April 2006 README FILE FOR TREASURY PROJECT DATA DELIVERED TO MAST J.C. Martin, K. Davidson (UMN), K. Ishibashi(MIT), & R.M. Humphreys (UMN) University of Minnesota (If you use information or advice from this memo, please acknowledge it and the net site http://etacar.umn.edu in any resulting publications; thanks. A full listing of the people who contributed to the Treasury Project is at the end of this document.) 1. INTRODUCTION This is our first delivery of data to the HST MAST archive by the HST Treasury Project for Eta Carinae. At this time the publicly-available archive includes all of the "Version 1.2.1" HST STIS/CCD spectroscopic data images, covering the approximate wavelength range 170-1000 nm at epochs from January 1998 to March 2004. STIS/MAMA UV Echelle data will be added soon. The HST/STIS CCD data include several improvements over the normal STScI pipeline and standard CALSTIS reductions. The data were reduced using CALSTIS_TP 1.0 which is CALSTIS 6.8 customized for the Eta Car Treasury Project. Some of the improvements are substantial and would be applicable to other STIS data. Details concerning the reductions are outlined in Section 2. Section 3 lists other caveats and warnings pertaining to this data. Additional processing and software will become available as time progresses. For more information on the full extent of the HST Eta Carinae Treasury Project please visit our web site: http://etacar.umn.edu When you use data from this archive in a publication, in addition to any acknowledgment to MAST, please mention the Treasury Project in the text and include the following or equivalent wording: "This paper has used data from the Hubble Treasury Program for Eta Carinae, http://etacar.umn.edu, supported by grant GO-9973 from the Space Telescope Science Institute. The STScI is funded by NASA." A full listing of all the co-investigators and contributors to the Treasury Project is given at the end of this document. Several additional detailed documents of interest are archived at: http://etacar.umn.edu/treasury/techmemos/ 2. ITEMS OF INTEREST The following items pertain to the release of Version 1.2.1 of the database to the MAST archive: a) FILE NAMING SCHEME Each data file has a name which indicates useful details of that observation. A file name such as "h_cd29_0020" has four parts: -- "h" is the archive code, merely signifying a high-level data product in the MAST archive.; -- "c" identifies the instrument (see below); -- "d29" is a simple three-character code for the observation date; -- "0020" is a running number for the observations in a given set. For clarity, two underscore characters ("_") always occur at the locations shown above. Details: -- The instrument code is "c" for the STIS/CCD or "e" for the STIS/MAMA/ECHELLE data. -- The 3-character date code indicates the observation year (1998=8, 1999=9, 2000=a, 2001=b, 2002=c, 2003=d, 2004=e) and fractional year, calculated from the Modified Julian Date (MJD): f(MJD) = 2000.0 + (MJD - 51544.5) / 365.25. Thus "a20" corresponds to 2000.20 on MJD 51616.5 or 2000 March 13th. -- The observation number (four characters) identifies the sequential order for each observation date. The first observation for a given date and instrument is normally labeled "0010," some numbers are skipped, and the sequential order is determined by the FITS header keyword TEXPSTRT. -- In some cases a suffix "p" is appended to the observation number to indicate that some original data pixels were saturated and have been patched (see note 2e below). In each of these cases the unpatched data file, without a "p" suffix, is also available. Two examples: h_cd29_0020 = STIS/CCD data, visit on 2003.29, second exposure in the set h_ce18_0250p = STIS/CCD data, visit on 2004.18, twenty fifth exposure in the set, overexposed pixels have been patched. b) IMPROVED PIXEL INTERPOLATION Improved techniques used to model and interpolate the pixels in the STIS CCD are outlined in Davidson (2005) and