My next project: tracking the ISS

[Vroobel]

Hi, 

As my GoTo is successfully finished (some cosmetic issues remain) I shoud focus my attention on planets' positions. I have proper source of information: fantastic book "Astronomical Algorithms" by Jean Meeus, thus I will sort the planets soon. But planets are not a challenge for me at this moment, they are just something obvious to do in my list. 

I have another idea and ambitious plan for next project within the year: locating and tracking the ISS to be able to make a video of its fly, not only transition. 

Similarly to other objects, I need some equations. I'm pretty sure they are available somewhere, because plenty websites or apps offer showing current position of the ISS. 

I will use this topic for sharing a progress ot the project. 

[JamesF]

Oddly enough I was considering how I might go about this earlier this evening, even adding it to one of my many "things to do" lists.

James

[Vroobel]

9 minutes ago, JamesF said:

Oddly enough I was considering how I might go about this earlier this evening, even adding it to one of my many "things to do" lists.

James

Just a coincidence

I will be happy to do it together with others. What kind of telescope do you have? 

[JamesF]

Just now, Vroobel said:

What kind of telescope do you have?

Well, I don't have a RASA or an RC

I'm guessing in my case the C9.25 would probably be the preferred choice, unless a wider field is required to stand a chance of actually finding the target.

James

[Vroobel]

46 minutes ago, JamesF said:

Well, I don't have a RASA or an RC

I'm guessing in my case the C9.25 would probably be the preferred choice, unless a wider field is required to stand a chance of actually finding the target.

James

I realised it was improper question. What kind of mount do you have? And what is your idea for the ISS tracking?

I think it's too fast to simply use any GoTo feature to catch it and follow, so my idea is to calculate its position for 5-10 minutes later, to aim the OTA there and to wait, then to start the tracking in exact time. 

Look please at this site:

https://www.n2yo.com/?s=25544

They evidently use some equations for achieving Ra and Dec of the ISS. I need them only for the job. 

Another question is its radial speed. 

Now look please at this video:

https://youtu.be/qROmsXf8-fg

I understand that the ISS pass about 0.5 degree within about 1 second, so my stepper motors will work really fast and the tracking will be similar to a slew... 

BTW, I have 10" Newtonian scope on a Dobsonian mount. 

Tomasz 

Edited August 25, 2019 by Vroobel

[JamesF]

I'll have a look at the videos later.  I'm in the observatory at the moment and my network connection from here is not great.

I can use either a NEQ6 or HEQ5 for this.

There are videos online of people tracking the entire ISS pass, which sounds like it would make imaging much more practical.  I'll see if I can find them when I'm back in the house.

James

[Vroobel]

Great! I'd like to ask an author or the site for his method of calculation the Ra and Dec. 

We have different mounts, so our ways will be different as well. I have the Raspberry Pi onboard and software written in Python completely by me, thus I can do whatever I want with my drives. But what are you gonna do to encourage your mount to find and track the ISS? 

Tomasz 

[JamesF]

At the moment I have no idea :D

First I think I need more information about the entire problem of tracking at speed.  I do recall the mount moving at quite a pace.  I imagine the sky model has to be pretty accurate to be able to hit the path of the ISS.

James

[Vroobel]

For you, as an owner of better EQ mount, it can be quite easy using dedicated software working with the mount in EQMODE (I'm sorry for mistake if happened - I don't have this kind of mount):

https://github.com/AstronomyLiveYt/SatTraker

I just seen the video about tracking the Starlink, looks good. 

https://www.cloudynights.com/topic/644609-presenting-satellite-chaser-a-free-software-made-for-tracking-and-guiding-on-satellites-using-ascom-compatible-equatorial-mounts/

or

I'm not so clever in EQ area, the links are given me by users of Polish astro-forum in similar topic which I created earlier. 

I have both easier and more difficult task. I don't need to use any MoveAxis mode/command to follow the object, just need to calculate its position and delays between subsequent steps. But how to get the sky model, as you wrote? 

Right, let's continue when we get something. Thanks James

Tomasz 

[Vroobel]

I've got it! 😊

https://www.n2yo.com/api/#positions

I gonna play with that tonight. 

[Guest]

Could be interesting for you all, https://sites.google.com/view/satellitechaser/home?authuser=0

[JamesF]

A quick hunt around this evening suggests that you just need to grab the latest TLE data for the ISS.  Then it's a case of finding it in the sky.

