ULDA Notes

1  Summary

2  Introduction

The Uniform Low Dispersion Archive (ULDA) is a software system which. in a single sitting, allows you to obtain copies on your computer of those IUE (International Ultraviolet Explorer) low dispersion spectra of interest to you.

The search programs and data base are maintained by the European Space Agency (ESA) at the ESA IUE Observatory Villafranca. Spain (VILSPA) and updates are sent from them to the national host sites. Regular updates to the user ID files are done by VILSPA. The NSSDC is currently running version ?.0 which contains approximately ? images.

The ULDA software is menu-driven from within the National Space Science Data Center (NSSDC) captive account (NODIS). The ULDA user does not need to have the IUE/ULDA Coordinator intercede in order to access the data or search routines. It is a reasonably self-contained series of programs and a data base.

NSSDC is the United States national host site for users within the U.S. and a temporary host to astronomers in other countries that do not have an established national site.

The overall concept is that research institutes within a country are served by a single national host which can be accessed using intercomputer links. This center holds the ULDA as a relational data base as well as the software to interrogate and select spectra from the data base. In order to ensure that everyone has access to the same data ,these national hosts are in turn served by a principal center.

Below are the two ULDA programs you will use:

The search program runs at your national center while UNSPL runs at your local computing center. The following section comprises an overview of how to use the ULDA and its components.

3  How to Use the ULDA (Overview)

Retrieving spectra in a form suitable for your image processing system comprises a two or three stage process, the number of steps required depending on the nature of your connection to your national center. The steps required are described below.

4   Stage 1: Search and Select Spectra

When you log in to your national center's ULDA account from your institute , you will enter the Query/Select program through the NSSDC NODIS account. A user can initiate a new search for data or recover a previous file that may not have successfully been transferred to the user's node

If a search is chosen. the program writes the data to several user files and returns the user to the SELECT menu. Here users may edit the files. view the files and/or save the files they wish to transfer to their host nodes.

The search menus start with a specific search target and regress to a wider ,more general search window with each new menu display. A "?" in any of the search displays could yield further examples and/or clarification of what is expected within that panel. This is particularly useful when you have reached the Object Class and Homogeneous Object ID menus.

5  Stage 2: Transfer Data to Your Computer

This stage is required only if you are not connected to a national center by DECnet. To begin this stage, copy the spectra you have selected (held on a single file) to your computer using either a filer transfer system (e.g.. KERMIT) or computer mail. The Query/Select program affords you the option of outputting the spectra in compressed ASCII encoded form which is essential if you intend to use Email or a file transfer system which does not support the transfer of K blocks of binary data. If you have saved auxiliary data (e.g. what a search found) these will be in ASCII format and should be downlinked as well using Email or a file transfer system.

6   Stage 3: Generate Spectral Files on Your Computer

To generate individual spectral files having a format suitable for your image processing system, run program UNSPL on your computer. UNSPL (UNscrambler/SPLitter) will downlink your data for you from the national host if you are connected to it by a DECnet link (these data comprise your spectra and auJdliary information). The following two diagrams illustrate what this stage involves if you are connected to the national host by a DECnet-like link or not.

No DECnet link to National Center

With a DECnet link to National Center

7   Overview of How to Run the Query/Select Program



If you ignore all but the absolutely essential actions required, you are left with a three stage process.


  1. LOGIN via the NSSDC IUE NODIS account. Choose option 8 on the main menu and then option 2 in the IUE main menu display.
  2. Give the ULDA your ID (If you do not have one ask ULDA to give you one of your choice). See section 4.1.2. (Getting an ULDA ID IRegistrationl) and section 4.1 (Functions Avallable During LOGIN) for more Information. This ID is used to form the names of your work files. If your ID is UKMIKE then the spectra which you eventually choose will be stored on a file called UKMIKE.SPE.
  3. Give your motivation (a number corresponding to an entry in a menu) and you will be presented with the main menu.


