by Marco Langbroek





* This guide was written after questions on how I work with NEAT data. It is the way
I work (other ways are possible) and can be a guide to novice users of Skymorph.
Much of the technique described in this guide, I learned from Rob Matson, whom I
heartfuly thank for his teaching!

* The guide focusses on the discovery of new objects in the NEAT archives, but also
serves as a guide to doing precovery work on known objects.

* It is assumed that the astrometric program used is Herbert Raab's very fine shareware
"Astrometrica", see http://www.astrometrica.at

* Also needed is Bill Gray's freeware "FindOrb", http://www.projectpluto.com

* If you want to do this kind of work, you should get yourself acquainted with the meaning
of orbital elements.

Note: Astrometrica has very good step-by-step tutorials in the "help" section. Do these
tutorials to learn to work with the software.


Okay, here it goes step by step:

Step Action(s)

0. Go to Skymorph, http://skys.gsfc.nasa.gov/skymorph/skymorph.html and select "look
for images by time and position". Fill in a position (both RA and declination should
be input in degrees) and time search definition. Tip: mid-August 2002 has a number of
good quality nights close together.

*For precovery work, omit this first "step zero", go to step 8 immediately and omit
step 5. At step 8b, fill in the object designation at the top of the page or its
orbital elements below if it is a very recent discovery not yet in the HORIZONS

1. Download triplet of 3 FITS files with about 20-30 minute time spacing from Skymorph;
FITS from station 644 have an a, b or c behind the image number, the other are images
from station 608. The 608 images have so many false artifacts (and worse limiting
magnitude) that I never use them for searching, only for additional positions after
a find. I usually download images of 1100 x 1100 pixels (you can set this before
downloading under "request parameters: NEAT pixels". Also un-tag "DSS-image" and
"NEAT/DSS SkyView overlay" as you won't need these)

2. Open them in Astrometrica (and type in image dates and time in the entry boxes
that appear); Make sure your configuration file used is the 644.cfg (you can
alter this through "settings"), an example of which can be found at

3. Make a blink of these 3 images (tools ---> blink image, or click the appropriate

4. Look by eye for moving objects (Rob and I never use the auto-detect function);
when you have one, check whether the movement is consistent with the image sequence;

5. If you have a moving object, check whether its a known asteroid or not using ASTPLOT:

*note*: the reason for using ASTPLOT and not the "known image overlay" option of
Astrometrca itself, is that the MPCORB file Astrometrica uses will be for the current
epoch and not the epoch of these "old" images, which will introduce discrepancies.

5a. Select the correct station code from the list in Astplot, fill in date, time and
image center of the middle image of the triplet, change limiting magnitude to
+21.5 and image vector to "1 hour" and then click "obtain plot".

5b. It takes a while, but after some 2 minutes or so it will generate a map.
Compare this map to your images.

6. If the object on your images is an unknown: measure 3 positions (3 images);

6a: click "astrometry ---> data reduction";
6b: enter coordinates of image center (leave "object" field blank);
6c: Astrometrica will now download the appropriate set of
reference stars from USNO-B1.0 catalogue over the internet;
6d: after downloading it will proceed to auto-match this set
to the stars in the images *This can take several minutes!*;
6e: After the fit to the stars has been achieved, click on the
object you want to measure in the (first of 3) image;
6f: Astrometrica will auto-match to the center of the object. If
you are satisfied with this fit (99 out of 100 times), enter
a preliminary designation of 6 characters (e.g. "YOU111") in
the most lower-right box (the right-most box under "object
designation": leave the left box blank) and then click "accept";
if you are not satisfied with the fit click "reject", zoom in on
the image and mmanually put your cursor where you want it, then click
while holding the "ctrl"-key;

*Note*: if the object you measure is not a new discovery but a precovery,
enter it's "packed" MPC designation in the lower right box;

6g: Repeat steps 6e and 6f for the other two images.
6h: Astrometrica will auto-write each measurement to the file MPCReport.txt
in the correct MPC report format;
6i: *Note*: when starting up Astrometrica, it may say "MPCReport already
exists" and asks whether you want to overwrite it. DON'T DO THIS! You
will loose all previous measurements in the file... Periodically
archive MPCReport.txt by renaming it and then let Astrometrica create
a new, blank one. Note that new measurements on new objects are appended to
the existing MPCReport file, it will not overwrite older data in the file.