described in detail in an upcoming publication (Davidson et al., in preparation) and Technical Memo #1 on our web site: http://etacar.umn.edu/treasury/publications/pdf/tmemo001.pdf This procedure has improved the effective resolution of the STIS CCD on both axes since the previously used scheme effectively acted as a position dependent blurring function. Data reduced by the new method are of a quality which surpasses the standard data processing techniques for STIS CCD data. The "scalloping" which is obvious in narrow extractions from data reduced by the former methods is mostly eliminated by this new method. A typical one row extraction made from CCD data reduced using the old interpolation technique has periodic noise with an amplitude from 20% - 30%. Our technique reduces this scalloping to less than a few percent. See the illustration in Technical Memo #1 and Davidson (2005). Also note that the reduced data have been rebinned so that one pixel in the raw data corresponds to two pixels in the reduced data. The effective pixel scale in the reduced data is therefore about 0.025035 arcsecond/pixel. In any discussion it is very important to distinguish between "reduced pixels" and "instrument" (or "original") pixels. c) INITIAL BAD/HOT PIXEL REMOVAL Bad pixels (hot pixels, cosmic ray hits, etc.) were identified in the STIS/CCD data by two independent techniques. First, we applied the standard CR-SPLIT method which is familiar to HST users. This method requires at least two exposures (a condition that was not always met) and can miss a few bad pixels. Therefore, we developed a second technique which identifies bad pixels without requiring multiple exposures. Our method relies on the fact that a legitimate point source is sampled by several pixels on the STIS CCD. Each pixel is compared with its neighbors in 5x5 square centered on that pixel. After rejecting the two highest and lowest neighboring values, we calculate the mean (V_avg) and the r.m.s. dispersion (D) of the 20 remaining comparison pixels. The R value of the pixel being tested (V_0) is assessed by the expression: R = |V_0 - V_avg| / D. In the cases where R exceeds a well-chosen threshold the pixel is flagged as bad. By iterating this process, sizable clusters of bad pixels can be identified and eliminated. This method was carefully tuned and tested so that it eliminates nearly all the obviously bad pixels in a single exposure without adversely altering the noise structure of the data. d) FLAGGING BAD PIXELS WHICH RESULT FROM EXPOSURE SATURATION Interactions between the distortion correction and the pixel interpolation technique make it a non-trivial task to identify pixels in the final reduced data which may have been adversely affected by interpolation. That includes pixels saturated by overexposure in the raw data. Most saturated pixels involve Balmer H-alpha or H-beta emission, see section 2e. Unfortunately, one cannot rely on the Data Quality (DQ) array in the FITS files to properly identify all bad pixels (see 3b in Caveats & Warnings section below). A list of the observations which have over-exposed pixels (identified from the raw data files) is given in the Appendix. In most cases the raw exposure with saturated pixels has a shorter companion exposure with the same grating tilt and target at the same epoch. In those cases, we were able to compare the reduced data from the long exposure (with the saturated pixels) to the shorter exposure (without saturated pixels) in order to determine which pixels have been adversely affected in the longer exposure. Very conservative tolerances were used to identify the bad pixels in the longer exposure. The pixels identified as having been affected by overexposure have been assigned a value of NaN (not-a-number) in the flux data array so that it is unavoidably obvious that those pixels are bad. We refer to NaN as defined in IEEE 754. We use a quiet NaN (QNaN) which is explicitly generated by the expression: NaN=(-999./0.0). Twenty