How to see deployed Starlink "Racetrack" flares

flarkey

Senior Member.
Staff member
Would I get good footage from where I am (on Tropic of Cancer in Baja California Sur)? Low light pollution and usually clear nights.

If so, tell me a time and date and I'll see what I can see.

hey @Rory - I now have a bit of a better understanding of how to see Starlink Flares, and in fact last night I saw three (one really bright and two others) after looking in the right place at the right time. This seems to be...
  • At the time that the sun is around 40° - 45° below the horizon (either after dusk or before dawn)
  • Look towards the direction of the sun (beyond the horizon) at this time.

So how do we find this info? The easiest way online is to use https://www.timeanddate.com to see the sun position for a location and time.

For your location of Baja California Sur, find a nearby city... I've selected Cabo San Lucas. On the Sun position graph move your cursor to either a time after sunset or before sunrise, and note the time it crosses the 40° line. The direction of the sun , i.e. the direction to look, is then given (highlighted in yellow below). The Starlink flares will appear here and will be pretty close to the horizon (within 5° to 10°), so you'll need a good view in that direction (out to sea would probably be best).

https://www.timeanddate.com/astronomy/mexico/cabo-san-lucas
1670496988776.png

So, if you have a free evening in the next few weeks and the weather is favourable - grab your favourite beverage, go outside at about 8.30pm and look West for about half an hour. You might see a starlink flare or two. Let me know if you have any success. ;)

Restrictions: There's a restriction on this - the observer's location, time of year, and the Starlink orbit inclination will affect the ability to see flares in paricular direction. Starlink flares are typically inclined at 53° so anyone who is located close to these N/S latitudes will only be able to see flares that are to the East or West of them. So for example, someone in London will be able to see flares in Winter as the sun is at -40° in the early morning and late evening, which puts it on either a W or E direction - where there are lots of Satellites. Conversely in the summer the sun may only get to -40° around midnight, which will be almost North, where there will not be satellites in the right position to make flares.W hereas, someone in Dakar (further south) may be able to see flares to the North East and West (depending on the season).

1670582495198.png

ffdit: Added restriction
 
Last edited:
Ay! Just under two weeks ago I left a place where I was living literally a two minute walk from a Pacific beach with uninterrupted views to the west.

Probably be back there at some point though to give it a try. Thanks for digging out the info.
 
I've been refining @Mick West 's GeoGebra model of the Starlink Flares as observed from aircraft (link: https://www.geogebra.org/calculator/fjpzsmws ) to try and understand the geometry of Surface observations.

This is my model: https://www.geogebra.org/calculator/xqtqgqzr

I have made it to scale, that is - the average radius of the earth of 6.371km is on the x & y intersection at 6.37, and the altitiude of the Satellite orbit of 550km put it at 6.871. I have put the observer's location at just above the earth's surface. This will account for terrain elevation and some oblateness of the earth. The satellite can be moved to show the conditions upon which a flare might be obsereved. The model calculates the angle to the sun, the elevation angle to the satellite and the distance on the earth from the observer where the satellite is overhead when the flare occurs.



1670779135877.png

A few things I have noted...

Sun Angle
We had previouslly noted that the angle to the sun (over the horizon) needed to be over -40 degrees for flares to occur. This is partly true. The angle is compounded with the elevation to the flare from the observer. The magnitude of flare elevation and the sun depression will always add up to 45 degrees.

Distance from observer
The distance to the flaring satellite is around 2000 km

Things I havent included in the model:

Atmospheric refraction.
The sun's light refracts about 1 degree throught the atmosphere to an observer on the ground. I'm not sure of the full effect on the angle to satellite.




So how much does atmospheric refraction impact on the times of sunrise and sunset? When the sunlight from the Sun's upper edge reaches the horizon at sunrise and sunset (see Figure 3), the actual altitude of the centre of the Sun is about -0o50’[1] (negative value means below the horizon). About -0o16’[2] of this actual altitude is the apparent radius of the Sun. The remaining -0o34’[2] is the effect of atmospheric refraction under average atmospheric condition, leading to the apparent effect of advanced sunrise and delayed sunset. For example in Hong Kong, the atmospheric refraction causes the sunrise and sunset to appear about 2 minutes early and late respectively when compared to the situation without the atmospheric refraction. In fact, the times of sunrise and sunset announced by the Hong Kong Observatory have already included the effect of the atmospheric refraction so that users do not need to calculate the times by themselves.

Source: https://www.hko.gov.hk/en/education...ction-on-the-times-of-sunrise-and-sunset.html

Altitude of Satellite
Although the starlink satellite publishrf altiitude is 550km there will be variations in each individual satellite's altitude (540 to 571km). This will affect the geometry of the flaring conditions. I don't know how this will affect the model.

If anyone wants to amend this model, please feel free. Comments and peer review are of course welcome.

