Salvador J. Ribas<=
span
lang=3DES style=3D'mso-bidi-font-size:13.0pt;mso-ansi-language:ES'>1,2,*, Jordi Torra2, Francesca Figueras2, Sergi Paricio3,
Ramon Canal-Domingo1
[1]<=
/a> 1Parc Astro=
nòmic
Montsec – Consell Comarcal de la Noguera, Àger E25691, Spain
2Institut de
Ciències del Cosmos (ICCUB-IEEC), Barcelona E08028, Spain
3Servei per=
la
Prevenció de la Contaminació Acústica i Lumínic=
a (Generalitat de Catalunya),
Barcelona E08028, Spain
Received
04 December 2016, Accepted 30 December 2016
Abstract
Lig=
ht
emitted or reflected up to the sky can interact with clouds or fog changing
Night Sky Brightness (NSB). So the evaluation of NSB in any place can be
affected by this meteorological features and needs to be analyzed. To evalu=
ate
this effect, data of NSB obtained with the Catalan Light Pollution Network
(XCLCat) has been analyzed accurately and shows how the effect is completely
different in urban and protected areas. The study, including for the first =
time
independent cloud coverage data from ceilometer in combination with Sky Qua=
lity
Meters (SQM), shows how in a city NSB increases dramatically and in a prote=
cted
area the sky appears darker than in cloudless situation.
1.
Introduction
Light
Pollution (LP) is an effect generated by artificial light at night (ALAN).
Normally LP is associated with astronomy, but there are many other effects =
in
natural environment, health or power consumption [1, 2, 3, 4]. One of the m=
ost
common features of LP is the emission of light to the sky, generating skygl=
ow.
The
evaluation of LP can be done measuring Night Sky brightness (NSB) using
ground-based instruments as telescopes with cameras or stand-alone devices =
for
this purpose. Light emitted or reflected up to the sky can interact with cl=
ouds
or fog changing dramatically NSB (see for example [5, 6, 7]). The interacti=
on
of artificial lights with clouds can only be evaluated with ground based
measurements because this interaction starts in the base of clouds layer, s=
o any
observation above the clouds will not provide clear information of this
interaction. So the evaluation of NSB from the ground can be clearly affect=
ed
by presence or absence of clouds. This effect is completely different depen=
ding
of the nature of the site: dark site or urban polluted site.
2.
Catalan Light Pollution Network (XCLCat)
There
are different approaches to evaluate NSB to estimate LP: from the space or =
from
the ground. The first option is based in the idea of using satellite data [=
8]
or data from the International Space Station [9]. The ground based observat=
ions
can be done using classical photometry with telescopes or using new special=
ly
designed devices. In our study we have worked with ground-based measurements
using Sky Quality Meter (SQM) devices. As it is explained above, ground
measurements are mandatory to evaluate interaction of artificial light with
clouds.
The
cooperation of Catalan Service against Light Pollution and Parc Astron&ogra=
ve;mic
Montsec (PAM) has leaded to a pilot plan of a network for monitoring and
evaluation of NSB in different places of Catalonia region [10]. Initially t=
he
network had just five devices but in 2016 the network increased to nine dev=
ices
and in the near future is expected to be integrated in the Air Quality Netw=
ork
(XVPCA) of the Government of Catalonia. Each station of the network is base=
d on
SQM devices, mainly SQM-LE version that works connected to Ethernet socket.=
SQM
devices are used pointing to zenith and also with its L (lens) versions hav=
e a
wide field of view of around 40¨¬ [11, 12]. The spectral response is close =
to photopic
eye response but is more sensitive to blue part of visible light, so this m=
eans
its response is not exactly the photopic band nor the V Johnson astronomical
filter band. Luminances are reported in magSQM/arcsec2 but can be converted=
to
mcd/m2 with some assumptions (see for example [13, 14]).
All
the procedure of acquisition and preprocessing of data of XCLCat stations is
done using PySQM software [15]. Currently the devices are installed with two
main focuses: natural or dark sky protected areas and places that contribut=
e to
NSB of these protected sites (see Fig.1 and Table 1).