You may search on one or more of the following:
  1. A list of cameras and/or image numbers and/or apertures
  2. A list of RA and Declination ranges
  3. A list of IUE object classes
  4. A list of homogeneous object IDs
You will always be prompted for the above four classes of input (carriage return <CR> for those you do not wish to use). If you reply to more than one of the above types of search criteria, the search will be for the intersection of the hits found for each set.

EXAMPLE: Find spectra taken with the SWP camera for objects with class 57 or 58.

Reply SWP to the camera/lmage number/aperture prompt and Reply 57,58 to the object class prompt.


Enter the SELECT command from the main menu and a description of the spectra found by the last search will be displayed a screen at a time. These descriptors (ie.. camera image number, aperture, homogeneous ID, coordinates, date of observation) take one line each and are prefixed by an Index (ranging from one to the number of spectra found) used to perform the selection.

EXAMPLE: Suppose you want the 5th. 6th. and 20th. spectra displayed:
Reply 5, 6, 20 OR 5-6,20

When you have picked the spectra you want, type E and the ULDA will terminate. From this you can see that SELECT is not cumulative (which is also true for the other data base commands; for example the output from a search overwrites the previous search's results).


Your output from the ULDA is stored on work files whose names comprise your ULDA ID (see section 3. [LOGIN]) and standard qualifiers. These files are listed bdow.

Note :

  1. The ULDA will normally delete and recreate your work files when you LOGIN to it. Therefore never use someone else's ID since you may destroy his output. For the same reason, do not use communal IDs.
  2. Work files have a lifetime (48 hours) set by your national host manager. After this time has expired, anyone using the ULDA will automatically delete the files.
  3. If a command which writes to a work file (e.g. SAVE DESCRIPTORS) is issued more than once, only the output from the last call is saved.

14  How to Use the ULDA (for Reference)


Logging in to the ULDA will normally involve entering your ULDA ID and a number from a menu. However, certain additional functions are also available to you at startup. For Instance:

16  Giving your ULDA ID

On entry the ULDA will ask you to enter your ID. Such an ID should have a two-letter country ID followed by up to six nonspecial characters and must have been given to you by the ULDA in a pre vlous session. Examples of valid IDs are: UKJO Jo in the UK, WGINGE Inge in West Germany.

Since IDs define your output file names under no circumstances, use someone else's ID since you will erase his data files. If you have forgotten your ID, sec section "Dlsplaylng the IDs and Ihelr Owners' Names for a Country", or if you do not have one, see section "Getting an ULDA ID (Registration)".

17   Getting an ULDA ID (Registration)

To avoid delays the ULDA allows you to register yourself, a process you need to do only once. When the startup prompt is displayed on the screen, reply "R" and the possible countries and their two-letter abbreviations are displayed (the two character codes will be the first two letters of your ID). You are then prompted in turn for each of the following:

  1. The ID of your choice (reflected if invalid or already exists). Keep your ID short and easy to remember. Do not forget it forms part of your output file names; hence you will need it for more than just logging in to the ULDA.
  2. If the first ID you choose already exists. you may look at the extant IDs for your country (see section "Display IDs and Their Owners' Names for a Country") to help you choose one of your preference.
  3. You will be asked for your work address and/or telephone number and/or computer address used if someone has to get in touch with you.
  4. You will be asked to confirm each reply to guard against possible typing errors.

18  Displaying the IDs and Their Owners' Names for a Country

Should you forget your ID or have difficulty choosing one use this command by replying "D" to the startup question as indicated on the screen. The possible countries and their two-letter abbreviations will be displayed.

ID - Country |ID - Country |ID - Country| |
AA - Austria |AR - Argentina |AU - Australia | |BE - Belgium |BR - Brazil |CA - Canada | |CH - China |CI - Chile |CK - Czechoslovakia| |CO - Colombia |DK - Denmark |EQ - Equador | |FI - Finland |FR - France |HL - Switzerland| |IN - India |IR - Ireland |IS - Israel| |IT - Italy |JA - Japan |LU - Luxembourg| |ME - Mexico |NL - Netherlands |NO - Norway | |PE - Peru |PO - Poland |PP - Portugal| |SP - Spain |SU - Soviet Union |SW - Sweden| |TW - Taiwan |UK - United Kingdom |US - United States| |VE - Venezuela |WG - West Germany || |


If you enter the country's ID only, it will display all the registered users of that country.