Edits: Typos and improvements to the model
 
Last edited:
Starlink Flare Spotting Update!

I've just been outside in the cold and on dark clear night here in the UK, and following the advice I gave above - looking in the direction of the sun when it was about 40° below the horizon (almost due West at 8pm) - I must have seen about 20 Starlink Flares in a 15 minute period. They were as bright as the star Altair , which is magnitude 0.8, so easily visible from the English countryside. And if I'm honest, at times they did actually look like circling Aircraft.

Pretty cool to have gone full circle on the Racetrack UFO Flap and now being able to predict them. Unfortunately I didn't manage to capture any with my phone camera. Might have to put a low light camera on my Christmas list.
 
Was there a distinct start and end time? (which would correspond to start and end sun elevation)
Hard to tell. I went out at 8pm and within a few mins I'd seen one, then another, and then a few together. I left at about 8.20pm because I was getting cold. They did seem to be getting lower as time went on.
 
I gave it a go, looking West from East of Sacramento at 8:15PM. I saw nothing. Possibly the trees covered to much of the sky. Possibly too much light pollution, as this is looking toward a densely populated area.

2022-12-12_23-18-05.jpg
 
Hmm. Ok might have to refine the spotting parameters. It might be, like you suggested previously, that we have to look towards the extents of the orbits. I'm at 52°N, so West is directly in line with this. Sacramento is about 38°N so you may need to look more North West...? Not sure yet.
 
Hmm. Ok might have to refine the spotting parameters. It might be, like you suggested previously, that we have to look towards the extents of the orbits. I'm at 52°N, so West is directly in line with this. Sacramento is about 38°N so you may need to look more North West...? Not sure yet.
Wouldn't it be the opposite? Someone near the equator would only have the hope of finding the sun below the horizon due east or due west. Nearer the north pole, in summer I've got a sun just out of sight due north at midnight. Were there starlink in such orbits, I'd expect to see them flare up north, as their annulus is outside the viewer's annulus. Wherever you are, the satellites are flaring further in that direction.
 
Wherever you are, the satellites are flaring further in that direction.
As with most things - it's not quite that simple. It is dependant upon the time of year and your latitiude. Where I am (eastern England latitude 52 °)) I can only see these flares in winter because in summer the sun only gets to ~40 ° below the horizon when it is north of me, but there isn't a high enough concentration of satellites to create the repeated flaring.



Conversely, the recent flap in Porto Alegre in Brazil (latitude -30 ° and in summer) was centred around due south and at midnight. However, in Porto Alegre in the height of summer (around now) the sun doesn't go below 40 ° even at midnight so they won't see them at all.
1670922561197.png
 
Last edited:
As with most things - it's not quite that simple. It is dependant upon the time of year and your latitiude. Where I am (eastern England latitude 52 °)) I can only see these flares in winter because in summer the sun only gets to ~40 ° below the horizon when it is north of me, but there isn't a high enough concentration of satellites to create the repeated flaring.

You've changed your argument from a qualitative one (where can they be) to a quantitative one (how many will there be).
 
Hmm. Ok might have to refine the spotting parameters. It might be, like you suggested previously, that we have to look towards the extents of the orbits. I'm at 52°N, so West is directly in line with this. Sacramento is about 38°N so you may need to look more North West...? Not sure yet.
I don't think so. Turns out my iPhone compass is off by 13° when set to use True heading, which oddly is the amount of magnetic declination here. Like it corrected for it twice.

So, I was zoomed in on the red area, which I was happy to see was clear down near to the horizon. But I should have been looking at the green, which sadly is covered by trees.
2022-12-13_04-32-17.jpg
 
You've changed your argument from a qualitative one (where can they be) to a quantitative one (how many will there be).
"Changed"...? I'd say 'refined'. ;) I accept I have added to the conditions under which they will be seen. My initial conditions only applied to the observer's location and direction of view, but the rate of occurrence of flares is directly related to the number of satellites that will be in that part of the sky. This is a complex model and is pseudo-random - in reality it is completely predictable and is dependent upon the satellites orbit. I'm not yet (and doubt I ever will be) able to predict which exact satellite is visible where and exactly when. But that is do-able by someone with the right programming skills.
 
I don't think so. Turns out my iPhone compass is off by 13° when set to use True heading, which oddly is the amount of magnetic declination here. Like it corrected for it twice.

So, I was zoomed in on the red area, which I was happy to see was clear down near to the horizon. But I should have been looking at the green, which sadly is covered by trees.
I think youre right . I think the star to the right in your picture is Altair...

1670936947366.png
Altair is a good reference point to use as the point directly above the sun. All the flares that I saw were right of and below Altair.

1670937247157.png
 

Attachments

  • 1670936953595.png
    1670936953595.png
    649.1 KB · Views: 89
I think youre right . I think the star to the right in your picture is Altair...

