Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/hst/nicmos/documents/isrs/isr_00_001.pdf Äàòà èçìåíåíèÿ: Mon Nov 19 17:07:39 2007 Äàòà èíäåêñèðîâàíèÿ: Sun Dec 23 01:18:44 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: zodiacal light

Instrument Science Report NICMOS 00-001

The NICMOS Exposure Time Calculator: Algorithms and User Interface
A.Sivaramakrishnan, S. Holfeltz, M. Sosey, B. Simon, and M. Robberto May 17, 2000

ABSTRACT This report describes the latest version of the NICMOS Exposure Time Calculator (ETC) developed to support NCS-era HST NICMOS planning. We added a thermal optics library to the ETC, and model various detector behaviors seen during the NICMOS end-of-life period. There are no ad-hoc parameter values in the thermal optics library describing the telescope and the NICMOS camera. This ETC supports imaging through all filters except the grisms. Exposure Time Calculators (ETCs) are used to estimate expected signal to noise ratios given a source brightness and size, and an expected sky background. In the infrared, a knowledge of the thermal properties of the optics is also required. There are two older NICMOS ETCs: a private one by Thompson (1996) and an STScI-supported one by Skinner (1996). The previous STScI-maintained version, written in FORTRAN by C. J. Skinner, and modified by L. Colina, D. Calzetti and D. Daou, uses pre-launch characteristics of the NICMOS camera, and its own collection of throughput and quantum efficiency data. In consequence, it will be hard to maintain when the NICMOS instrument is revived by the NCS cryocooler. Furthermore, instrument and telescope parameters in this ETC were tuned to match observations, especially in the thermal infrared. This replacement ETC uses measured values of all quantities except for Point Spread Function (PSF) information. In particular, the measured emissivities of the HST primary and secondary are used. These result in good agreement with observations (Robberto et al., 2000). Here we describe the user interface of this new NICMOS ETC, which uses

Copyright© 1999 The Association of Universities for Research in Astronomy, Inc. All Rights Reserved.


instrument throughput and data files maintained and updated by the Space Telescope Science Institute (STScI) and data from NICMOS end-of-life observations (Boeker et al., 1999). The NICMOS detector showed an unexpected bump in dark current during warmup: it is not known whether this bump will recur if the instrument is cooled down again. The bump peaks at around 7e per second per pixel at about 82K. Thus cycle 10 proposals should use the `dark current bump present' case when planning observations, as this is the worst-case scenario. The observer can also look at the more optimistic `bump absent' case if they wish. Offsetting this predicted increased dark current is the fact that the DQE gets better with higher temperatures. The expected operating temperature for the detector in the cryo-cooler era is about 75K (cycle 7 temperatures were between 61 and 64 K). This version of the ETC will be available for planning cycle 10 observations with NICMOS. We also describe the core algorithms used by the ETC. For software details see the ISR devoted to maintenance, support and installation (Sivaramakrishnan et al., 2000a): all the source code is made available to anyone interested in inspecting or modifying the implementation details on their own, though STScI does not provide development support for such activity.

Description of the Calculator
This version of the NICMOS ETC uses throughput data in the Calibration Data Base System (CDBS) at STScI instead of separately maintained private data files. It uses the STSDAS package SYNPHOT as its throughput calculator. SYNPHOT uses the CDBS instrument data files, and thus brings NICMOS exposure time calculations into line with other HST instruments (notably STIS, ACS and WFC3). It is hoped than this unification will make for easier data management. All the code used is non-proprietary, and can be distributed without concern for licensing issues. Copies of the CDBS data base are also freely available (eg., through STScI's Starview software package). While this NICMOS ETC derives its high level software architecture and execution thread from a STIS ETC (Hack et al., 1996, with subsequent modifications and redesign by B. Simon), it differs from the STIS and other optical ETCs in that it has to estimate a noise component due to thermal contributions from the telescope and the NICMOS foreoptics. SYNPHOT is not capable of handling internal sources of radiation yet. In order to accomplish this a library of NICMOS-specific thermal and optical transfer routines were added to the ETC. The NICMOS thermal model implemented in this library differs from the one described in the NICMOS ISR-14 in that it contains no ad-hoc adjustment factors to match observations, and it models various components of the HST pupil as seen from NICMOS, following the work of Krist et al. (1999). Background from sky is presumed to be zodiacal in origin. A model of zodiacal light from the previous STScI NICMOS ETC is used (cf. the one described in NICMOS ISR14). This is an average of zodiacal light at 45 degrees from ecliptic. Table 1 indicates the values used in photons per second per steradian per unit area per unit wavelength interval.