19   Recovery Mode

If you are unable to downlink all the spectra you chose, you may send the ULDA into recovery mode by giving your motive for using the ULDA as recovery" (in answer to the second question). If you ask for recovery, you will not go through the normal pathway but will be sent directly to the SELECT command where you will be able to choose spectra from those selected during the previous run.

EXAMPLE: Suppose you only managed to downlink and unscramble (using UNSPL) the first two spectra you selected. In this case, reenter the ULDA, go into recovery mode and reply as follows to the Select prompt : 3-100

Your id.SPE work file will now hold only those spectra you were unable to download.

Please note that, when in recovery mode:


Before describing the SEARCH function, it is worth remembering that your goal is the selection of spectra which, as illustrated in the diagram below, the ULDA helps you do in a two-step procedure. First, do a data base search for a set of spectra which includes those you want and secondly select those spectra you want. See the graph below.

The search itself is invoked from the main menu in two forms, differing only in the amount of explanatory text which is displayed. Because of the easy nature of the search and because the "short" format form gives free examples and prompts, this is how you will start your searches except for the first couple of times .

21   Overall Search Rules

You may search on one or more of four types of criteria, each of which you will be prompted for in turn. A carriage return will skip the search menu currently being displayed. The possible search class are as follows:

Putting one search element per line is often the best option. If you make a detectable typing error, the line is discarded. Furthermore, the separation between requests stands out better.

22   Search on Camera and/or Image Number and/or Aperture

Each search element comprises a maximum of three fields written as a contiguous string. These fields are below:

  1. Camera
    1. SWP - ShortWave Prime
    2. SWR - Short Wave Redundant (not used)
    3. LWP - Long Wave Prime
    4. LWR - Long Wave Redundant
  2. Image number = an integer from 1000 upwards.
  3. Aperture: S= Small, L= Large

If all three of the above are given (e.g.. SWP12345L), you will specify a spectrum uniquely (the maximum number of spectra found will be one). However , any combination of one to three of the fields is also valid, provided the order of [Camera] [Image number] [Aperture] is preserved within a query element.

Examples of some valid search requests follow:

In all search requests you can specify a list of criteria separated by commas, spaces, semicolons, or (preferably) by entering each element on a separate line, in which case the list elements will be joined by logical ORs.

EXAMPLE: Find all spectra with image numbers 12345 or 12551 or spectra taken with the SWP and having image number 1673:

CAMERA IM.NO.AND. AP'S: 12345,12551,SWP1673 <CR>

Your understanding of the logic is in good shape if you see that "LWRL" is not the same as "LWR L."

LWRL = Find all spectra taken with the LWR and the large aperture.

LWR L = Find all spectra taken with either the LWR or with a large aperture (useless request)

23   Search on RA and Declinations

Search requests are always composed of pairs of coordinates in the order RA then Declination. Both coordinates of such a pair must be given on the same input line and be separated by one or more blanks. If you wish to enter more than one pair, they should preferably be on separate input lines, though they may be on the same line, separated by one or more spaces commas. or semicolons. Valid coordinates are one- two- three- four- or six-digit integers which may be signed for Declinations only. In addition, either coordinate may be entered as "ALL" or "*" (in which case the search will be made on the other coordinate only).

EXAMPLES: 5 hrs. 8 mins. (of RA) must be represented by 0508; however 5 hrs. of RA can be entered as either 5 or 05.


