This site is to document how to track satellites - computerized mounts and telescopes help. It might be called "how to track satellites more easily, using common cameras or even computerized tracking systems" but that name is too long. At first this site is going to describe how to track with a commonly available camera and it will also cover (in a linked page) the use of a Meade drive and their Autostar software (see below for why that was chosen). Probably people with these systems know a lot about them, including how to observe astronomical objects, but the techniques and sources of tracking satellites may be new. The Meade user's manual has pretty pathetic instructions for people, and it could have easily included far better web page URLs. Hopefully soon I can include how to add a telescope on the Meade drive, this allows a site to track much fainter satellites.

There are a number of sites that talk about tracking and even show the history or results, but I have not yet found one that tells someone how to set up a site and get started. I got hints about Astrometry.net, I got hints about FITS files, I got hints about SAO Image DS 9 - but putting that all together has been very frustrating. So I created this site, with inexpensive cameras and available software anyone can track satellites and generate observations.

By the way the REAL trackers use video cameras, they allow you to produce many observations per night. But it requires more money and more expertise, I might get around to documenting some of that.

Now there are a few older sites that show the early results using custom software, hand built drives, etc - we owe a LOT to the pioneers.

Caveat

You can take observations with common hardware - a digital camera, a smart phone, a laptop, and a tripod. This will only let you track the brighter satellites, later I will describe how to find them. I have a used Nikon D200 with a 50mm fixed focal length 1.8 lens (after starting with a 1:3.5-45 35-105 zoom lens) and a (not needed after all) GP-1 (GPS) unit. My total cost so far is about $385 - I should have started with the faster lens but thought that I might need to use a longer focal length like 70mm. But when you zoom out like that the lens is very "slow" and the images are very dim. I would like to upgrade to a newer camera body at some point so I can record video; the D200 cannot take video. I will almost certainly keep the D200 as well since it works great. To point my camera I use an application on my iPhone, Orbitrack. It shows me the path that the satellite will take and helps me find a star to point at, one that the satellite will go past.

I started with a couple of simpler cameras, then I thought that I needed to buy a camera with a GPS module - to take photos with sufficient time accuracy. You need to know, within a tenth of a second, the time the shutter opened. But the GPS unit does not help with that, the GPS has turned out to not be necessary but it is darn handy if you are moving and want to record your location (when you are tracking satellites you do not move!).

If you are doing this and want to know how to get accurate time - send me an email. There are some things you need to know about shutter opening and closing times.

To record the precise time you need an application that reads the Exif information for the photo - I use EXIFTool. See "Generating Observations" below for how to take a photo and get observations from it. To accurately record my location I use an application on the iPhone.

I hope to update these pages over time to tell people how to calculate an orbit for a satellite and to later update that orbit.

As I said, you do not need a telescope but computer controlled mounts are widely available so let's give some information about a common example in case you can get one. A telescope will allow you to track much dimmer objects, satellites that are deeper in space. I will later put in some thoughts about what a telescope vs camera does to your field of view. The linked page has some details about using a computer controlled pointing system.

My efforts are all using Macintosh computers but many people use Unix or the other major operating systems, see the Hardware/Software section below for more explanation of the satellite prediction software.

Motivation

Governments that launch satellites have a responsibility to ensure that people on the ground below are notified if objects that they launch are about to impact the ground. The US government took it on itself to be the world's authority on what is in space and along with that came the ability (and perhaps the responsibility) to alert the world when objects are known to be reentering. One of my tasks, early in my career, was to track the Skylab space station and help predict its reentry. I want to track satellites to gather more positional information which will help predict impact times (and general locations) of some objects that will NOT appear in the available warnings of reentering satellites. Our government pretends that a whole class of satellites does not exist and so they do not warn people when they are about to reenter. The following article is one that I wrote to point out that the US government needs to do a better job of warning people about satellites that are likely to survive to hit the ground.

https://www.thespacereview.com/article/3397/1

Overlap

This site does overlap somewhat with a couple of other sites and projects I have, and I will (when appropriate) link to those sites. But the intention here is to help people get started tracking satellites.

Hardware/Software

A simple system could consist of a tripod, a digital camera, a smart phone, and a laptop. I have gotten some very acceptable results (I have had to wait weeks until the sky was clear enough to see any stars). You need software to tell you where and when to look, see the link to another of my pages just below; I use Macintosh computers to track objects and have been very happy with the ease of use and availability of software.

With the complicated setup that we are using, you first need to get some satellite prediction software running, the software I use is on my Macbook Pro, it's Gpredict. The site linked next tells you how to get Gpredict running.

How To Use Macintosh Computers To Track Satellites

A more capable system could include a computerized drive system like the Meade LXD 55 mount - and Autostar software. So the initial page will concentrate on those but hopefully will grow to give directions for other systems. If you have a different application - look around for similar options and similar menus, many writers have based their software on just a few original applications. So the applications that I have seen all have very similar dialog boxes and labels. If you have Autostar you should be able to read the user's manual and get it running if you don't already know how.