Fig.
1. Map of the current XCLCat stations (2016). The highest concentration is =
in
STARLIGHT protected area of Montsec
Table 1. Mean Values of NSB
during astronomical night in 2015 for the first five operative stations of =
XCLCat.
It
is clearly visible how in Lleida or Barcelona there is no difference when
moonlit data are not included, so the lunar cycle will be poorly visible. In
the other hand the first two sites listed, belonging to PAM sites, are
extremely dark and the Moon effect is clear.
|
Station |
GPS Position Latitude & Longitude=
|
Population (2014) |
NSB Total Sample |
NSB Moonless Sample |
|
|
|
|
|
|
|
Àger – Centre d¡¯Observaci&o=
acute; de l¡¯Univers (COU) |
42.02444N 0.73611E |
5=
94 |
20.24 |
21.38 |
|
Sant Esteve de la Sarga – Obs. Astr. del Montsec (OAdM) |
42.05140N 0.72962E |
1=
32 |
20.25 |
21.48 |
|
Balaguer |
|
|
18.13 |
18.43 |
|
Lleida |
|
|
16.67 |
16.71 |
|
Barcelona – Fac. Física |
|
|
16.67 |
16.75 |
All
the instruments of the XCLCat have been intercalibrated in the specially
designed station of Centre d'Observació de l'Univers. Intercalibrati=
on consists
in a cross-calibration method based in pairing all the instruments, so with
this strategy all the devices will show the same value for the same situati=
on.
The initial accuracy of a SQM is around 10% and with the intercalibration c=
ould
be improved to 1% [10, 16, 17].
3.
Interaction of clouds with NSB
One
of the difficulties of compare NSB data with clouds is the quality of clouds
data, sometimes only synoptic information has been used [5] and not direct
measurements of clouds properties. Fortunately XCLCat has some stations that
are installed in areas used for environmental studies. This is the case of
Montsec Starlight protected area and the city of Barcelona. Both places have
installed a ceilometer, an IR laser device that provides real time informat=
ion
of clouds and aerosols [18]. The ceilometer installed in Montsec is a Jenop=
tik
CHM15K Nimbus that measures up to 14 kilometers and the ceilometer installe=
d in
Barcelona is a Vaisala CL13 that operates up to 8 kilometers [10]. So for t=
he
first time SQM and ceilometer data have been combined to evaluate the inter=
action
of clouds and NSB, and it has been done in completely different situations =
of
LP: dark and urban places.
The
presence of clouds and fog generates oscillations in NSB values obtained wi=
th
SQM's. In the absence of clouds NSB is more stable in comparison with the
presence of clouds as is shown in Fig.2. In the case of cities or light
polluted areas the effect of clouds is to increase the brightness of the sk=
y [5,
6].
Fig. 2. Comparison of two nights of measurements
with SQM. Top panel shows a night with clouds and bottom panel shows one
without clouds. The night with clouds shows some oscillations on the NSB
created by clouds.
The
effect of increasing NSB is extremely visible in case of fog, also without =
the
use of specific clouds data, just with evaluation of NSB in yearly plots as=
in
Fig.3. This is the case of the cities of Balaguer, a small city with not ve=
ry
high LP, and Lleida, a medium size city placed 25 km south of Balaguer. Both
cities are located in an area where it is possible to have huge episodes wi=
th
permanent fog. In Fig.3 are clearly visible three periods with permanent fog
showing an important increase of NSB up to 14-15 magnitudes per square
arcsecond (red vertical stripes). It is also well identified because the red
stripes (fog episodes) are completely simultaneous in both cities and could=
not
be related to any local effect. This fog effect is not visible in the case =
of a
site located in the mountains of Montsec, as the station placed in Observat=
ori
Astronòmic del Montsec (OAdM) at 1600 meters above sea level (asl). =
In Fig.4
all the lunar cycles are clearly visible as diagonal stripes and, because t=
his
site is one of the darkest in southern Europe, NSB reaches values as dark a=
s 22
magnitudes per square arcsecond.