Your first set of measurements is now ready! This was the easy part. Now comes the
hard part.....

You now have to find the object back on images of at least two other nights, one of
which must be close to the first night (e.g. no more than a week later, preferrably
less...). For this, you have to "predict" where to look...

7. Switch to FindOrb software (www.projectpluto.com), which is an orbital propagator.
FindOrb will try to calculate a very preliminary orbit from a set of measurements.
It reads files in MPC format so it can directly read the file MPCReport.txt made by

7a: start-up FindOrb and open the file "MPCReport.txt" (which you can
find in C:/Program Files/Astrometrica);
7b: you will see a list of all objects in the file. Click on the name of the
object you want to process;
7c: you will now see a preliminary set of orbital element;
7d: fill in a value for the perihelion distance which is sensible for a
main belter in box "R1";
7e: click "Vaisala";
7f: You will now get a set of estimated orbital elements, and an
indication of the fit of the observations to this estimate.
*note 1* a Vaisala orbit is an orbit based on the assumption that
the object is in perihelion.
*note 2* you need to have some knowledge of orbital elements to
judge whether the FindOrb results make sense (they don't always do).

8. Now go back to the Skymorph website and:

8a: Choose "moving targets".
8b: scroll down and enter the orbital elements obtained by FindOrb into
the data input boxes, then click "send"

9. Skymorph will now return a list of images which *might* contain the object,
based on the preliminary orbit you submitted. This list should include the original
triplet of images which you measured, otherwise there is a mistake in the data input.

10. Go back to step 1, and this time use a second night as close to your first one as
possible: repeat the whole sequence up to step 4;

11. Now you will either find an object of correct magnitude *and moving in the correct
direction* in the set of images, or you won't. This is the hard part. Sometimes you
simply do not find the object. Bad luck then. Note that the object need not be in the
image center, it only will be if the preliminary orbit obtained by FindOrb is a quite
good estimate.

12. If you are lucky and do recover the object: repeat the whole sequence up to step 7c;

13. After step 7c, this time check all planets for nclusion of their disturbing factors,and then click "Autosolve" instead of "Vaisala". You will get a new set of improved orbital elements.

Take a carefull look now at your residuals. These should be below 1.0 arcseconds
for each data line. When they are (much) bigger, this is an indication that something
is wrong: for example that the object of the second night is *not the same* as that of
the first night (this is very well possible!)! Or maybe something went wrong when measuring an image. In this case, do *not* proceed to step 14!

14. You will now have a new set of improved orbital elements. Go back to step 8 and repeat. Try to do this again and again untill you have data from as many nights as possible. You'll also notice that the orbital solution will get better and better, and after 3-4 nights of data you'll start to recover your object in the center of the images.

At some point now, you'll not be able to get more nights. If you have at least 3 nights
including two that are close together (within a week), you can submit your data to the MPC now (in the case of recovery of a known object instead of discovery, you can report single night data).

15. *BEFORE* you start to submit data to the MPC, carefully read their instructions at
http://cfa-www.harvard.edu/iau/info/ObsFormat.html ! I also advise to read

16. Astrometrica already formats the data in the correct MPC format. Copy and paste the
data you want to sent from MPCReport.txt (which is in C:/program files/Astrometrica)
into a blank e-mail making sure you maintain the correct format (i.e., don't change
spaces, indentation etc.) and *include the correct header details*. Send the e-mail
to mpc@cfa.harvard.edu

*Note* sent the e-mail as plain txt, *not* HTML, and remove any signatures from the mail.