nine of the overexposed observations have no companion short exposure, so we cannot easily identify the pixels which are adversely affected by overexposure in those cases. A history entry has been added to the primary and FLUX HDU of these FITS files to note that they possibly contain bad pixels which have not been properly identified. e) PATCHING OF SATURATED 6768AA AND 4961AA EXPOSURES The STIS CCD observations of the Hydrogen Balmer lines at 6768AA and 4961AA are pairs of one long and one short exposure. Each long exposure overexposed the core of the Balmer lines from the central star in order to obtain a satisfactory signal strength at other wavelengths. We have combined these exposure pairs into single images with a larger combined dynamic range than a single STIS CCD exposure allows. Each of these higher level science products contains an extra FITS image header (HDU #5 [MASK]) which is the array of weights used to combine the pixels from the long and the short image. In the case of these combined images the error (ERR) array contains the standard deviations of the weighted individual pixel values that went into building the flux array. The names of the files affected are marked with a suffix "p" for "patched." The unpatched data are also present in the archive. f) WAVELENGTH CALIBRATION The wavelength calibration has been rigorously verified by measuring the wavelengths of narrow stellar photospheric spectral features in standard stars reduced in the same manner as the Eta Carinae data. The current wavelength calibration meets or exceeds the standard set by the STScI pipeline and CALSTIS. Remaining wavelength errors may result from imperfect pointing of the STIS slit. A number of the observations have no standard WAVECAL data. For them the wavelength scale was determined by cross correlating with a known feature (i.e. lines in the Weigelt knots). See Section 3a. g) FLUX CALIBRATION The flux calibration of the data was checked by comparing standard stars BD +75 325 and AGK +81 266, reduced by our technique, with data published by Bohlin et al. (2001). The current flux calibration meets or exceeds the standard set by the STScI pipeline and CALSTIS for extended source spectral flux calibration across the entire range of wavelengths and epochs covered by this data. Caveat: The absolute fluxes are not of photometric precision because the target is spatially complex and a narrow slit was used. A possible point of confusion regarding the flux array has been clarified by adding notes to the FITS headers. The units in the FLUX array (as given by the BUNITS keyword) are: ergs/sec/cm^2/Angstrom/reduced-row "Reduced-row" refers to a single row in the reduced data array NOT an original CCD row. We have chosen this convention so that the flux can be integrated across the rows in the FLUX array without applying a scalar correction to account for the pixel interpolation. However, the units on the ERR array are different from the FLUX array. They are in ergs/sec/cm^2/Angstrom/original-CCD-row. The reason for this is discussed in section 3d. Note that when a 1D spectral extraction is made from the 2D spectra in the database the extracted spectrum is still uncorrected for the extraction height. To convert this data to flux units of erg/sec/cm^2/Angstrom/arcseconds^2 divide by a factor C: C = (effective slit width, arcseconds) x (effective pixel size, arcseconds) For the 52x0.2 slit C ~ 0.005007 arcseconds^2 and for the 52x0.1 slit C ~ 0.002504 arcseconds^2. However, be aware that those are only approximations of C. The actual value may vary significantly due to changes in the effective width of the slit and/or the illumination of the slit which can be wavelength dependent and cannot always be predicted. i) CENTERING OF THE NOMINAL TARGET ON THE CROSS-DISPERSION AXIS We have determined the location (fractional row number) of the nominal slit center for each observation. The following algorithm was used to determine these reference rows: 1) If the central star is present on the slit, then its actual position (as determined from the average row of the cross dispersion peak in the continuum across the CCD) is used as a reference point to determine the fractional row number for the slit center. 