Each sky position you enter actually is a search window whose dimensions are determined by the accuracy with which you specify the coordinates. For example, if you give hours of RA, only you will get a larger range of RA (by default) than if you specify hours and minutes, which in turn give a larger default range than if you specify the RA down to seconds. The "fuzziness factors" for the various previsions of coordinates are given in the following table, which you can display anytime with the RA./Dec. help (i.e,"?").


|| Search Window Sizes || ||
||R.A. ||DECLINATION || ||
Accuracy Used ||Window Size ||Accuracy Used ||Window Size|| ||
Hours Only ||+30 Mins. ||Degrees Only ||+4 Degs.|| ||Hours + Minutes ||+10 Mins. ||Degrees + Minutes ||+/-25 Mins.|| ||Hours +Mins +Secs ||+30 Secs. ||Degrees + Mins + Secs. ||+/-1 Min.|| ||

See the following diagrams for an illustration of how implicit ranges work

Search Windows for Two Different Requests for RA= 8 Hrs Dec.= 6

For request = 86 For request = 08006 Search window: Search window


Window generated by "8 ALL"


24   Search on IUE Object Classes

IUE object classes are integers in the range 0-99 whose meaning you can display within a search by using the object class search query's help (i.e. "?"). If you want to search on more than one class at a time, separate them by one or more commas, blanks or semicolons , or enter them on separate lines.

EXAMPLE: Search for pulsars and supernova remnants.


|OBJECT CLASSES 0 - 29 | |0 Sun |10 WC |20 B0-B V-IV | |1 Earth |11 WN |21 B3-B5 V-IV | |2 Moon |12 Main Sequence O |22 B6-B9,5 V-IV | |3 Planet |13 Supergiant O |23 B0-B III-I | |4 Planetary satellite |14 Oe |24 B3-B5 III-I | |5 Minor planet |15 Of |25 B6-B9,5 III-I | |6 Comet |16 sd O |26 Be | |7 InterplaneL medium |17 WD O |27 Bp | |8 Giant red spot |18 |28 sdB | |9 |19 UV-strong |29 WDB| |
| OBJECT CLASSES 30 - 59 | |
30 A0-A3 V-IV |40 F0-F9 |50 R,N,S TYPES | |31 A4-A9 V-IV |41 F3-F9 |51 Long Period Variable| |32 A0-A3 III-I |42 Fp |52 Irregular Variable| |33 A4-A9 III-I |43 Late-type degen.stars |53 Regular Variable| |34 Ae |44 G IV-VI |54 Dwarf Novae | |35 Am |45 G I-II |55 Classical Novae| |36 Ap |46 else K IV-VI |56 Supernovae| |37 WDA |47 K I-III |57 Symbiotic Stars| |38 Horizontal Branch |48 M dwarf |58 T Tauri | |39 Composite |49 M I-III |59 X-ray | |
|OBJECT CLASSES 60 - 79 | |
60 Shell Star |70 Central Star of PN || |61 Eta Carinae |71 Planetary Nebula || |62 Pulsar |72 H II Region || |63 Nova-like |73 Reflection Nebula || |64 Stellar objects |74 Dark Cloud || |(others) |(absorption spectrum) || |65 misidentified |75 Supernova Remnant || |66 Interacting Binaries |76 Ring Nebula || |67 |77 || |68 |78 || |69 Herbig - Haro object |79 || |
|OBJECT CLASSES 80 - 99 | |
80 Spiral Galaxy |90 Intergalactic Medium || |81 Elliptical Galaxy |91 || |82 Irregular Galaxy |92 || |83 Globular Cluster |93 || |84 Seyfert Galaxy |94 || |85 Quasar |95 || |86 Radio Galaxy |96 || |87 BL Lacertae object |97 || |88 Emission Line Galaxy |98 Wavelength Calibration || |(non Seyfert) |(NASA log) || |89 |99 NULLS & FLAT fields || | |(NASA log) || |

25   Search on homogeneous Object IDs

The IDs comprise a catalogue (one to four letters) followed by the object's ID within it. If the latter has two parts which are not a pair of celestial coordinates (e.g.. BD -17 4554), these must be separated by one or more spaces. Otherwise, do not insert blanks. Spaces between the catalogue and the ID are optional except for catalogues which are prefixed by AOO (Any Other Oblect) when a space is required after the catalogue.


  1. NGC6960 is equivalent to NGC 6960
  2. BD-17 4554 is equivalent to BD -17 4554 however BD-17 4554 is incorrect.

If you specify more than one ID to search on, it is preferable to put them on separate input lines, though you can enter them on a single line separated by commas or semicolons. Please note that unlike all the other search queries a space is not a request separator.