1670936947366.png
Yup, if I just shift the image over 13° then Altair is exactly right at that time, and the Sun is at -40°
2022-12-13_08-20-25.jpg

I guess I'm going to have to get my chainsaw out...

Altair there is at +8° elevation, what's the highest Starlink flare we've seen?

I might try again tonight, wider angle lens. I'm using a Sony A6400, which has pretty good low-light abilities. Might catch something through the trees.
 
There's a lot of foreground lights in the night shot, the webcam may not be picking them up, removing them as noise etc.
 
It may be that the conditions for the flares are not good for California?
I tried my balcony viewpoint again tonight, looking West at 8:16PM, Sun at -40°.

Got nothing. A couple of what looked like faint satellite flares, but after syncing it with a Stellarium replay, they did not appear to be Starlink. Too high, and descending (All the starlinks we rising)

Here's the only flare, in the upper left region, not very bright. Time would be 20:23:00

Regarding light pollution. Altair is visible in the tree at the start, and the dim constellation Sagitta is visible in the upper right

But in the full clip, I see the star Almizan IIi (Magniture -1.52) descending, but when it gets below about 4° it is no longer visible. This may be related to my low altitude (250 feet above sea level), low level haze, and looking in the direction of a lot of light pollution.
 
I managed to capture these three from the Norfolk countryside. Saw a total of 5 in about 2 mins of spotting.



These are three satellites, seen at 20.14UTC

 
Last edited:
Saw about another 30 flares tonight between 8 - 8.30pm. I used @Creamy Pasta 's realtime Flare plotter ( https://a-j-k.github.io/starlink/index.html ) and it worked really well - once I'd worked out how to use it. Very accurate plots - I was able to use it to expect a flare and look in that particular spot. Didn't get any good videos, but for a few momemts I did see 4 flares at once.
 
I was recently sent a timelapse video from a meteor-cam in the UK which has captured the Starlink flares quite well. This shows about 30 mins of activity looking West from about 1930hrs on December 26th.


 
Last edited:
I was recently sent a timelapse video from a meteor-cam in the UK which has captured the Starlink flares quite well. This shows about 30 mins of activity looking West towards from about 1930hrs on December 26th.

Nice. I suppose in this configuration then people will not mistake them for "racetracks"
 
Nice. I suppose in this configuration then people will not mistake them for "racetracks"

Yeah, I think the apparent appearance is dependent upon viewers location and the perspective they get of the satellites orbit. For example the flares in this example are mostly vertical, but in the Jimmy Church video the flares are all moving horizontally so could be confused for aircraft on a racetrack pattern.
 
I was recently sent a timelapse video from a meteor-cam in the UK which has captured the Starlink flares quite well.
They've got a network of meteor watchers, and can triangulate the position from several locations. That enables them to find fresh meteorite landings within a very short time after they fall. I found out one of the leaders is a cousin of mine.
 
Note the imidiate "pavlov" wrong assumption in the tweet, of it impossibly being satellites:

seeing something crazy here, 3 bright dots that just keep increasing and decreasing in brightness? they're slowly moving around too, but they aren't satellites. so confused.
 
Note the imidiate "pavlov" wrong assumption in the tweet, of it impossibly being satellites:
Although this is a "wrong" assumption it is understandable because these flares are moving and flaring unlike any other satellites he has seen to date. It is equally understandable when pilots make the same error.
 
He posted a long exposure:
FnA4VleXoAAIc3C.jpg

And I did a "log exposure" of the Starlinks in Stellarium (Grey = theoretical sunlit portion, not the glares)
FnA71gVaAAEUtFg.jpg
That's not matching the time, but, it's a pretty good match for the two directions.
 
I love the certainty in "but they aren't satellites". Also, are you sure he's an astronomer, as "European Launch Correspondent for @TLPN_Official" sounds more like "journalist" to me.
He self-identifies as an Astronomer. Who am I to argue?

Screenshot_20230122_151530_Twitter.jpg
 
@Ravi - I think you're right, and I've commented in that Reddit thread already. However, as you can see in the video, the elevation is quite high, so although it probably is Starlink satellites flaring i dont think we'll see the Sun at 40 degrees below the horizon

The reddit post give 6.45pm on Feb 16 as the time of the video:

1677834117329.png

https://en.wikipedia.org/wiki/List_of_Starlink_and_Starshield_launches

I think this was probably Group 5-4 that was launched on 12 Feb.

1677833749493.png

I'll check the pass predictions when I have a little time later.
 
The video metadata says

Create Date : 2023:02:17 01:42:26
Modify Date : 2023:02:17 01:42:26

I assume these are in UTC?
 
Just an update on this - I've tried to find accurate Satellite orbital data for the Group 5-4 that was launched on 12 Feb for the time of the sighting ( 01:42am UTC on 17 Feb 23) but the closest I can find is the predicted data after the launch on the 12th. Will keep trying.
 
Last edited:
Back
Top