This chapter is a tutorial explaining how to use Smeg as a tool for visual satellite observation. If you're into radio predictions, please read this chapter for its discussion of Smeg basics, then read the next for detail about radio predictions. If you only want to know about visual predictions, you can read this chapter and skip the next. If you'd like to use Smeg for both purposes (or you just thirst for knowledge), you should read both chapters.
These tutorials assume you've started smeg already (you can do so by
running the smeg command. They also assume that you'll be
giving all the commands as part of the same session. (A safe assumption,
as Smeg currently can't save state between sessions.)
Commands you're to give will be prefixed by the smeg> prompt, and
output will be shown unprefixed.
The first step in predicting satellite passes is to define the observer's position. This is done using three commands: set lat latitude, set lon longitude, and set alt altitude. The latitude and longitude are given as signed decimal degrees. For latitude, positive is north, and for longitude, positive is east. (Please note that this means observers in the Americas will have to enter a negative number for longitude. Altitude is given as a decimal number representing kilometers above mean sea level. To check the current settings, one can run the command show location, which will print out the current location.
The following example will change the location for predictions to that of a beach near Tom's house. (It then displays the location for confirmation purposes.)
smeg> lat 40.9266
smeg> set lon -73.3123
smeg> set alt 0
smeg> show location
The current location is:
Latitude: 40.9266 N
Longitude: 73.3123 W
Altitude: 0.0000
A satellite's orbit can be characterized by a set of orbital elements. These elements are usually produced by NORAD, but occasionally are created by amateur observers (as is occasionally the case for classified satellites.). Before Smeg can be useful, a set of these elements must be loaded. This is accomplished by the tle filename command. (Tle stands for two-line elements, the format NORAD uses for element files.)
For prediction of visual passes, using two line elements with embedded
magnitude information is highly recommended. (This is currently the
only way one can load magnitude information into Smeg.) The tle file
of choice is the McCants file. It's available from http://www.fc.net:80/~mikem/tle.html as mccants.tle. Included in the Smeg distribution is a
perl script, gettle that will fetch the file automatically.
To use it, simply run (from the shell) the command
gettle mccants.tle. Then, in Smeg, to use the tle file:
smeg> tle mccants.tle
Loaded 1462 elements.
The McCants file is only updated about once a week. You may need to download other files in order to have accurate predictions of manuvering satellites and spacecraft. (Such as the space shuttle, soyuz capsules, and progress freighters.)
This loads the elements from a file, and builds a satlist structure in memory. Depending on your circumstances, this may be enough. However, in some cases, it may be desirable to filter out entries from the satlist. This is done using the command filter satlist expression. The expression used varies depending on exactly what filtering is desired on.
One useful filter command is filter satlist minmag < 4, which will remove from the list satellites that have a theoretical maximum brightness in perfect conditions of greater than 4th magnitude. (Remember, a lower magnitude means a brighter star or satellite.) Another useful command is filter satlist norad = 16609, which selects Mir, which has the NORAD catalog number of 16609. Of course, the number can be changed to that of any other satellite. Conditions can be connected by the boolean operators and and or.
Filtering can only remove satellites from the list, not add them. You can add all filtered-out satellites back to the list by running the command unfilter satlist. You can also see the current contents of the satlist by running the command show satlist.
The next step is to predict when the satellites in the satlist will be in your line of sight. This is modified by a minimum elevation, which is set using the command set minelev elevation. I find that set minelev 10 works well for visual observations from my site, your mileage may vary.
Once the minimum elevation is set, you can run the predict passes command. This will generate a list of passes for the next 24 hours. (This can be modified using timespecs, as discussed later in this manual.) Smeg will print out a progress report, and will eventually tell you the number of passes that it found.
smeg> set minelev 10
smeg> predict passes
Prediction complete. 6427 passes found.