A typical, correctly formatted e-mail with data and correct header will look like
this example:

COD 644
CON *your name and address*你的姓名和地址
CON *your e-mail in square brackets [] *用方括号括起来的你的Email地址
OBS R. Bambery, E. Helin, S. Pravdo, M. Hicks, K. Lawrence, R. Thicksten
MEA *your name*你的名字
TEL 1.2-m Schmidt + CCD
ACK *fill in a code or whatever that identifies this set for you*你的代号或你认可的身份标识
LAMA90 C2002 08 08.44720 22 18 41.93 -13 37 36.8 20.7 R 644
LAMA90 C2002 08 08.45810 22 18 41.35 -13 37 37.6 20.8 R 644
LAMA90 C2002 08 08.46909 22 18 40.75 -13 37 40.0 20.6 R 644
LAMA90 C2002 08 18.42448 22 09 15.73 -14 02 21.6 20.7 R 644
LAMA90 C2002 08 18.43537 22 09 15.09 -14 02 22.8 19.4 R 644
LAMA90 C2002 08 18.44572 22 09 14.47 -14 02 24.7 20.1 R 644
LAMA90 C2002 08 26.17449 22 01 26.68 -14 20 27.2 20.7 R 644
LAMA90 C2002 08 26.24252 22 01 22.42 -14 20 36.8 19.9 R 644
LAMA90 C2002 08 26.31396 22 01 17.94 -14 20 45.6 20.2 R 644
LAMA90* C2002 08 29.36705 21 58 14.54 -14 26 57.7 19.8 R 644
LAMA90 C2002 08 29.38826 21 58 13.20 -14 27 00.2 20.1 R 644
LAMA90 C2002 08 29.41086 21 58 11.81 -14 27 02.7 20.6 R 644

In the above example (which is data I submitted for a new object that based on this
dataset got the official MPC designation K02Q66R), "LAMA90" is a preliminary personal
designation for the object. Use your own 6-digit personal designation here.

*Do not call your designations "LAMAxx" but design your own ID code!*.
"LAMAxx" is my personal system of temporary designations for my objects.

Under "ACK", fill in for example your personal object(s) designation(s) in this
dataset (see point 17).

17. After receipt of your e-mail, MPC will within a few minutes automatically send back
an acknowledgement, which will have as subject line the line you specified under "ACK".

18. After some time (a few hours or up to a few days, this depends) someone at the MPC
will sent you a mail called "designations". You will see a line with your temporary
designation, followed by the official designation assigned by the MPC. This of course,
assuming the MPC accepts your report. To understand the meaning of the coded content
of this message, read http://cfa-www.harvard.edu/iau/info/Astrometry.html#des

19. The next day after MPC mailed you the designation, you'll probably find it listed in
the Daily Orbit Update (DOU). You can find these at

20. You can check whether the object has additional single night observations by other
stations linked to it at http://cfa-www.harvard.edu/iau/MPEph/MPEph.html ;
fill in the packed designation, check "MPC 8 line" format and check the "show
residuals" block and then click "get ephemerides"

* NOTE * After recent MPC policy changes, you will no longer get official credit for new
objects for which a previous single night report exists. Credit instead will go formally
to NEAT or the other station that sent in the single night detection.
This is a bit of a disappointent after your hard work in prying out all these
undiscovered detections from the archive but then, that's the way it is. NEAT archive
hunters now count their finds unofficially as 'discoveries' just as SOHO comet hunters do.


Okay, hope this helps!!!! Happy hunting!


- Marco Langbroek
Leiden, 09/08/2005

Note 1: My sincere thanks to Rob Matson, who introduced me to the wonderful world
of SKYMORPH precovery and discovery. Almost all of the things this guide
contains, I learned from him.

说明1、我真诚地感谢Rob Matson,是他给我介绍了如何利用SKYMORPH进行发现和前期发现,本指南中几乎所有的内容我都是跟他学习的。

Note 2: if you mirror this guide, please acknowledge the source


Dr Marco Langbroek
Leiden, The Netherlands

Volunteer image reviewer FMO Spacewatch Project
NEAT archive hunter
Admin FMO Mailing List

e-mail: meteorites@dmsweb.org
private website http://home.wanadoo.nl/marco.langbroek/asteroid.html
FMO Mailing List website: http://home.wanadoo.nl/marco.langbroek/fmo.html

--- end of file ---