2) If the slit for an observation does not include the central star but that observation is immediately preceded or followed by a matching observation with the same grating parameters which has its reference row determined by method #1, then the reference row from the matching observation is used. In other words we trust the stability of the instrument and the HST offset parameters. 3) If neither of the first two methods applies, then a plot of the cross dispersion axis from the observation is matched with a slice extracted from HST ACS/HRC images to determine the reference rows. This method has a much lower accuracy; up to plus-or-minus half a row. The WCS information in the FITS headers has been edited so that the CCD row where the nominal target appears is identified by the keyword CRPIX2. Also, as per the findings of Bowers & Baum (1998), the cross-dispersion pixel scale has been set to 0.025035 arcseconds/reduced-pixel (CD2_2 FITS header keyword). As a result, the WCS information should return the proper coordinate along the slit, with the nominal target located at zero. In most cases, the centering of the nominal target is accurate to better than +/- 0.0127 arcseconds (half a reduced pixel). It is important to note that some observations which are nominally targeted on the Weigelt blobs are not centered on those features. Those observations did sample the Weigelt blobs, but we usually tried to include more than one of them so that the resulting slit position was a compromise. This means that sometimes the slit includes the blobs even though it is not centered on them (the reference row does not pass through them). j) WORLD COORDINATE SYSTEM (WCS) Standard World Coordinate System (WCS) information is included in each of the image headers for the STIS/CCD data. The dispersion axis coordinate (X = columns) is set to return the position of each pixel in Angstroms. The cross-dispersion axis coordinate (Y = rows) is set to return a value in degrees from the slit center (see 2i above). The WCS information on both axes is defined by first order linear terms. CAUTION: So far as we know, the absolute position of Eta Carinae (the central star) has never been measured with high absolute astrometric accuracy because the bright ejecta makes measuring the star's position difficult. Our relative X and Y grid centered on the star is good to about 0.01 arcseconds, but the absolute RA and Declination of (X,Y) = (0,0) may be uncertain by 0.1 arcsecond or worse. For more information see: http://etacar.umn.edu/etainfo/basic/astrometry/ 3. CAVEATS AND WARNINGS a) IN SOME CASES THERE IS NO WAVECAL There is no formal WAVECAL (STIS/CCD wavelength calibration) associated with 92 of the observations in the data archive. In those cases the zero point for the wavelength calibration is uncertain. The cases where this occurs have been noted by adding comments to the CRVAL1 and WOFFSET values in the FLUX HDU of the FITS files which clearly state that there is "NO WAVECAL" for that dataset. The data affected are listed below. Ninety two (92) of the STIS CCD spectra were taken with no WAVECAL calibration data sets in order to optimize the scientific return in a limited number of available HST orbits. In most cases we have been able to calibrate the affected data using narrow, stationary nebular emission features (e.g., Weigelt Blobs). We have verified the accuracy of this method by comparing the diffuse interstellar bands in data with and without formal WAVCALs. Note that the stellar spectrum (or the reflection of it) is not a suitable source for wavelength calibration; hence we avoid using it for the purpose. In 78 of the 92 cases we used the narrow emission spectra of the Weigelt