Before we get down to the actual commands needed to filter passes, it's good to find out in exactly what conditions a satellite is visible. The basic rule is that a satellite is potentially visible if the observer is in darkness while the satellite is in sunlight. As a practical matter, for darkness to reign the sun's elevation has to be less than -6 degrees beneath the horizon.
Two expressors have been added to smeg to represent these two conditions. The first, sunel is a float representing the elevation of the sun, in degrees. The second, lit, is a boolean that is true when the satellite is lit at either the beginning or the end of the pass.
Using this pair, and the filter passes command, we can easily select only potentially visible passes. The passes that are not filtered out after the following command will always be potentially visible (although perhaps they may be very dim.).
smeg> filter passess sunel < -6 and lit
2143 passes passed filter.
After passing the filter, all the satellites seen are potentially visible. Most, however, are very dim. To determine which can be easily seen visually, we can step through the list of passes, calculating the magnitude at various points along each one, printing out a line when a condition is met. This is done using the command step passes expression, as in:
smeg> step passes mag < 4
25407 Cosmos 2360 r (11.055 days old)
19990717 22:30:21 62.1 NE 44.5 +3.9
19990717 22:30:51 77.0 E 43.2 +3.8
19990717 22:31:21 90.1 E 40.1 +3.8
19990717 22:31:51 100.6 E 36.1 +3.9
19990717 22:32:21 108.7 E 31.8 +4.0
19120 Cosmos 1943 r (11.173 days old)
19990717 22:29:21 11.2 N 59.9 +3.8
19990717 22:29:51 35.1 NE 68.6 +3.4
19990717 22:30:21 77.2 E 70.9 +3.2
19990717 22:30:51 110.4 E 64.3 +3.1
[etc...]
Note that this usually produces copious amounts of output. You can redirect
this output by prefixing the command with less or print.
less pipes the output to the pager (usually less,
appropriately enough.), while pager sends the output to
a printer, or at least the lpr program.
For each pass, one header line is printed, and one or more data lines are printed. The header line consists of, from left to right: the satellite's norad number, the satellite's common name, and the age of the elements used to make this prediction. The data lines contain the following: The date (in YYYYMMDD format), the local time, the satellite's azimuth, a compass heading roughly indicating that azimuth, the satellite's elevation, and the satellite's visual magnitude, if available.
Probably the most useful thing to use in these expressions is the mag expressor, which allows you to compare against the satellite's current computed magnitude. (Remember, the lower the magnitude, the brighter the satellite is.) If the element files omit magnitude, this will lead to an empty least. In that case, leave the expression blank. (A blank expression is always true.) You'll be stuck having to use other means to find the magnitude of the object, but at least you'll get output that's correct in other respects.
The previous step should be enough for someone with an accurately set watch to observe satellites in the field. However, for someone with a laptop, Smeg has even more benefits. These are the live displays.
The first is useful if your laptop is fast enough to do the predictions itself. In that case, one simply runs the live passes command, which will display predicted passes. Please note that this is only filtered by the filter passes command, but not the results of step passes. Pass selection is accomplished using the up and down arrow keys, 'q' exits the mode.
The second, and more complicated, procedure is useful if the laptop is not fast enough to run predictions. (This is especially the case if it doesn't have a FPU.) First, complete steps 1-5 on a fast machine. Then, complete steps 1 and 2 on the slow laptop. Enter the command noradfilter. Enter the norad numbers produced by the step passes, one per line. End with a blank line. Then, finally, run live satlist. While this discards pass times, it does provide a way to run the live display on a slower laptop.
The live displays are based on system time. It's highly reccomended that you use a protocol such as NTP to synchronize your system's clock with a reference time source.
It's been found that these live displays work best when used in conjunction with two person teams. One watches the laptop (destroying night vision), and calls out the bearing to the object. The other scans the sky. When he finds it, he'll then point it out to the first observer, who won't need to look at the laptop to find it.
Of course, the best laid plans can be destroyed by clouds or light pollution. That's just life. Still, we wish you luck in your observing efforts!
tom-smeg@onegeek.org