Fig. 3. Evolution of close=
to
one year of NSB data in two sites. Top panel shows data from Balaguer and
bottom panel from Lleida. In Balaguer few lunar cycles are still visible but
not in Lleida. The most visible features are the three red vertical stripes
related to fog episodes.
Fig. 4. Evolution of close=
to
one year of NSB data in a very dark place, OAdM (c. 1600m asl). All the lun=
ar
cycles are clearly visible as diagonal bands and there are not fog episodes=
in
mountain location.
Another
possible approach to detect clouds or fog effects is evaluating how NSB is
distributed during the night, with accumulative data of one year, showing a
different shape depending of the amount of LP present at the site. In Fig.5=
it
is plotted the distribution of NSB in a protected dark area as OAdM site an=
d it
is clearly visible how the main distribution is around 21-22 mag and scatte=
red
measurements are moving to dark part of the plot (higher magnitudes). This
scattered data could be related to measurements with clouds, so this is a f=
irst
probe of clouds darkening in dark sites. This effect of darkening linked wi=
th
clouds has been also recently detected with SQM data in Lake Stechlin (Germ=
any)
[7].
Fig. 5. Distribution of NSB data during nightti=
me
for one year in OAdM (c. 1600m asl). The data are mainly centered in 21-22 =
mag
with scattered measurements moving towards the darkest (top) possible value=
s.
Opposite
to what is happening in dark sky places, in Fig. 6, there is the case of
Barcelona with a brighter distribution dominated by measurements around 17 =
mag
and scattered values appearing in the bottom part of the plot, reaching ext=
reme
values up to 13-14 mag. The origin of this effect is directly linked with
clouds and how they reflect artificial light at night and how clouds are
amplifying ALAN effects on NSB [1, 2].
Fig. 6. Distribution of NSB data during nightti=
me
for one year in the city of Barcelona. The data are mainly centered in 17-18
mag with scattered measurements moving towards the brightest values (13-14
mag).
Because
the stations placed in Parc Astronòmic Montsec (COU and OAdM) and in=
Barcelona
have a ceilometer device, there is available information of the presence of
clouds that can be combined with NSB measurements from SQM devices. So addi=
ng
available information of ceilometer data is possible to do a combined study=
to
evaluate accurately how clouds can change NSB in both situations. The study=
of
combined data in Barcelona city center, in the period August-November 2015,=
leads
us to a clear example of how the clouds are amplifying the effect of LP. In=
the
case of Barcelona, as seen in Fig.6, NSB could be six times brighter (more =
than
2 mag) with clouds in comparison of a clear (no clouds) night.
It
is possible to compare the effect of different kinds of clouds (see Table 2=
).
Ceilometer data provides the height of clouds and it is possible to combine=
NSB
data, not only with presence or absence of clouds, with the levels of clouds
(low, medium and high). This analysis shows how low clouds are clearly lead=
ing
to the biggest increase of the effect of light in an extremely polluted are=
a as
Barcelona. In the other side high clouds are generating a tiny effect on the
measurement of NSB in this city.
Table
2. Mean magnitudes of NSB measured in Barcelona with different configuratio=
n of
clouds
|
Sample=
|
Number=
of
measurements |
NSB |
smag |
|
|
|
[mag/arcsec2] |
[mag/arcsec2] |
|
Total |
27645 |
16.79 |
1.21 |
|
Without Clouds |
14375 |
17.71 |
0.40 |
|
With
Clouds |
13270 |
15.79 |
0.99 |
|
Low Cl=
ouds |
8427 |
15.53 |
0.96 |
|
Medium
Clouds |
3643 |
16.07 |
0.84 |
|
High
Clouds |
1200 |
16.81 |
0.63 |
For
the first time the evaluation of this effect in a natural protected area is
done. The case study is Montsec mountain range as specially protected dark =
sky
area. In this place clouds are not amplifying the effect of light. The effe=
ct
is just in the opposite direction because clouds can block natural sources =
and
NSB can be reduced to extremely dark measurements (see Fig. 5). In this case
the period evaluated with ceilometer data is from November 2014 to April 20=
15
and also different samples are available to evaluate NSB in different
situations according to different kinds of clouds (see Table 3).