2


Point source signal is estimated in apertures of 1, 1 and 2 arc seconds diameter in the NIC 1, NIC 2 and NIC 3 cameras respectively. Signal for point sources is the number of photons (or electrons --- the numbers are equal for HgCdTe FPA's) detected in the brightest central pixel, assuming the image is centered on that pixel. The Tiny Tim PSF simulator is used to calculate the point spread function. COUNTRATE provides the pivot wavelength for the source/filter/detector combination, which is parsed out of the SYNPHOT results on the source. The aperture correction is estimated using a grid of aperture corrections obtained from Tiny Tim simulations. These simulations reside in the SCIDATA area of STSDAS (Bushouse, 1998). Table 1. Zodiacal background as a function of wavelength
Wavelength microns 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 Zodiacal Background (number s-1cm-2sr-1¨-1) 5.0e+08 4.3e+08 3.7e+08 3.2e+08 2.8e+08 2.5e+08 2.3e+08 2.0e+08 1.9e+08 1.7e+08 1.6e+08 1.4e+08 1.3e+08 1.2e+08 1.1e+08 1.1e+08 1.0e+08

Signal for an extended source is the number of electrons detected in a pixel fully illuminated by the source. The noise considered here is due to sky brightness, thermal radiation from telescope and instrument optics, dark current inherent in the HgCdTe detector, and Gaussian noise characteristic of the readout electronics. Like the STIS ETC, the controlling executable is compiled from a C program that parses name-value pairs typical of an HTML post. That is, a web GUI interface can create

3


the appropriate input file, and post the information to the ETC executable, which resides in a world-accessible cgi-bin directory. This executable program then calls the STSDAS.SYNPHOT.COUNTRATE task with the appropriate parameters, and calculates the expected signal to noise ratio (SNR). Two calls to COUNTRATE are made by the ETC: one for the target and one for the sky background. COUNTRATE output is the number of source and sky background photons detected per second. These numbers are post-processed by the high level ETC code to account for DQE variations with temperature before signal and noise are estimated. The DQE at the pivot wavelength of the source is used in this post-COUNTRATE processing. Given the size of the pixel-to-pixel variation in the DQE, this approach is considered sufficient. Output from the ETC executable is in the form of HTML, which is displayed on the browser in the normal operating mode. Use of the ETC from the UNIX command line is possible, will be described in a companion ISR (Sivaramakrishnan et al., 2000).

Theory
Signal Sky Dark Thermal Read noise ns nsky nd nt rn electron / pixel / second electrons / pixel / second electrons / pixel / second electrons / pixel / second Double-correlated sampling read noise standard deviation in electrons

The NICMOS focal plane array (FPA) produces one electron per detected photon. For an exposure time of t seconds the target signal is n t [electrons/sec/pixel] and the total

s signal is nt [electrons/sec/pixel], where n = ( n + n s sky + n d + n t ) . Assuming Poisson statistics apply, the standard deviation of the total signal is = deviation of the total noise is

nt . The standard

2 nt + rn , yielding a signal to noise ratio of

ns t S = ------------------------ , 2 nt + rn
where

rn , the standard deviation of the double-correlated sampling read noise, is a

function of the number of non-destructive reads performed. Solving for the exposure time t yields

4


22 2 2 n s t = nt + rn S 2 22 222 nS + n S + 4 n s S rn t = ---------------------------------------------------------------------2 2 ns The thermal optics library
At the time of development of this ETC SYNPHOT does not handle thermal radiation by components of the optical train. Each optic in the train has an area, a temperature and an emissivity. Usually the emissivity is the quantity with the greatest uncertainty. Dust on an optical surface can also be a source of thermal radiation: on HST the dust appears to be well-coupled to the mirror surfaces, and is assumed to be at the same temperature as the optical element itself. Consider an optical train with n elements. Each optic has a throughput factor