B and D blobs to calibrate the wavelength scale of the STIS CCD. In those cases the Weigelt blob spectrum was calibrated using the data 1998 March 19 where they were observed with roughly the same slit orientation. This class of calibration includes observations made on: 1999 February 21 (files beginning with h_c914), 2000 March 20 (files beginning with h_ca22 numbered 0010 through 0350) and 2004 March 7 (all files beginning with h_ce18). For the rest, there are no WAVECAL or Weigelt knot spectra. Therefore we have used fainter nebular emission lines (sometimes the reflection nebula) to determine the zero wavelengths. This type of calibration can be very uncertain and should not be implicitly trusted under any circumstances. However, observations of diffuse interstellar bands in these data lead us to believe that the possible errors are no worse than 10 -- 20 km/s and are probably just as good as the other methods. This affects the following datasets: 1998 November 25 (files beginning with h_c890 numbered 0010 to 0110), the G230MB (cenwave= 1713AA) spectrum in the visit on 2000 March 13 (file a_ca20_0010), and STIS CCD data in the visit on 2000 October 9 (files beginning with h_ca77). b) DO NOT TRUST THE DATA QUALITY (DQ) ARRAY At present, the data quality (DQ) array in the FITS files follows a convention established by the STIS Instrument Development Team (implemented in the later versions of CALSTIS including CALTIS 6.8 and CALSTIS_TP 1.0) rather than the convention established in the STIS Instrument Handbook. This means that only the "worst" thing that occurred to a pixel during reductions is recorded in the DQ array rather than an audit trail of all of the errors encountered by that pixel. We also have some justifiable concerns that the DQ array might not properly identify all the pixels in the reduced image which have been adversely affected through interpolation by overexposed pixels in the raw image. For example, not all of the pixels in the long exposures affected by overexposure have received the correct DQ flag. As a result, we suggest that users consider the DQ array a best guess as to the status of a pixel and place more trust in flagging of bad or overexposed pixels which has been done by us post-production. c) BEWARE POSSIBLE PSF CHANGES ON THE STIS CCD The PSF of a point source on the STIS CCD depends on grating, spectral wavelength, and column number on the CCD. We plan to address these issues in a later release of the database. In the mean time, we have documented several of these effects on our web site in Technical Memo number 2: http://etacar.umn.edu/treasury/publications/pdf/tmemo002.pdf d) CAVEATS CONCERNING THE ERROR ARRAY A typical HST data file customarily includes a FITS header-and-data-unit (HDU) labeled "ERR," giving an estimate of the statistical counting noise in each pixel. Essentially this is just: {(number of counts in the pixel) + (readout noise)^2 }^(1/2) multiplied or divided by various efficiency factors so that it is expressed in the same units as the main data stored in the "SCI" HDU (usually fluxes) (see page 12 of the CALSTIS Users Manual). ERR is a somewhat misleading label, since it refers only to the simplest form of statistical noise estimate, and ignores many systematic, calibration, or instrumental errors that can and do occur. ERR files are most applicable to faint objects where the S/N ratio is fairly low, while other sources of error, far more difficult to quantify, often dominate for measurements in high-S/N data. The STIS data have been rebinned and interpolated to correct for instrumental distortions, rotations, etc. Unfortunately rebinning and interpolation introduce serious ambiguities or even fallacies in assessments of statistical counting noise, since by nature they correlate values in adjacent pixels. Thus, if we generate an ERR file from the formally "correct" standard errors for individual interpolated pixels, then those values depend in a complex way on local