Table
3. Mean magnitudes of NSB measured in OAdM with different configuration of
clouds
|
Sample=
|
Number=
of
measurements |
NSB |
smag |
|
|
|
[mag/arcsec2] |
[mag/arcsec2] |
|
|
66492 |
21.47 |
0.59 |
|
|
31752 |
21.44 |
0.56 |
|
With Clouds |
34740 |
21.50 |
0.62 |
|
Low Clouds |
7837 |
21.82 |
0.88 |
|
Medium Clouds |
15537 |
21.50 |
0.60 |
|
High Clouds |
11366 |
21.29 |
0.21 |
The
effect of darkening by clouds is especially important in OAdM site in the t=
op
of the mountain, without no lights in the surrounding, where low clouds sam=
ple
is showing the darkest mean value, around 0.4-0.5 magnitudes darker than the
sample without clouds. It is also possible to see, in Fig. 7, how the low
clouds are leading to extreme NSB values (up to 24 mag) what is the zero le=
vel
of SQM device.
Fig. 7. Distribution of NSB
measured with presence of clouds. The different subsamples are plotted and
clearly low clouds can lead to darkest possible measurements of the SQM dev=
ice.
4.
Conclusions
For
the first time Catalonia has a permanent network to evaluate NSB, currently=
it
is equipped with nine stations. All the SQM devices have been previously
intercalibrated. These devices are installed in a mix of protected and urban
areas so comparison studies can be done. The presence of fog or clouds modi=
fies
the measurements of NSB obtained with SQM. Using only SQM data it is easy to
identify intensive fog periods in the cities because fog generates a huge
increase of NSB in light polluted areas. Measurements could reach 14-15
magnitudes per square arcsecond. Also without ceilometer data it is observed
how in the protected areas scattered measurements are moving towards the
darkest part of the plot. In the other hand in light polluted areas scatter=
ed
values are appearing in the brightest parts of the plots.
For
first time a combination of SQM and independent cloud data from a ceilometer
has been used. With this combination of data any clouds effect is easy to
determine. In the case of Barcelona (light polluted area) measurements show=
an
increase up to 2 magnitudes of NSB due to the presence of low clouds. Also
medium and high clouds have an impact increasing NSB. In the case of Montsec
(protected area) clouds can darken the sky giving the darkest possible valu=
es
of an SQM. The reason is clear, in a protected area without LP: clouds are =
not
reflecting artificial lights and are also blocking natural features as Milky
Way, Zodiacal light or stars from the sky, generating an important darkenin=
g of
the sky.
Acknowledgements
Authors
want to thank the institutions involved in the project of measuring and
evaluating LP in Catalonia: Parc Astronòmic Montsec - Consell Comarc=
al de la
Noguera, Institut de Ciències del Cosmos - Universitat de Barcelona =
(ICCUB-IEEC)
and Servei per la Prevenció de la Contaminació Acústic=
a i Lumínica (Generalitat
de Caalunya). Also authors want to thank all the institutions and researche=
rs
that have provided ceilometer data: Dr. Yolanda Sola from Meteo Group of
Universitat de Barcelona and Dr. Marco Pandolfi and Dr. Andres Alastuey from
Institut de Diagnosi Ambiental i Estudis de l¡¯Aigua (IDAEA) of Consejo Sup=
erior
de Investigaciones Científicas.
Authors
want to thank the anonymous reviewer and editors for their comments to impr=
ove
this paper.
SJR
acknowledges the European COST Project ES1204 Loss of the Night Network for=
the
chance to network and improve his research on this topic.
References
[1] Longcore, T., & Ric=
h,
C. (2004). Ecological light pollution. Frontiers in Ecology and the
Environment, 2(4), 191-198.
