readme.html

ObsReduce is an MS Windows program that reduces observations of satellites relative to the backgroun

<P>
In case of a coincident appulse, i.e. the satellite appeared to pass directly over the star, enter zero as the miss-distance. The co-ordinates of Star A will be reported as the satellite's coordinates, and the position-angle will be ignored.
<P>
If you judged the miss-distance as the fraction of the distance between the appulsed star, A, and another star in the FOV, then the input box below the Pos Angle box will be labelled "Frac AS/AB", and you should enter your estimate there.
<P>
If you judged the miss-distance as the fraction of the distance between two other stars in the FOV, then the input box will be labelled as "Frac AS/BC"
<P>
ObsReduce computes the miss-distance internally, based upon the fractions you input, but has no room to display the result; however, you can peek at the value by placing your mouse cursor over the "Frac AS/AB" or "Frac AS/BC" label beside the input box.
<P>
The miss-distance has default units of degrees, but you may change this to arc minutes by entering min in ObsReduce.ini at line 13.
<P>
<B><a name="PosAccy">D.9.1.4 &nbsp;Pos Accy Text Box</B>
<P>
This is your estimate of the accuracy of your positional observation. You may set a default accuracy value in ObsReduce.ini at line 7. The default units are degrees, but you may change this to minutes by entering min in ObsReduce.ini at line 6.
<P>
Entering a positional accuracy of zero, instructs ObsReduce to insert its estimate the positional accuracy into an IOD or UK formatted report. It is computed as a fraction of the distance of between star A and B, or in the case of an appulse, as the fraction of the Miss Distance. The fraction used depends upon the geometry.
<P>
The defaults are 0.05 when the satellite passes on-line and between two stars, 0.07 when it passes on-line and outside two stars, 0.10 when it forms a right-angled triangle with two stars, and 0.15 in the case of an appulse. In the case of a coincident appulse, i.e. zero miss-distance, the default uncertainty is fixed at 0.005 deg. You may change the defaults, by editing lines 8 through 12 of ObsReduce.ini.
<P>
Even if you do not use ObsReduce's estimated positional accuracy within an IOD or UK report, the value will appear in the box labelled <a href="#Standard">"Std. Pos Accy +/-"</a>, located near the bottom of the window.
<P>
<B><a name="Camera">D.9.2 &nbsp;Camera Geometries</B>
<P>
The <a href="#Traditional">traditional geometries</a> are ideal for non-camera observing, but they occur too infrequently to be relied upon for use with cameras. ObsReduce supports two special geometries ideally suited for use with cameras: satellite located to the right or left of a pair of stars, or any configuration of a satellite and three stars.
<P>
The reduction algorithms of both geometries require the physical length on the image between one pair of stars, and between each star and the satellite. The program accepts lengths measured using a ruler or calipers, or the pixel coordinates of the satellite and stars, from which it will compute the required lengths.
<P>
<B>D.9.2.1 &nbsp;Measured Length</B>
<P>
To work with lengths, select option "Length" in the Geometry frame. The Position frame will display several input controls, depending on the number of stars selected in the FOV.
<P>
In the case of a satellite located to the right or left of two reference stars, the Position frame displays input boxes for the measured length on the image between stars A and B, star A and the satellite, and star B and the satellite. Use the option buttons that appear at the bottom of the frame to specify whether the satellite was "Right of AB", or "Left of AB".
<P>
In the case of a satellite and three reference stars, the Position frame displays input boxes for the measured length on the image between stars A and B, star A and the satellite, star B and the satellite, and star C and the satellite.
<P>
<B>D.9.2.2 &nbsp;Pixel Coordinates</B>
<P>
To work with pixels, select option "XY" in the Geometry frame. The Position frame will display several input controls, depending on the number of stars selected in the FOV.
<P>
In the case of a satellite located to the right or left of two reference stars, the Position frame displays pairs of input boxes for the x and y pixel coordinates on the image of the satellite and stars A and B.
<P>
In the case of a satellite and three reference stars, the Position frame displays pairs of input boxes for the x and y pixel coordinates on the image of the satellite and stars A, B and C.
<P>
<B>D.9.2.3 &nbsp;Accuracy Tips</B>
<P>
In principle, using three stars should be more accurate than using two stars, but both can yield accuracy significantly greater than 0.01 deg cross-track, when used with suitable reference stars.
<P>
Measured lengths are scaled according to the distance between stars A and B. For this reason, it is recommended to choose star A as the one closest to the satellite and star B as the furthest from the satellite.
<P>
For two reference stars, A and B, if the satellite is between them, and is close to star A then an exact in-line or right angle configuration is best. Exact in-line can be forced by setting the measured length BS to the smallest non-zero value, i.e. 0.0001, but this should only be done when the satellite is between A and B. If the satellite is outside of AB, setting BS to zero will cause unreliable results.
<P>
If the satellite is between A and B and not in-line, then an equilateral geometry would minimize error. Close to in-line compounds the measurement uncertainty.
<P>
Using three reference stars, it would be optimum for minimizing measurement errors to get the satellite within the triangle formed by A, B and C, and all the lengths, AS, BS, and CS close to the same size.
<P>
If the satellite is near star A and the other reference stars are positioned to help reduce measurement error, a three star position will be superior to a two star. For instance, this will work well:
<pre>
                                                 S
                                                 A