0 t i ( ) 1 (reflectance or transmittance), and an emissivity i ( ) , where is the
wavelength of the radiation passing though the relay. If there is any scattering,

t i + i < 1 .
We use a quantity known as the Ètendue, which is an invariant of the optical train, to calculate the thermal contribution of each optic to the detector. Regardless of intermediate magnifications and beam f-ratios, the Ètendue, or the product of the area of optic and the solid angle of the detector as seen by the optic, is preserved. The Ètendue can be calculated by forming the product of the area of the entrance pupil, A, and the solid angle of the pixel on the sky, . The entrance pupil area for NICMOS is the HST primary's visible area as seen from the detector). Thompson (1996) gives the NICMOS 1,2 and 3 camera pixel Ètendues as:

NIC 1 = 1.75 e NIC 2 = 5.29 e NIC 3 = 3.61 e

­ 13 ­ 13 ­ 12

2 m sr 2 m sr 2 m sr

5


The first optic in the train, M1 (the HST primary) contributes

n B ( , T 1 ) [ DQE ] A --------------------- 1 t i h i=2

es

­1

pixel

­1 ­1

counts to each detector pixel at the final image because of it thermal radiation. Similarly, M2, the secondary, adds

n B ( , T 2 ) [ DQE ] A --------------------- 2 t i h i=3

es

­1

pixel

­1 ­1

counts to the thermal background. Each optic down the relay is treated similarly. We present the values of the various optics' emissivities and temperatures used in appendices A1 and A2. As is well-known, HST's NICMOS is not well-aligned with the telescope optics. There are several warm opaque surfaces visible from the NICMOS detectors. The bestcharacterized misalignment is that of camera 2. To calculate the thermal contribution from the primary and its surrounding opaque, warm structure, geometric details of the entrance pupil (Krist et al., 1998), and representative structure temperatures (Bacinski, 1999) were used to calculate the effective contribution from the misaligned HST pupil as seen from NICMOS 2 in the thermal library as the M1 contribution to thermal background. Camera 3 has more warm obstructions in its field of view. These are not treated yet --- the NIC2 pupil description is used for all three cameras at present. Upgrading the values for NIC3 could be a future development task for the ETC. We present the various areas and their temperatures in appendix A1. However, we note that the dominant sources of thermal background are still the large, warm, reflective mirror surfaces of the telescope. In the cases when the optical surface is composed of various areas at different temperatures and emissivities (e.g. the primary `optic' being the circle around the outer edge, with baffles, pads, spiders and cold stops intruding into this area), we weight the corresponding Planck function (evaluated at the temperature of the appropriate pupil component) by the filling factor of the area at that temperature, and its emissivity, to calculate its contribution to the thermal background. The thermal library needs data files with the products of all filter transmissions and detector quantum efficiencies (tabulated as a function of wavelength) in order to calculate the internal optics' thermal contribution. To this end, SYNPHOT's CALCBAND task is used to create all 57 possible `filter times DQE' throughput tables. These tables are read in from a file specified in the call to the public functions offered by this library, and the thermal background for the instrument is returned.

6


World Wide Web Implementation and Usage
The ETC tool can be found on the NICMOS web site at http://www.stsci.edu/instruments/nicmos/topnicmos.html. The GUI will evolve in time as more features are added. Figure 1 through Figure 4 below show the current graphical user interface. Figure 1: ETC Web interface

Section 1: Select a detector and available filter. Section 2: Specify wether you want your observation parameters to be calculated for a given exposure time or a required signal to noise limit.

7


Figure 2: ETC web interface

Section 3: Specify the parameters for the source you wish to observe. · user supplied spectrum: This is accomplished by placing the user input spectrum in an ftp staging area, which the program will load for the simulation. This "staging area" is the anonymous ftp directory:
ftp.stsci.edu:/outside-access/in.coming.