For the computerized tracking system - I put the satellite data we need into Autostar but we have not yet loaded and worked through any of the other available programs that come from other companies besides Meade. I have not had a chance to use the Meade drive to show which satellites are visible, so for the time being I use my Macintosh computer and two programs (Gpredict and JSatTrak) to get ready to track. Also we should be able to drive the mount from Autostar, probably with a telescope on it that has the camera connected to it. Hopefully we will figure out how to use Autostar to tell us when satellites are going to be visible.

When I have the photos I use a web Astrometry program and an application like SAO Image DS 9 (see below), I will continue to update these directions.

Now Let's Take Photos Of Some Satellites

Go to this page Examples to see how to use "Sky At A Glance" to see when satellites will be available, this allows you to plan your observing time. I also use the Orbitrack app that shows me the sky above with the stars and the satellite - it is good for "instantaneous" tracking but it is slow to plan observing for the next couple of hours. It also does NOT show some of the satellites that are very interesting but I am not ready to look for them yet anyway. With Orbitrack you can see the path that the satellite will take, and I pick out bright stars near that path that I use to point my camera.

Camera settings: I use an application (cleverly named Time) on my iPhone to get the most accurate time and then set the camera clock manually. Set your camera to take a 10 or 15 second exposure, after a 2 second delay. So when you push the button to take a photo there is a 2 second delay to allow the camera vibrations to damp out. ISO rating should be set at 400, this is pretty high. But you will have dim satellites and stars in the image and they will not show up unless you use a high ISO. When you get the image, there is software (to read the EXIF data) that allows you to get the time of the exposure. Which is plenty accurate! By the way do not synch the time between your computer and the camera, my experience with computer time can be off by a few seconds.

Select an upcoming object that has a good elevation pass - like 20 degrees or more - and will be visible for several minutes. In the evening, passes to the West will be back lit, those have not been useful. But if the satellite is East, North, or South of you that can easily provide good illumination. Higher elevation passes are better but I have gotten some good photos at low elevations to the North and South.

If you are using a camera on a tripod, point the camera towards a point that you know the satellite will pass through. I pick a bright star that the object will go close to (using Orbitrack of course) and point the camera towards the star so I can see it in the view finder, then I sometimes have to nudge the camera towards the satellite track. Orbitrack gives you real time elevation and the iPhone has an app that will let you check the camera elevation against the value you need but I have not used that yet.

I am hoping to track more distant satellites, but have not yet.

When the object should be coming into the field of view, push the button to take a photo. The idea is to NOT follow the satellite, we want a photo where the stars are points but the satellite is a streak. So the camera will NOT follow the satellite (in the below 20 second exposure I don't see any smear due to "sidereal" movement in that short time). I think that there is some other reason that the stars show short streaks.

Here is a sample that show some initial results - you can see stars in the photos as well as the streak from satellite number 31114, in the right bottom of the photo (it is dim). I used "Photos" to brighten the photo to help me see the streak. This was shot at ISO 800 and that produces an acceptable photo.

You can see that we have a lot of light pollution in my area, to get any usable photos I have to wait for a cold front to come through and then take advantage of a few clear nights after it.

Below is USA 281, Future Imagery Architecture Radar, NROL-47, going by; satellite number 43145.

Generating Observations

Once I get usable photos I need the exact time of each end of that streak, the time is stored in the photo in "EXIF data" and I use an EXIF data reader which makes this easy to get. The Nikon D200 gives the time of the shutter closing (instead of shutter opening as later Nikons do) and then the EXIF data also has the exposure duration.

Then I upload them into nova.astrometry.net and this "resolves" the image; it identifies stars and gives angles to them. It is an image with the angles to recognizable stars imbedded in it. A pal wrote a Bash script for me which loads a folder of images and gives me the needed results quickly.

Next I go to SAO Image DS 9 (which is an application on my Mac), that gives me a window with the image in it. I had to get that application, expand it, and install it on my Mac. With the output from nova.astrometry.net I just double click on the image results file and SAO Image DS 9 opens.

The small window on the upper right is an expanded view, put the cursor on the ends of the trail in that window and note the FK5 numbers - these are Right Ascension and Declination of that point. You can get RA and Dec for the beginning end of a streak like this and figure the time of each point.

Next you format the observation to mail it - I am using the IOD format - and type the observations into an email. Send that email to the SeeSat list, Mike McCants, and Ted Molczan. After MUCH experimentation I found the right combinations to get usable observations.

So after a lot of looking around and getting software and installing it and making lots of mistakes, this process is getting more automated. Yes I am going to add some illustrations here! Sorry that I am updating this when I have available time.

This site is NOT complete but it might help people get started - you might be able to fill in the gaps that I have left. I continuously update this site as I improve my techniques.

If you have a Meade computer controlled drive system and the Autostar software, go here.

Of course this site is copyrighted by Charles Phillips, you may not use any part of it without permission.