interpolation details. It is common for strange large-scale patterns to occur while is difficult to deduce the local amount of this effect based only on the ERR file. In general, it is very difficult to assess the true r.m.s. uncertainty of a local pixel sum or average from an ERR file of this type. For practical reasons, therefore, the Treasury project data has ERR arrays which represent the statistical noise in the area sampled by one original pixel. Our subpixel-modeling technique (an unusual form of interpolation) has a subtle technical advantage in this connection: the ratio of the actual interpolated ERR value to the ERR value calculated from the original pixels ~ 0.8 everywhere in each data image. Since the reduced pixels are half the size of the original pixels, it is also necessary to renormalize them so that they are expressed in the same physical units as the data values. A full detailed explanation of the contents of the ERR array is given in Technical Memo #7: http://etacar.umn.edu/treasury/techmemos/pdf/tmemo007.pdf It is preferable to use some alternative means to measure the noise in the data (i.e. the standard deviation of flux values in a patch of continuum) since: 1. it is difficult to interpret the ERR array correctly due to pixel interpolation and 2. the ERR array does not account for all the sources of error. e) SCATTERED LIGHT/DIFFUSE BACKGROUND AND GHOST IMAGES A scattered light/diffuse background is present in the STIS/CCD data. The level rises with increasing wavelength and can be significant at long wavelengths in exposures which are long and/or have a very bright source on the slit. A detailed explanation of this affect is found in Technical Memo #6: http://etacar.umn.edu/treasury/techmemos/pdf/tmemo006.pdf There are also ghost images in the STIS CCD optical system (Gull et al., 2002) present in the data. The most obvious place example is in exposures of the central star including the strong Balmer H-alpha line. In those cases a ghost appears below and to the right of the actual feature (decreasing row number, increasing column number). Do not mistake this ghost for diffuse emission from the surrounding nebula! We have produced software which removes the H-alpha ghost (see http://etacar.umn.edu/software). See Technical Memo #10 for details: http://etacar.umn.edu/treasury/techmemos/pdf/tmemo010.pdf 4. REFERENCES Bohlin, R.C., Dickinson, M.E., & Calzetti, D. 2001, AJ, 122, 2118. Bowers, C.W. & Baum, S. 1998, "Plate Scales, Anamorphic Magnification & Dispersion: CCD Modes," STIS Instrument Science Report 98-23, (STScI: Baltimore). Davidson, K. 2005, "Some Neglected Pixel Problems," presented at the 2005 HST Calibrations Workshop. (A. Koekemoer, L. Dressel and P. Goudfrooij, eds.) Gull, T., et al., 2002, "The STIS CCD Spectroscopic Line Spread Functions" presented at the 2002 HST Calibrations Workshop. (S. Arribas, A Koekemoer, and B. Witmore, eds.) Lindler, D. 2003, "CALSTIS Users Manual," NASA Goddard Spaceflight Center: Greenbelt. APPENDIX The following is a complete list of observations which are overexposed in part of the observed spectrum described in Section 2d. The list was made by checking the raw HST data for pixels with values over 30,000 counts. Saturated pixels which were caused by cosmic ray hits or hot pixels were ignored in this analysis. The column listing the number of saturated pixels is included to give a hint as to the degree that observation is over exposed. The more saturated pixels, the more overexposed the spectrum is. A star (*) next to the Treasury Project (TP) Label denotes that long exposure has no corresponding short exposure. Therefore, the all _bad_ pixels in those data affected by interpolation of the overexposed pixels remain unidentified. Raw HST # of Pix Dataset TP Label lambda_0 Saturated =========================================== o4j802080 h_c800_0080 6768 69 o4j8010q0 h_c821_0220 6768 115 o4j8010s0 h_c821_0240 6768 90 o4j8010t0 h_c821_0250 6768 305 o4j8011d0 