James

[Vroobel]

1 hour ago, artem said:

Could be interesting for you all, https://sites.google.com/view/satellitechaser/home?authuser=0

Thanks artem. 

The software is for equatorial mounts:

"Satellite Chaser is a software made for tracking and guiding on satellites using ASCOM compatible equatorial mounts." 

I pasted link for it earlier, so maybe James or other owners of equatorial mounts will be interested. I have motorised my Dobsonian mount, which is quite unique, so I have to find another way. 

[Vroobel]

28 minutes ago, JamesF said:

A quick hunt around this evening suggests that you just need to grab the latest TLE data for the ISS.  Then it's a case of finding it in the sky.

James

I agree with you, James. If I understand it properly we can get a kind of set of positions for next 300 seconds. It's not too much. I think the API gives data about now + 300 seconds, but I'd like to have data about 10 mins later + next 300 second. Without some attempts I can't say anything. 

As I read earlier in the linked sites, the software for equatorial mounts use the TLE data as well.

Tomasz 

[Vroobel]

At a second look it seems to be not so easy.

First of all you have to provide coordinates of your location. It can be your home or some other places, more then one, but you have to choose one as default. Every time you request for data from the server sending dedicated license key you do it for the selected location. I just checked: added my home coordinates taken from GPS in my telescope and also another location outside the city taken from Google Maps - the most recent location is always the selected one, but you can change the selection. Thus you have to be prepared and location has to be provided and selected if you want to track the ISS outside your home.

Every request for positions of the ISS is limited to 300 sets of data, every position is for one second, thus you have set of data for following 5 minutes. It not too much. You can ask for the data maximum 1000 times within 1 hour. My Az drive allow to revolve the telescop 360 degrees within 5 minutes (I know, it's not fast...), I shoud reach an opposite (faretst) position within 2.5 minutes. I should figure out how to aim the telescope in position, where the ISS will be after maximum 2.5 minutes, then the tracking shoud do its job within remaining 2.5 minute. It's long enough to make quite nice video. 

Not bad, but it's still not set of desired equations...
 
I'll try to play with that a little bit, but will still seek another way.  How the Stellarium is doing that? It use TLE data as well, but I launched it with no internet conection, so it couldn't get data from the serwer, but it is working despite no link with serwer. Interesting...

[JamesF]

If Stellarium is working without an internet connection, perhaps it may not be accurate?  Or perhaps Stellarium has some other special arrangements.  However, as it is open source, you could at least have a look at the code and see what is going on.

James

[Vroobel]

BTW, a video linked below shows how it should work. There is Starlink recorded, but the idea is exactly same. I expect a little bit longer delay between the slew and the tracking related with my software, but can't be sure. 

You can hear single "beeps" with decreasing intervals, so I think the mount reaim an OTA several times, then the proper tracking starts. Unfortunately palm-trees are disturbing. 

A most interesting part begins after 16.00 minutes. 

Tomasz 

[Vroobel]

On 27/08/2019 at 00:09, JamesF said:

If Stellarium is working without an internet connection, perhaps it may not be accurate?  Or perhaps Stellarium has some other special arrangements.  However, as it is open source, you could at least have a look at the code and see what is going on.

James

I think the Stellarium is accurate enough within 30 days since update of TLE. Wht is the TLE?

https://en.wikipedia.org/wiki/Orbital_elements#Two-line_elements

You can get it from the https://www.n2yo.com site after registering. Every time I get it I have same numbers, because possibly it's valid for some time or forever - I don't know yet. I saved a file with the data and will check it for every several days. I read on Wiki that it becomes inaccurate after 30 days.

I found similar need of the Ra and Dec coordinates here:

https://www.physicsforums.com/threads/satellite-right-acension-and-declination-calculations.220399/

then jumped to mentioned subject:

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

and found the description of the TLE. After then I found the pdf file: spacetrk.pdf - it's there: 

https://www.celestrak.com/NORAD/documentation/spacetrk.pdf

And here my headache began...

I know that the TLE includes coded orbit of the ISS, but is here any matematician / physicist / astrophyisicist able to pass through the equations and receive the Ra and Dec? 

But I found also this: 

https://github.com/brandon-rhodes/python-sgp4 - "Python implementation of most recent version of the SGP4 satellite tracking algorithm".

So will try to sort it out.

Tomasz

Edited August 27, 2019 by Vroobel

[Davey-T]

NASA publish TLEs daily for the ISS and more than once daily if it's docking.