·

Kurucz Model: These model spectra are calculated from the Kurucz database (Dr. R. Kurucz, CD-ROM No. 13, GSFC) which have been installed in the Calibration Database System (CDBS). HST Standard Star Spectra: These spectra are available in CDBS and were chosen from the paper Turnshek, et al., 1990, An Atlas of HST Photometric, Spectrophotometric, and Polarimetric Calibration Objects. Real Object Templates: Observed object spectra that are on-line.

·

·

8


Figure 3: ETC web interface

Section 4: Normalizing the source flux. Whether supplying your own spectra or using one of the supplied model spectra, the source's continuum flux needs to be normalized at some wavelength. This wavelength needs to be within the wavelength range of the input spectrum. The ETC will use it only for normalization and calculate the appropriate flux values for the wavelength range of the observations. If your object is point-like, it can be normalized to a magnitude at a particular Johnson band, or it can be normalized to a flux [in Janskys] value at a given wavelength. If you supply your own spectrum or use one of the HST calibration sources, you can either normalize this spectrum to a fixed value, or you can use the "Do not renormalize" option on the form. For an extended source, you must specify the surface brightness. E(B-V): The flux is normalized after the extinction is taken into account so that it always corresponds to the observed flux. The ETC supports two different extinction laws: -An average Galactic extinction law taken from Seaton (MNRAS, 187, 73p, 1979) -An LMC extinction Law taken from Koorneef & Code (ApJ, 247, 860, 1981).

9


Figure 4: ETC web interface

Section 5: Specify the expected background levels. Currently only one model is provided, regardless of choices made here: a zodiacal light level corresponding to a 45 degree elevation above the ecliptic. In future more choices (low and high zodiacal levels and earthshine) may be added to the calculator. Sections 6 and 7: The detector temperature option was added to allow users to estimate exposure times in the NCS era of NICMOS operations. During end of life, a large rise and decline in the dark current (known as "the bump") was observed over the predicted temperature range of NICMOS operation in cycle 10. Since it is currently unknown whether this event will repeat itself after NCS is installed, the user should choose to include its effect in the exposure time calculation for phase 2 of the proposals. A detector temperature of 75K should be used for Cycle 10 proposals.

10


Figure 5 shows an example output page that is returned to the user. It contains the suggested exposure time, target signal to noise and the chosen observation parameters. Figure 5: Returned output to the user

Acknowledgements
Thanks are due to J. Bacinski, L. Bergeron, T. Boeker, H. Bushouse, C. Hanley, P. Knezek, J. Krist, and R. Shaw from STScI, R. Thompson of Steward Observatory and M. A. Pahre of the Harvard-Smithsonian Center for Astrophysics for detailed and helpful discussions. The observed elliptical galaxy spectrum was provided by M. Rieke. The cool star/brown dwarf spectra were obtained from I. N Reid and B. R. Oppenheimer.

11


References
Boeker, T., Bacinski, J., Bergeron, E., Gilmore, D., Holfeltz, S., Monroe, B. and Sosey, M. NICMOS ISR 001-1999 (revision) Bushouse, H. 1998. SYNPHOT Users' Guide, STScI. Bergeron, E. 1999, private communication Hack,W.,Sahu, K., Kinney, E. and Bohlin, R. 1996, STIS ISR-020 Krist, J. E., Golimowski, D. A., Schroeder, D. J. and Henry, T. J. 1998, PASP 110, 1046-1058 Robberto, M., Sivaramakrishnan, A., Bacinski, J., Calzetti, D., Krist J. E., MacKenty, J. M., Piquero, J., and Stiavelli, M., Simon, B., 2000, SPIE Conference proceedings v4013 (Munich) Simon, B. 1998, STIS ETC C and HTML software Sivaramakrishnan, A., Holfeltz, S. T., Sosey, M. 2000, In preparation Skinner, C. J., NICMOS ISR 14-1996, and associated software Thompson, R. I. 1996 NICMOS Performance Predictor IDL software