                                 B                              C
</pre>
<P>
WARNING! Avoid using three reference stars that are exactly in line. This will probably never occur in actual practice, but using it in a reduction would yield unreliable results.
<P>
Pixel coordinates may yield greater accuracy than measured lengths, but both will yield excellent results, when used with care.
<P>
In tests, repeated measurements of an on-screen length using calipers accurate to 0.001 inches, varied by 0.002 or 0.003 inches, indicative of the actual accuracy. This corresponds to a linear angular error of 5 seconds of a degree. Depending upon the geometry chosen to calculate a position from these measurements, error is estimated as anywhere from 60 seconds of a degree, worst case for an unfavorable geometry, to 20 seconds worst case for a more favorable geometry. Most situations will be more accurate than this, on the order of 10 seconds of a degree.
<P>
Due to the expected consistently high accuracy of camera observations, there is no provision to enter estimates of position uncertainty on a per observation basis. Instead, the program uses the same fixed value as for coincident appulses, i.e. zero miss-distance. The default uncertainty is fixed at 0.005 deg, which may be changed by editing line 12 of ObsReduce.ini.
<P>
<B>D.9.3 &nbsp;Important Notes</B>
<P>
IMPORTANT: ObsReduce interprets traditional geometry selections and position entries in accordance with the orientation of the FOV, as discussed in <a href="#FOVsettings">Section D.16. FOV Settings Frame</a>.
<P>
IMPORTANT: Moving the satellite may result in automatic clearing of positional entries, and  other data, as discussed in <a href="#Autoclear">D.15.1 &nbsp;Automatic Observation Clearing</a>.
<P>


<B><a name="Optical">D.10 &nbsp;Optical Data</B>
<P>
If you seldom report optical data, you may instruct ObsReduce to skip past all of the optical fields during data entry, by entering n at line 24 of ObsReduce.ini.
<P>
<B><a name="Magnitude">D.10.1 &nbsp;Magnitude Text Boxes</B>
<P>
Magnitude is entered via two text boxes, the meaning and labeling of which is determined by the reporting format.
<P>
In IOD format, the first box is labelled "Magnitude", and is used to enter the satellite's magnitude at the time of the observation; the second box is labelled "Unc +/-", and is used to enter the magnitude uncertainty. The default uncertainty is 1, which can be changed by editing line 23 of ObsReduce.ini.
<P>
In UK and RDE format, the boxes are labelled Brightest and Faintest, and are used to enter the satellite's magnitude extremes during the minute centred on the time of the observation. In the UK format, if both magnitudes were the same, only the brightest is reported, and the field for faintest is left blank. ObsReduce respects those rules.
<P>
<B><a name="Period">D.10.2 &nbsp;Period Text Box</B>
<P>
Use this box to enter the satellite's period of brightness variation, in seconds.
<P>
<B><a name="OpticalCode">D.10.3 &nbsp;Optical Code List Box</B>
<P>
The Optical Code specifies the object's optical behaviour. The choices depend upon the reporting format. You may set the default value at line 25 of ObsReduce.ini. If you usually do not report this or any other optical data, you may prefer to set the default to a blank space.
<P>