The catalogues ordered by their priority are as follows:

IUE |IUE Related (e.g.WAVECAL's) |HD |Henry Draper| |BD |Bonner Durchmusterung |CD |Cordova Dutchmusterung| |CPD |Cape Durchmusterung |V* |Variable Stars (Kukarkin)| |WD |White Dwarfs (Villanova Un.) |GD |Giglas White Dwarfs| |NGC |New General Catalogue |IC |Index Catalogue | |PK |Perek-Kohoutek (Plan.Neb.) |SK, AZV, LMC, SMC, FD |(Magellanic clouds)| |MRK |Markarian |QSO |Hewitt + Burbidge Quasar Catalogue | |MCG |Morphol. Cat. of Galaxies |ESOBE SO(B) |Uppsala Catalogue | |PG |Palomar Green Catabgue |ABCG |Abell Rich Clusters | |Z |Solar System |NOVA |Novae | |SN |Supernovae |AOO |Any Other Objects (AOO)| |

AOO is composed of catalogues not listed above plus those objects which could not be uniquely identified.

Note: Spaces are optional between the above catalogue IDs and their entry but mandatory between an AOO's subcatalogue and Its entry.


DISPLAY is invoked from the main menu and shows what the last executed search found, one screen at a time, up to a host center dependent number of records (400 when distrlbuted). See the following sample output (the search was for oblect class 40).

Ref |Cam+Image |Homogen- |R.A. |Decl |Observ. |Exp.T.|Ob |Usage |Code| |No |+Aperture |eous Id |hh:mmss|dd:mmss |Date |Secs. |Cl |Dubious || |
58|LWR10004L |HD15164 |2:2406 |10.2130 |24FEB81 |600.0 |40 | || |59|SWP13426L |HD99028 |11:2119 |10:4815 |7MAR81 |220.0 |40 |D || |60|LWR10090L |HD99028 |11:2119 |10:4815 |7MAR81 |55.0 |40 |D || |61|SWP13668L |HD118216 |13:3234 |37:2616 |6APR81 |1500.0 |40 |U || |62|LWR 5230L |BD+39 4926 |22:4354 |39:5059 |2AUG79 |420.0 |40| |40| |63|SWP 6044L |BD+39 4926 |22:4354 |39:5059 |2AUG79 |2400.0 |40| |40| |64|SWP 6512L |HD178524 |19:0646 |-21:0559 |14SEP79 |1260.0|40||702 | |65|SWP 6514L |HD163506 |17:5323 |26:0259 |14SEP79 |1500.0|40| |501 | |66|SWP 7344L |HD118216 |13:3233 |37:2559 |8DEC79 |2100.0|40| |741| |67|LWR 6595S |HD132322 |14:5725 |-63:4359 |9JAN80 |300.0|40| |503| |68|LWR 6595L |HD132322 |14:5725 |-63:4359 |9JAN80 |600.0|40| |703| |69|SWP 7619S |HD132322 |14:5725 |-63:4359 |9JAN80 |900.0|40| |502| |60|SWP 7619L |HD132322 |14:5725 |-63:4359 |9JAN80 |1800.0|40 ||702| |61|SWP 7793S |HD 93614 |10:4556 |2:3900 |27JAN80 |220.0|40 ||501| |62|SWP 7793L |HD 93614 |10:4556 |2:3900 |27JAN80 |260.0|40 ||701| |63|LWR 6005L |HD 31964 |4:5822 |43:4500 |1NOV79 |120.0|40| |801| |64|SWP 7063S |HD 31964 |4:5822 |43:4500 |1NOV79 |960.0|40| |721| |65|SWP 7063L |HD 31564 |4:5822 |43:4500 |1NOV79 |1200.0|40| |821| |66|SWP 9375L |HD175813 |18:5521 |-37:1027 |26JUN80 |3000.0|40| || |

The Usage/Dubious column contalns flags, "U" and "D" respectively, whose presence indicates that auxlllary lnformation is available and what type it is (see section: (Auxiliary Information: Dubious and Usage Data). Usage data wlll tell you who has selected thls spectrum prevlously while "Dubious" comments usually indicate why the flux units of the spectrum mnay be in doubt or should be treated wlth care. e.g.. trailed spectrum. no exposure time avallable.