A1: HST physical values used
All values are in SI units unless the units are explicitly stated in the variable names.
9.520e-01 3.000e-02 2.400e+00 4.524e+00 288.0 9.520e-01 3.000e-02 7.904e-01 4.907e-01 290.0 3.036e+00 7.600e-01 3.580e+00 256.0 1.000e+00 290.0 1.000e+00 272.6 9.500e-01 253.0 hst->primary_reflectivity hst->primary_emissivity hst->primary_diameter hst->primary_area hst->primary_temperature hst->secondary_reflectivity hst->secondary_emissivity hst->secondary_diameter hst->secondary_area hst->secondary_temperature hst->primary_visible_area hst->secondary_clear_fraction hst->RC_platescale_as_mm hst->spider_temperature hst->spider_emissivity hst->pad_temperature hst->pad_emissivity hst->hole_baffle_temperature hst->hole_baffle_emissivity hst->edge_baffle_temperature

12


9.500e-01

hst->edge_baffle_emissivity

Note that if some of the dust particles are smaller than a wavelength in size, the emissivity and reflectivity do not have to add to unity Currently the following values are used for all three NICMOS cameras, though they were measured for NIC2. Areas are also given as percentages of the primary area. See Krist et al. (1998) for details.
4.520e-02 2.220e-02 4.700e-02 1.603e-01 1.108e+00 hst->nic_hot_pad_area[NIC2] 1.00% hst->nic_hot_spider_area[NIC2] 0.49% hst->nic_hole_baffle_lune[NIC2] 1.04% hst->nic_edge_baffle_lune[NIC2] 3.54% hst->nic_pupil_coldmask_area[NIC2] 24.49%

Varying the temperatures of the primary and secondary, and assuming an emissivity of 3%, we obtain the following thermal background contributions (in electrons per second per pixel) from the longest wavelength filters. These mirror temperatures are in line with the range of measured temperatures of the mirrors. The following three tables show the thermal background count rates in electrons per second in a pixel in two longer wavelength filters of NIC2. Table 2: Primary Temperature 283K
Secondary Temp. F222M F237M 280 6.3 26 285 6.8 28 290 7.5 31

Table 3: Primary Temperature 288K
Secondary Temp. F222M F237M 280 6.9 28 285 7.4 30 290 8.1 33

Table 4: Primary Temperature 293K
Secondary Temp. F222M F237M 280 7.7 31 285 8.2 33 290 8.9 36

13


Table 5: Previous thermal background estimates in electrons/sec. pixel
NICMOS Handbook v.II F222M F237M 8.0 31.0 NICMOS Handbook v.III 9.26 35.6 STScI Recommendation 8.8 34.0

The primary is assumed to have a temperature of 288K, and the secondary 290K in the this ETC. The mirror temperatures are known from the engineering telemetry data from HST. The secondary temperature can vary somewhat (depending on attitude): the ETC attempts to model the telescope's average thermal behavior. The previous STScI-maintained ETC assumed emissivities of 4.8%, and primary and secondary temperatures of 291K and 288.5K respectively, though the calculation of the secondary mirror's thermal contribution underestimated the flux reaching the detector by a factor of a few. We find that both HST mirrors have similar contributions to the thermal background. Recent work (Robberto et al., 2000) indicates that the 3% emissivity figure is within the expected range of mirror emissivity given the history of HST.

A2: NICMOS physical values used
nic->darkcurrent_bump nic->detector_temperature nic->foreoptics_temperature nic->bend1_mirror_reflectivity nic->reimaging_mirror_reflectivity nic->pupil_mirror_reflectivity nic->image_mirror_reflectivity nic->paraboloid1_mirror_reflectivity nic->paraboloid2_mirror_reflectivity nic->bend2_mirror_reflectivity nic->dewar_window_transmission nic->pixel_size_um nic->pixel_size_as[NIC1] nic->pixel_size_as[NIC2] nic->pixel_size_as[NIC3] = = = = = = = = = = = = = = = BUMP_PRESENT; 70.0; 270.0; 0.985; 0.985; 0.985; 0.985; 0.985; 0.985; 0.985; 0.93; 40.0; 0.043; 0.075; 0.20;

