Starlink satellite constellation: simulating its impact on optical observatories

SEA Icosaedro simulations

May and June 2020

For the moment, all of them are based on the constellation profile starlink.dat, that considers some 12 thousand satellites.

File: bv2020_microsatelites.pdf, report published in the Butlletin of the Spanish Astronomical Society (in Spanish language)

All these tests include a photometric model of intermediate complexity, that takes into account distance to the observatory, phase angle, extinction (0.12 mag/airmass was selected for this bunch of simulations) and geometry of the shadow cone of the Earth. This model is very similar to that of Hainaut & Williams (2020) and our results may be directly compared to theirs.

Affectation depends on the observatory latitude, but not on longitude. However, the main factors are the width of the field of view (FOV) and integration time (T). We performed tests for several different observatories available to the SEA community and this certainly illustrates latitude effects, but our results underline the strong importance of FOV and T and they are useful mainly to analyse these factors.

Observatories and latitudes (degrees):
Calar Alto      +037
Javalambre      +040
La Palma        +029
Montsec         +042
Paranal         -024

For each observatory (i.e., each latitude) we perform simulations of two kinds:

1) All-sky simulations:

Counting the number of satellites visible as a function of time along one complete night

--> For five different Sun declinations: +023, +012, +000, -012, -023 degrees

--> For two different elevations over the horizon: satelites visible above +000 deg, and above +030 degrees

So, in total, ten simulations are done at each location.

File naming conventions:

S_+LLL_+DDD_+HHH_NNN.xxx

S: means "Starlink"
+LLL: observatory latitude, degrees
+DDD: Sun declination, degrees
+HHH: Elevation over which satellites are counted, degrees
NNN:  Number of individual interations that are averaged out (050 in all cases)
xxx:  Type of file:

xxx = jpg, ps, pdf, graphs with number of visible satellites as a function of time, time is measured from the previous noon and is given in mintutes, vertical lines indicate the instants of beginning and end of civil, nautical and astronomical twilights, and midnight.

xxx = mp4, animation with the apparent magnitude histogram of visible satellites at one minute steps; background colour means: white in daylight, grey in civil twilight, light blue in nautical twilighg, deep blue in astronomical twilight, black during astronomical night

xxx = dat, files with very detailed information about the simulation, not included here, but may be provided, with indications about their contents

Example:

S_+037_+012_+030_050.mp4

S: Starlink
+037: Calar Alto latitude
+012: Solar declination intermediate north +12 degrees
+030: Counting satellites at 30 deg or more above horizon
050:  50 simulations were averaged
mp4:  Movie with the histogram of apparent magnitudes

2) Pointing-oriented simulations:

We select a FOV in arcminutes and an integration time T in seconds. Then, five observing directions are predefined:

N, S, E, W at 45 deg elevation, and zenith

For the given latitude we select Solar declination (0, +23, -23) and, both fixed, we study the five fields of view for three different solar elevations: -12, -25, -37, both PM and AM.

This means that for one given observatory (latitude, FOV, T), 5 x 3 x 3 x 2 = 90 configurations, see:

5 pointing directions
3 solar declinations
3 solar elevations
2 for am/pm conditions

This produces quite a large amount of information that is organised in form of detailed files and summary tables.

Detailed files are:

P_+LLL_+DDD_+HHH_pm_+aaa_+hhh_FOVi_iTim_NCRO.dat
p_+LLL_+DDD_+HHH_pm_+aaa_+hhh_FOVi_iTim_NCRO.dat

P: very detailed output, p: less detailed output
+LLL: observatory latitude
+DDD: Sun declination
+HHH: Sun elevation
pm = "pm" or "am"
+aaa: observation azimuth from the South (O = S; 90 = W, 180 = N, 270 = E)
+hhh: observation elevation (45 deg for NSEW, 90 deg for Z)
FOVi: field of view in arcminutes
iTim: integration time in seconds
NCRO: number of crossing (multiple shots are simulated until NCRO sat crossings are registered, or until 1000 shots have been simulated)

Detailed files are probably intersting only for very technical analysis, so they are not included here, but they may be provided upon request.

The main results are contained in pdf tables that are, hopefully, self-explaining.