This function, which is executed from the main menu, saves the descriptors of what the last search found, onto your work file id.DES (where "Id" is your ULDA ID). The format on the file is almost identical to DlSPLAY's wlth the text prefixed by your search request. Like DISPLAY only the first 400 finds are saved.

One possible applicatlon of the command is for those who would like to look at the list of spectra at their lelsure.

Note: If you issue the command more than once, your work file will hold the result of the last Save Descriptors only. However, remember the descriptors of any spectra you subsequently select will be appended automatically to your descriptor work file.


SELECT, which is started from the main menu enables you to pick those spectra of interest from those found by the last search. The spectra you choose will be stored in your work file id.SPE (where id is your ULDA ID), their descriptors in id.DES and their usage data appended to id.AUX.

Before starting the selection you are asked if you want the spectra that you will select to be encoded in compressed ASCII format (the default output being 2K binary blocks).


If you Intend to E-mail the data to your computer or will be using a KERMIT of limited power then ask for ASCII encoding (each spectrum, its header and epsilons call be compressed and converted into universally recognised ASCII characters and altogether will take between 20 and 26 lines of text, each 72 characters long). Please note that ASCII encoded spectra will be illegible to all programs except the UNSPL which will convert spectra into the format you want.

SELECT displays the descriptors of the spectra found by the last search in a format and a manner identical to those used by DISPLAY. The only difference is the bottom prompt lines which include an example and read as follows:


You select spectra by using the reference numbers in the first column of the display. You may either select spectra individually or as ranges by entering the reference numbers or ranges separated by commas spaces or semicolons or by putting them on separate input lines.

EXAMPLE: Choosing individual spectra: 3, 7, 9, 14
EXAMPLE: Selecting ranges of displayed spectra: 8-14

Should you specify a spectrum more than once, only one copy will be made.


  1. When SELECT ends so does the ULDA and you are automatically returned to the IUE main menu. If you want to SAVE DESCRIPTORS or look at Dubious Informatlon, this must be done before starting SELECT. However, SELECT will automatically append the descriptors of any spectra you selected to your descriptor work flle.
  2. There is a country dependent upper limit on the number of spectra you can select at a time (about 40).
  3. If you have a line error while downloading your spectra, you can specify RECOVERY when loging in. You will be sent directly to SELECT, where you will then be able to select spectra from those you selected last time (ie. normally those that you failed to download).
  4. To select all the spectra found just enter 1-1000. (If the upper limit exceeds the number of spectra you have available to choose from, it is taken to mean "to the end.")
  5. Each selection of a spectrum you make will be recorded in a "usage data" file and be flagged as having been used. These flags will show up in subsequent DISPLAYs
  6. If before terminating you choose any spectra which have "usage data " SELECT will inquire if you want to see it. If you say "yes", the dates selected, motivations and names and addresses of those who have previously chosen the spectra are displayed for each spectrum. In addition, the information is appended to your id.DES work file.

29   AUXILIARY INFORMATION (Dubious and Usage Data)

Two forms of auxiliary information are available as described below. Both are being output to your id.AUX work file (where "id" is your ULDA ID), if they are displayed.

For dubious comments display and filing use the DUBIOUS command in the main menu. All the dubious comments are displayed and filed (on your id.AUX work file) for the first 400 spectra found by the last search. For usage data display and filing use the SELECT command. When this terminates, if any spectra have been selected that have usage data, you will be asked if you want to see it. If you reply "yes ", it will be shown on your terminal and appended to your id.AUX file.

30  How to Convert Selected Spectra into a Usable Form (UNSPL)

Since the output from the ULDA will be in a compressed format, a program (UNSPL) is supplied by your national host site to unscramble your file of selected spectra and split it up into separate files, one per spectrum, with a format suitable for immediate processing. Every end user institute will have its own version of UNSPL running on its computer tailored suitably if necessary to its particular needs.