<B><a name="Reduce">D.11 &nbsp;Reduce Button</B>
<P>
Pressing the Reduce button results in the following actions:
<P>
- reduces the positional observation based upon the coordinates of the selected reference stars and specified geometry and position
<P>
- formats the reduced observation in the chosen reporting format, and displays the result
<P>
- if the object's elements have been loaded, computes the cross-track and time difference between prediction and observation. See below.
<P>

<B><a name="Error">D.12 &nbsp;Obs - Pred Frame</B>
<P>
Obs - Pred consists of the position cross-track and time difference between observation and prediction, with the track compensated for Earth's rotation in the interim. The cross-track difference is displayed in the same units as the position uncertainty, as specified in the ObsReduce.ini file at line 6.
<P>
These values are computed whenever the Reduce button is pressed and orbital elements have been loaded. Their sole purpose is to aid in finding and confirming the reference stars used in an observation. If you make several observations of the same object on a pass, it reasonable to expect them to have roughly similar cross-track and time differences, If not, then it is worth reviewing the input data to ensure that no data-entry or other processing error has been made.
<P>
One pitfall to be avoided is attempting to "improve" observations on the basis of the predicted track or the cross-track and time differences. Observations are made at the eyepiece, and cannot be improved afterward, by ObsReduce or any other means.
<P>
The only meaningful measurement of observation error is through the orbit determination process, in which a number of observations made over a period of time are analysed to produce a new set of orbital elements that best fit the observations.
<P>
The difference between predictions made with the new orbit and the observations, called residuals, are the best indication of the absolute and relative accuracy of the observations.
<P>
Differences between prediction and observation are to be welcomed - indeed they are the reason that we observe - to provide the data by which ever-changing orbital elements may be updated. Attempting to "correct out" the differences during the reduction process obliterates the very data that we seek to obtain.
<P>

<B><a name="UseAs">D.12.1 &nbsp;Use As Early/Late Button</B>
<P>
If the difference between the observed and predicted time is significant and you have additional observations to reduce, pressing this button will copy the time difference into the <a href="#Early/Late">Satellite Early / Late text box</a> (located to the left of the Plot Satellite button, so that all subsequent satellite track plots are compensated for Earth's rotation corresponding to early/late interval. In the vast majority of cases, this will place the reference stars near the centre of the FOV (field of view), thus speeding the task of identifying the correct stars.
<P>

<B><a name="Elset">D.13 &nbsp;Elset Frame</B>
<P>
The elset frame displays the age of the orbital elements used to compute the satellite's track, and an estimate of the uncertainty in the predicted time of the pass. This information is intended as a rough guide to the reliability of the prediction; it is probably most useful to observers who have a practical understanding of orbital elements, and how their accuracy degrades over time.
<P>
The elset age is the number of days from the epoch of the orbital elements to the time of the observation.
<P>
The prediction time uncertainty, shown in the box labelled Tunc, is the number of seconds that the object would be early or late, assuming that its rate of orbital decay is uncertain by 10 percent.
<P>
The prediction time uncertainty provides a reality-check of the time difference between observation and prediction, displayed in the box labelled Time, found in <a href="#Error">Obs - Pred</a> frame.
<P>
For example, if the prediction time uncertainty is 1.5 s, then I should not be surprised to find that the time difference between my observation and the prediction is of the same order of magnitude.
<P>
Of course, the decay-rate uncertainty itself is uncertain, so it may differ greatly from the 10 percent assumed by ObsReduce, which in turn affects the time uncertainty estimate.
<p>
Fortunately, the prediction time uncertainty varies in direct proportion to the decay uncertainty, so if there is reason to believe that the decay uncertainty differs from 10 percent, it is easy to mentally adjust the time uncertainty displayed by ObsReduce.
<P>
For example, if the displayed time uncertainty is 0.5 s, but you suspect that the decay uncertainty is 50 percent, not the 10 percent assumed by ObsReduce, then mentally multiply 0.5 s by 50/10, which yields a time uncertainty of 2.5 s.
<P>
If the elements are fairly recent, and known to be reliable, then ObsReduce's assumption of a 10 percent decay-rate uncertainty probably is reasonable. Otherwise, the uncertainty is likely to be much higher, but that is a matter of judgemental.
<P>

<B><a name="Centre">D.14 &nbsp;FOV Centre Frame</B>
<P>
The FOV Centre frame contains the controls to centre the field of view for a given date and time, anywhere in the sky, in coordinates of R.A. and Dec or AZimuth and ELevation.
<P>