h_c821_0420* 6768 10 o55601hnq h_c890_0020 6768 185 o55601040 h_c890_0040 4961 20 o556010b0 h_c890_0110 6768 135 o556020m0 h_c914_0220 6768 183 o556020n0 h_c914_0230 6768 4508 o556020r0 h_c914_0340 4961 35 o55602150 h_c914_0410 6768 181 o55602180 h_c914_0440 6768 77 o5f102030 h_ca20_0030 6768 477 o5f102070 h_ca20_0070 7283 117 o5f1020e0 h_ca20_0140* 9336 16 o5f103010 h_ca77_0010* 6768 35 o5f103020 h_ca77_0020 6768 137 o5kz01cqq h_ca22_0010 6768 12232 o5kz010l0 h_ca22_0210 6768 159 o5kz010o0 h_ca22_0240* 4451 24 o5kz010u0 h_ca22_0310 4961 199 o5kz02020 h_ca22_0370 6768 268 o5kz02030 h_ca22_0380 6768 13922 o5kz02040 h_ca22_0390* 6768 757 o5kz02050 h_ca22_0400* 6768 240 o5kz020s0 h_ca22_0630 6768 14848 o5kz020t0 h_ca22_0640* 6768 626 o5kz020u0 h_ca22_0650* 6768 57 o5kz020v0 h_ca22_0660* 6768 3 o5kz021b0 h_ca22_0820* 6768 9 o5kz021f0 h_ca22_0860* 6768 4 o5kz021i0 h_ca22_0890 4961 1366 o5kz021j0 h_ca22_0900* 4961 46 o5kz022c0 h_ca22_1190 4961 1369 o5kz022d0 h_ca22_1200* 4961 11 o5kz022e0 h_ca22_1210* 4961 6 o5kz02360 h_ca22_1490 6768 278 o5kz023a0 h_ca22_1530 6768 282 o62r01010 h_cb29_0010 6768 5038 o62r010k0 h_cb29_0200 6768 174 o62r010v0 h_cb29_0310 4961 206 o6ex03010 h_cb75_0010 6768 641 o6ex03030 h_cb75_0030* 9336 926 o6ex03080 h_cb75_0080 7283 7718 o6ex03090 h_cb75_0090* 7283 14 o6ex030a0 h_cb75_0100* 4451 53 o6ex030c0 h_cb75_0120* 5216 129 o6ex030d0 h_cb75_0130* 2375 43 o6ex030e0 h_cb75_0140 4961 328 o6en01010 h_cb75_0160 6768 o62r02010 h_cb90_0010 6768 628 o62r020d0 h_cb90_0130 6768 54 o6ex02010 h_cc05_0010 6768 622 o6ex020k0 h_cc05_0200 6768 603 o6ex020v0 h_cc05_0310 4961 240 o6mo02010 h_cc51_0010 6768 592 o6mo02030 h_cc51_0030* 6768 27 o6mo020v0 h_cc51_0310* 7283 6 o6mo021m0 h_cc51_0580 4961 246 o8gm01010 h_cc96_0010 6768 574 o8gm01060 h_cc96_0060 6768 17 o8gm12010 h_cd12_0010 6768 627 o8gm120v0 h_cd12_0310 4961 208 o8gm21010 h_cd24_0010 6768 580 o8gm41010 h_cd34_0010 6768 483 o8gm330m0 h_cd37_0220 4961 179 o8gm32010 h_cd37_0290 6768 480 o8gm32030 h_cd37_0310* 6768 8 o8gm52010 h_cd41_0010 6768 497 o8gm52030 h_cd41_0030* 6768 2 o8gm521k0 h_cd41_0560 4961 202 o8gm630c0 h_cd47_0120 4961 201 o8gm630h0 h_cd47-0170* 6252 188 o8gm62010 h_cd47_0200 6768 462 o8gm620v0 h_cd47_0500* 7283 4 o8ma72010 h_cd51_0010 6768 475 o8ma82010 h_cd58_0010 6768 395 o8ma821o0 h_cd58_0590 4961 180 o8ma92010 h_cd72_0010 6768 605 o8ma920w0 h_cd72_0320 4961 347 o8ma83010 h_cd88_0010 6768 695 o8ma940e0 h_ce18_0140* 2697 9 o8ma940f0 h_ce18_0150* 4451 17 o8ma940p0 h_ce18_0250 4961 468 o8ma940r0 h_ce18_0270* 6252 26 o8ma940t0 h_ce18_0290 6768 887 o8ma940v0 h_ce18_0310 6768 88 o8ma940w0 h_ce18_0320* 6768 104 ========================================= CONTRIBUTORS TO THE ETA CARINAE HST TREASURY PROJECT =================================================== The Hubble Treasurey Program for Eta Carinae was supported by grant Go-9973 from the Space Telescope Sciences Institute and based at the University of Minnesota, Twin Cities Campus in the School of Physics and Astronomy. Our website is at http://etacar.umn.edu. Principle Investigator: Kris Davidson (University of Minnesota) Co-Investigators: T.R. Gull (NASA Goddard Space Flight Center) K. Ishibashi (M.I.T.) D. John Hillier (University of Pittsburgh) Roberta M. Humphreys (Universiry of Minnesota) Augusto Damineli, (University San Paulo, Brazil) Michael Corcoran (NASA Goddard Space Flight Center) Otmar Stahl (Landessterwarte, Heidelberg, Germany) Kerstin Weis (University of Bochum, Germany) Sveneric Johansson (Lund University, Sweden) Fred Hamann (University of Florida) Nolan Walborn (Space Telescope Science Institute) Manuel Bautista (Inst. Venezolano Invest. Cientifica, Venzuela) Henrik Hartman (Lund University, Sweden) Project Staff at University of Minnesota: Kris Davidson (Professor and PI) Roberta M. Humphreys (Professor and Co-I) John C. Martin (Post Doctoral Research Associate) Michael Koppelman (Undergraduate Research Assistant) J.T. Olds (System Administrator) Matt Gray (System Administrator)