsec² darker than BBG. This disparity is likely influenced by several factors, including Bogotá’s size, which is 4.4 times that of Montreal.

Fig. 17 Comparison of MRP and BBG data, with analogous information to Figure 15. The color figure can be viewed online. 

Figure 18 compares the NSB maps of MOS and CG, highlighting the values of the darkest points, which are 17.75 mag/arcsec² and 19.00 mag/arcsec², respectively. The photometric indicator (30° elevation ring) shows values of 17.27 mag/arcsec² and 18.39 mag/arcsec², respectively. According to this indicator, CG presents a sky approximately 1.12 mag/arcsec² darker than MOS. This difference could be attributed to the location of CG on the outskirts of Bogotá, in contrast to MOS, which is situated within the city’s interior. Both locations provide panoramic views of the city: CG, situated at an elevation of approximately 3,273 meters above sea level, and MOS, at 167 meters. From MOS, one can observe the city’s most iconic buildings and urban landscape, while CG serves as a natural look-out that also offers opportunities to observe local wildlife.

Fig. 18 Comparison of MOS and CG data, with analogous information to Figure 15. The color figure can be viewed online. 

4.3. GAMBONS Model

The night sky is not completely dark due to contributions from several natural sources, including integrated starlight, di use galactic light, extragalactic light, and zodiacal light. Additionally, phenomena such as airglow result from the interaction of particles in the upper atmosphere. These natural effects create a background glow that is always present in the night sky.

The GAMBONS model, as detailed in ^{Masana et al. (2021}, ²⁰²²⁾, accounts for various factors to simulate natural NSB under conditions devoid of clouds, the Moon, and planets. A key innovation of GAMBONS is its seamless integration of the Gaia-DR3 and Hipparcos catalogs, combined with detailed modeling of di use galactic and extragalactic light, zodiacal light, airglow, and atmospheric attenuation and scattering effects. The model operates with default airglow and aerosol parameters, providing a reliable initial approximation of the typical conditions at our monitoring sites.

Using these parameters, GAMBONS generates an all-sky NSB map for a specified time and location. The presence or absence of the Milky Way and the elevation of the ecliptic above the horizon are the primary factors influencing this model. Additionally, the model can estimate synthetic NSB values in different photometric pass bands, such as the SQM, TESS, or specific photometric bands. This feature is precious for theoretically assessing natural light contributions and comparing them with observed data, thereby facilitating the estimation of the percentage of ALAN at each site.

Table 5 compares values measured with the SQC at the zenith and all-sky, alongside the values obtained from the GAMBONS model (GMB). The final column indicates the estimated percentage of ALAN at each location. This estimation is derived from the model’s values, with conversions following the methodology outlined in ^{Kolláth et al. (2020)}. It is important to note that the GMB model tends to overestimate natural NSB near the horizon and underestimate it at the zenith by up to 0.1 mag/arcsec² (^{Masana et al. 2022}).

TD and BBG are the only sites for which prior ALAN data were available in the literature, all recorded using the SQM photometer (21.26 mag/arcsec² for TD and 16.31 mag/arcsec² for BBG). These historical data are utilized here for a direct comparison with our SQC measurements, providing a potential means to estimate the evolution of ALAN over the past years (^{Urrego Guevara 2016} (hereafter ^{UG, 2016}); ^{Góez Therán & Vargas Domínguez 2021} (hereafter ^{GT&VD, 2021})).

Table 6 compares the values reported in the literature (SQM LIT) with those obtained using the GMB models (SQM GMB) for the same location, date, and time. While direct measurements taken with a SQM were not available in this study, the SQC methodology enables the calculation of a synthetic SQM value based on SQC data. This synthetic value is included in the SQM LIT column. Conversely, the SQM GMB column reports the results of the GMB model simulations for the dates and times analyzed in this study.

Although a direct comparison is challenging due to variations in dates, times, measurement locations, instruments used, and meteorological conditions, the obtained values provide valuable insights into the evolution of NSB at both locations. To account for intrinsic and previously unknown variations in natural conditions, the original SQM measurements reported in the literature and our synthetic values derived from SQC were compared with the theoretical predictions generated by the GMB SQM model.

For this purpose, a column labeled Δm is included, representing the difference in NSB between the SQM GMB and SQM LIT values. Similarly, the corresponding row includes ΔM, which indicates the time difference between the SQM GMB and SQM LIT values. These metrics not only facilitate the evaluation of discrepancies and similarities between the simulated data and the literature-reported data but also enable an analysis of the temporal evolution of NSB.