[2] Hölker, F., Moss, T., Griefahn, B., Kloas, W., V=
oigt, C.
C., Henckel, D., ¡¦ Tockner, K. (2010). The Dark Side of Light: A Transdisciplinary Research Agend=
a for
Light Pollution Policy. Ecology and Society, 15(4). Retrieved from
http://www.ecologyandsociety.org/vol15/iss4/art13/
[3] Gaston, K. J., Visser, =
M.
E., Hoelker, F. (2015). The biological impacts of artificial light at night:
the research challenge. Philosophical Transactions of the Royal Society of
London. Series B: Biological Sciences, 370.
[4] Sánchez de Migue=
l, A.,
Zamorano, J., Castaño, J. & Pascual, S. 2014. Journal of Quantit=
ative
Spectroscopy and Radiative Transfer 139, 109.
[5] Kyba, C. C., Ruhtz, T.,
Fischer, J., & Hölker, F. (2011). Cloud coverage acts as an amplif=
ier for
ecological light pollution in urban ecosystems. PloS one, 6(3), e17307.
[6] Kyba, C. C. et al. (201=
5).
Worldwide variations in artificial skyglow. Scientific reports, 5, 8409.
[7] Jechow, A., Hölker=
, F.,
Kolláth, Z., Gessner, M. O., & Kyba, C. C. (2016). Evaluating th=
e summer
night sky brightness at a research field site on Lake Stechlin in northeast=
ern
Germany. Journal of Quantitative Spectroscopy and Radiative Transfer, 181,
24-32.
[8] Falchi, F., Cinzano, P.=
, Duriscoe,
D., Kyba, C. C., Elvidge, C. D., Baugh, K., ... & Furgoni, R. (2016). T=
he
new world atlas of artificial night sky brightness. Science advances, 2(6),
e1600377.
[9] de Miguel, A. S. (2015).
Variación espacial, temporal y espectral de la contaminación =
lumínica y sus
fuentes: Metodolog©¥a y resultados (Doctoral dissertation, PhD thesis,
Universidad Complutense de Madrid).
[10] Ribas, S.J. (2015).
Caracteritzacio de la Contaminacio Lumnica en Zones Protegides i Urbanes. P=
hD
thesis, Universitat de Barcelona, Barcelona.
[11] Cinzano, P. (2005). Ni=
ght
sky photometry with sky quality meter. ISTIL Int. Rep, 9.
[12] Cinzano, P. (2007). Re=
port
on Sky Quality Meter, version L.
[13] Puschnig, J., Posch, T=
.,
& Uttenthaler, S. (2014). Night sky photometry and spectroscopy perform=
ed
at the Vienna University Observatory. Journal of Quantitative Spectroscopy =
and
Radiative Transfer, 139, 64-75.
[14] Hänel, A. et al. =
(2017 (in
preparation)). Current techniques and problems involved in measuring the ni=
ght
sky brightness.
[15] Nievas Rosillo, M.,
Zamorano Calvo, J. (2014). PySQM the UCM open source software to read, plot=
and
store data from SQM photometers. E-prints Universidad Complutense de Madrid=
. Available
at: http://eprints.ucm.es/25900/
[16] Bará, S. (2016).
Anthropogenic disruption of the night sky darkness in urban and rural areas.
Royal Society Open Science, 3(10), 160541.
[17] den Outer, P., Lolkema,
D., Haaima, M., Hoff, R. V. D., Spoelstra, H., & Schmidt, W. (2011).
Intercomparisons of nine sky brightness detectors. Sensors, 11(10), 9603-96=
12.
[18] Pandolfi, M. et al. (2=
013).
Continuous atmospheric boundary layer observations in the coastal urban are=
a of
Barcelona during SAPUSS. Atmospheric Chemistry and Physics, 13(9), 4983-499=
6.
S.J. Ribas et =
al.
/ International Journal of Sustainable Lighting 35 (2016) 32-39