Did this awhile ago using Meade Wedge mounted SCT,  I had to enter 8 lines of code as a user object, things like AOS (acquisition of signal) set minimum AOS, Epoch (year) (day) from TLE rounded to 4 decimal places to give accuracy of a couple of minutes, inclination, RA ascension node, eccentricity, argument of perigee,  mean anomaly, mean motion etc.

IIRC it was lacking a LOS (loss of signal) to stop it so had to abort it manually otherwise it kept tracking below horizon.

Dave

[Vroobel]

Thanks, Dave. 

Could you please share some details with us? I'm interested in and technically ready tor the accuracy of seconds. 

Tomasz 

Edited August 28, 2019 by Vroobel

[Davey-T]

Hi Tomasz, I can't pretend to understand the mathematics involved and just followed instructions on how to fill in various parameters using the TLE data.

Data is supplied to a few decimal points and rounded to the desired level of accuracy, IIRC I rounded to 4 points.

This was specifically for a Meade Alt / Az mounted scope, not sure if it applies directly to EQ mounts.

I have a 10Micron EQ mount that does a similar but much more accurate thing, just download updated ISS TLEs into the mount and it will track the ISS perfectly.

Some years ago there was a Windows Java script app that downloaded the TLEs and sorted them into correct format to enter into the mount, around Win XP times I think, not much good now I guess as Microsoft have abandoned Java support.

I'll see if I can find any links to the old Meade site but don't think it's active anymore.

Good luck

Dave

Edited August 28, 2019 by Davey-T

[Vroobel]

Something for Python lovers:

https://rhodesmill.org/skyfield/earth-satellites.html

I have to install this package and try.

[ngc3031]

Hey guys,

Developer of Satellite Chaser here. I’ve added support for Alt/Az mounts. I don’t have an Alt/Az mount, so this could be buggy, please give me feedback if you try my software.

In case someone does try my software with an Alt/Az mount: Make sure that in the mount tab N and S correlate to Altitude movement and W and E correlate to Azimuth movement. Also do a test run as detailed in section 9.1 of the manual. In the guide and cockpit tab Ra is synonymous with Azimuth and Dec is synonymous with Altitude.

Specifically to your homemade mount: It needs to support arbitrary MoveAxis commands to work with my software. I too have made my own mount, mine is equatorial, but it also works with a Raspberry pie and python. Nice to see someone else with a similar setup.

A bit about how I calculate the position of the satellite in my software: I use the python library pyephem, the script is only about 30 lines. It takes time, TLE and Longitude/Latitude as Input and gives the Altitude, Azimuth, RA and DEC for a pass as output. Satellite Chaser then uses that info, takes the first derivative of the position data (which results in speed) and then feeds these speeds to the mount 10 times per second.

Because it hasn’t been mentioned yet: The easiest way to get the newest TLEs is via celestrak: https://www.celestrak.com/NORAD/elements/stations.txt

Edited August 31, 2019 by ngc3031

[Vroobel]

ngc3031, thats very clever!

Quote

>>> import ephem
>>> mars = ephem.Mars()
>>> mars.compute('2008/1/1')
>>> print(mars.ra)
5:59:27.35
>>> print(mars.dec)
26:56:27.4

I just checked, the Mars show's little difference in comparison to the Stellarium, but the Jupiter is nearly exactly same! I'm impressed! I can use it instead of typing hundreds line of code based on the "Astronomical Algorithms".

After a check of all planets I think is more accurate for the gas giants then other planets, but I gonna play with it! Even with the tiny inaccurates I can use a wider eyepiece to find the planet and then use another one.

Let's try the ISS...

[Vroobel]

Oh, mate, that's awsome! 

I just checked the ISS. Accuracy of the library is equal around 1 arcsecond! 

Regarding your software, the Satelite Chaser, I found out about it from one of our Polish astro-amateur, but I couldn't use it because of not equatorial mount. Moreover, my mount is not compatibile with any existing standard (that can change...).

Actually I am happy to have fantastically working library giving me the required Ra and Dec. 

Now I have to think hard how to make the proper slew to expected position and wait for the ISS. There is no place for mistakes, I can make 1 attempt per sesion, as I read the ISS is well visible once per 63 days. Maybe it's visible through the telescope, byt not so good. It will take some time.

https://rhodesmill.org/pyephem/

Best regards for all engaged!

Edited September 2, 2019 by Vroobel