Figures 19 and 20, generated using VIIRS Day/Night Band (DNB) data sourced from the Light Pollution Map platform, illustrate the temporal evolution of NSB for the years 2014 and 2022, respectively. Figure 19 covers an area of 1600 km² and shows a radiance of 2840.6 · 10^-9 W/cm²·sr in 2014, decreasing to 1595.3 ·10^-9 W/cm²·sr in 2022. This suggests a reduction in the way the desert is being illuminated, which can be visually observed as a decrease in radiance in the lower-left part of the image. However, this area corresponds to the Dina Field, one of the oil and gas production fields operated by Ecopetrol, the main oil and gas company in Colombia. This observed reduction may be attributed to the adoption of LED technology, the predominant lighting source during the measurement period. However, Table 6 reveals an increase in NSB, whereas VIIRS data suggest an apparent decrease. This discrepancy is primarily due to the limited sensitivity of the VIIRS DNB channel, which excludes wavelengths predominantly emitted by LEDs. Consequently, NSB tends to be underestimated in regions dominated by LED lighting. For a comprehensive discussion on the characteristics and limitations of VIIRS, refer to ^{Cao & Bai (2014)}.

Fig. 19 Comparison of NSB using VIIRS/DNB radiance data in a 40x40 km area centered on the 2022 measurement point at TD (red pin) for the years 2014 (left) and 2022 (right). The color figure can be viewed online. 

Fig. 20 Same as Figure 19, but focused on BBG and centered within a 2 × 2 km area. The color figure can be viewed online. 

On the other hand, Figure 20 covers an area of 4 km² and shows an increase in radiance, from 796.0 · 10^-9 W/cm²·sr in 2014 to 905.3 · 10^-9 W/cm²·sr in 2022. The increase in NSB is attributed not only to the 558 lamps²⁹ within the BBG, but also to the contributions from all the surrounding neighborhoods.

Figures 21, 22, 23, and 24 display the all-sky images generated by the GAMBONS model, alongside the images obtained by the SQC for each site. For each GAMBONS image, spatial and temporal data were replicated, while for the SQC image the NSB value was recalculated, including the contributions of the resolved stars without applying smoothing. A detailed description of each site is provided below.

Fig. 21 Comparison of the NSB all-sky images generated by the GAMBONS model (left) and the SQC (right) from AOUTP. For more details, see § 4.3 and Table 5. The color figure can be viewed online. 

Fig. 22 As in Figure 21, but for TD. For more details, see § 4.3 and Table 5. The color figure can be viewed online. 

Fig. 23 As in Figure 21, but for BBG. For more details, see § 4.3 and Table 5. The color figure can be viewed online. 

Fig. 24 As in Figure 21, but for CG. For more details, see § 4.3 and Table 5. The color figure can be viewed online. 

In the AOUTP case, as illustrated in Figure 21, 98% of ALAN in Pereira originates from artificial contributions to NSB. This dominance is evident in Table 5, Column 7, where the impact of artificial lighting on sky glow is clearly observed. ALAN results from a combination of natural NSB contributions, estimated using the GAMBONS model, and artificial NSB contributions, derived from our data. The overwhelming presence of artificial light high-lights the profound transformation of the night sky in urban environments like Pereira. In Figure 22, it is evident that, with the naked eye, one can still appreciate the stars and our galaxy, the Milky Way in TD. However, the NSB from the cities almost completely dominates the horizon. Table 5 indicates that approximately 59% of the sky is dominated by ALAN, which is highly concerning for a Starlight destination whose mission is to conserve and protect the night sky.

Raising awareness among people is crucial, as most inhabitants make their living from tourism, with astrotourism being the main attraction for visitors to the desert (^{Montealegre Vega & Garavito González 2022}). Efforts should prioritize Ecopetrol, a private company, and the city of Neiva, as these are the primary sources of light pollution in the desert. This is especially important during lighting transitions, such as the shift to LED technology. While LEDs provide significant economic bene ts and enhance conditions for nighttime activities, it is essential to ensure these lights are properly directed toward the ground and avoid over-illumination of space. Protecting the night sky is not only vital for environmental conservation but can also boost the local economy by attracting more tourists interested in astrotourism, thereby fostering sustainable development in the region.

Figure 23 shows that the sky at the BBG is dominated by ALAN. This result was predictable, given that BBG is located within one of the largest metropolises in Latin America. However, this does not imply that the night sky should not be protected. In this particular case, studies are recommended to evaluate how ALAN is affecting the plants in the various ecosystems present in the garden (^{Cadena-Vargas et al. 2020}). Additionally, measures should be taken to protect those collections and plants that are highly sensitive to light changes.

Moreover, the sky over CG (see Figure 24), like that of the BBG, is dominated by ALAN due to its proximity to the city of Bogotá. However, a 5% decrease in ALAN is observed compared to BBG. This is because CG is 700 meters higher in elevation and almost 10 km away from the urban center. This difference in altitude and distance allows for better night sky quality in CG, reducing light interference. Additionally, the area’s vegetation and natural relief help to mitigate light pollution. For these reasons, CG could be considered a strategic point for astronomical observation and ecological studies. In summary, the analysis of NSB using the GAMBONS model highlights the significant impact of ALAN on astronomical observations and ecological systems. Locations like the TD and BBG demonstrate varying levels of light pollution, emphasizing the need for targeted measures.