A useful additional feature of UNSPL for users connected to their national center by DECnet is the optional automatic down link of their work files:


To execute the UNSPL, type in "UNSPL". UNSPL will ask you a few simple questions to determine what it has to do.

  1. UNSPL has to know in what form the input data are, i.e., binary input in 2K byte blocks or compressed encoded ASCII. The input data form is what you requested of ULDA's SELECT command. The first question is: UNSPL: DID YOU ASK QUEST FOR ASCII CONVERSION |Y/N| ?
  2. Next you are asked what output format you want: FITS, ASCII or MIDAS, which are described in Appendix A, may change with subsequent updates of software. In addition, a dummy format called LOCAL is available which you can use to install your institute's own output format. If you choose MIDAS, then UNSPL must have been started under a MIDAS session.
  3. Please contact your ULDA host for user directory location information. See Appendix C.
  4. If you replied N to the downlink question, give your input file name. For example: UNSPL: NAME OF USER WORK FILE:. UNSPL will retrieve your ULDA data files from the national host site unscramble/unsplit the data and write them out to your directory on your host computer. The files will be named as follows: ULDAuserid.SPE, ULDAuserid.AUX. ULDAuserid.DES


32  Appendix A: UNSPL Output Format Descriptions


The user's compressed spectral work file is split into a number of ASCII files containing one spectrum each. Such files have a table format: the first column contains wavelengths, the second contains absolute fluxes (ergs/sec/cm2/A), and the third contains quality flags in standard IUE format (epsilons). This format can be used as input to a surprising number of image processing systems and it allows printouts of spectral data.

The name of each output file is built with the camera, image number and aperture keywords; the file extension is ASC (e.g.. the file concerning image number 12460 obtained from camera SWP with large aperture, will be assigned the name SWP12460L ASC). Records may be read subsequently using the FORTRAN statement : READ (n.*) WAVELEN,FLUX,IEPSIL

34   FITS FORMAT (IUE Extension)

Another standard output format provided by UNSPL is FITS, with the "Image' extension. Each output file contains one spectrum in the format described in the following.

Block 1 of the FITS file contains the file header. The following FITS keywords are filled, some with fixed values, some with the contents of the user work file record header.

Block 2 of the FITS file contains the spectrum pixels as a string of binary half words, the number of pixels being given by the NAXIS 1 keyword in the FITS header.

Block 3 of the FITS file contains the extension header. The following FITS keywords are filled, some with fixed values, some with the contents of the user work file record header

Block 4 of the FITS file contains the error flags, one per pixel, as a string of binary half words. The number of pixels is given by the NAXIS1 keyword in the FITS header.

The name of each output file is built with the camera image number and aperture keywords; the file extension is .ULD (e.g. the file concerning image number 12460 obtained from camera SWP with large aperture will be assigned the name SWP12460L.ULD).

35  Appendix B: MIDAS FORMAT

MIDAS is a data processing system commonly available in Europe. The compressed user work file is split into a number of files in MIDAS format: The data of each spectrum constitute a one- dimensional "bulk data frame" (BDF) file, while an additional file of type "table" is provided to hold the epsilons of each spectrum. A table is composed of three columns: The first contains the wavelengths of all the data flagged as of "bad quality" data, the second contains the value of the quality flag and the third contains the "bad quality' data themselves.

The name of each output file is built with the camera, image number and aperture keywords; the file extensions are .BDF for the bulk data frame file and .TBL for the table file (e.g.. the files concerning image number 12460 obtained from camera SWP with large aperture will be assigned the names SWP12460L.BDF and SWP12460L.TBL respectively).

36   Appendix C: LOCAL FORMAT

At UNSPL installation time, itis possible to provide a routine allowing output of spectral files in a locally-defined format. If such a format is available at your institute, refer to your own system's data processing documentation for additional information.

© Université de Strasbourg/CNRS

    • Contact