Relative DQE is measured at two wavelengths, lrefmu[0] and detector, at various temperatures. A quadratic fit (in temperature) private communication) produces the three coefficients a[0, 1, 2]. pivot wavelength of the source/filter under consideration is found at the particular detector temperature being used. For more detail
/* at lrefmu[0] microns: */ measured_rdqe[0] = a0[0] + /* at lrefmu[1] microns: */ measured_rdqe[1] = a0[1] + nic->rel_dqe_l_mu[NIC1][0] a1[0] * T a1[1] * T + +

lrefmu[1], for each to these data (Bergeron, The relative DQE at the by linear extrapolation see Boeker et al., 1999.

a2[0] * T * T; a2[1] * T * T;

= 1.1;

14


nic->rel_dqe_l_mu[NIC1][1] nic->rel_dqe_a0[NIC1][0] nic->rel_dqe_a1[NIC1][0] nic->rel_dqe_a2[NIC1][0] nic->rel_dqe_a0[NIC1][1] nic->rel_dqe_a1[NIC1][1] nic->rel_dqe_a2[NIC1][1] nic->rel_dqe_l_mu[NIC2][0] nic->rel_dqe_l_mu[NIC2][1] nic->rel_dqe_a0[NIC2][0] nic->rel_dqe_a1[NIC2][0] nic->rel_dqe_a2[NIC2][0] nic->rel_dqe_a0[NIC2][1] nic->rel_dqe_a1[NIC2][1] nic->rel_dqe_a2[NIC2][1] nic->rel_dqe_l_mu[NIC3][0] nic->rel_dqe_l_mu[NIC3][1] nic->rel_dqe_a0[NIC3][0] nic->rel_dqe_a1[NIC3][0] nic->rel_dqe_a2[NIC3][0] nic->rel_dqe_a0[NIC3][1] nic->rel_dqe_a1[NIC3][1] nic->rel_dqe_a2[NIC3][1] nic->aomega[NIC1] nic->aomega[NIC2] nic->aomega[NIC3]

= = = = = = = = = = = = = = = = = = = = = = =

1.6; -4.1412421; 0.11773110; -0.00057135511; -3.4798412; 0.10636465; -0.00055803535; 1.1; 1.6; -3.1131409; 0.093230650; -0.00044176987; -2.5612221;; 0.083302061; -0.00042354937; 1.1; 2.2; -5.1221150; 0.14595018; -0.00077193987; -1.8900719; 0.072383888; -0.00041987179;

= 1.75e-09 * 1.0e-4; = 5.29e-09 * 1.0e-4; = 3.61e-08 * 1.0e-4;

Our Ètendue include opaque obstructions within the outer perimeter of the primary. Dark currents are recalculated on the fly using NICMOS end of life warmup data. See Boeker et al. 1999 for details.

A3: Real object templates
Elliptical -- unpublished elliptical galaxy spectrum, M. Rieke Gl 229B -- Cool brown dwarf spectrum, Oppenheimer et al., 1998, ApJ 502, 932 GD 165B -- Brown dwarf spectrum, Kirkpatrick et al., 199N, ApJ 519, 834 Gl 752B -- M8 spectrum, H. Jones Gl 406 -- M6 spectrum, H. Jones Gl 411 -- M2 spectrum, H. Jones Other objects will be offered in the future.

A4: The high level routine imgston() pseudocode
Get the get set set detector name from input filtername camera number camera-dependent 'qetfactor_key' (for temp dependent QE multiplicative factor) set camera-dependent 'dark_key'

15


(for temp-dependent dark current) Get Get Get Get Get Get the point source / extended source switch value from input read noise for double-correlated sampling from 'mode file' pixel scale from 'mode' file pivot wavelength of source from synphot output detector temperature from input dark current bump key (yes/no switch) from input

Calculate sky background from synphot output on background and pixel scale ** * * * * * * Calculate thermal background from optics: initialize the NICMOS and HST objects set_nic_physical(...detector temperature) set_nic_physical(...presence/absence of dark current bump) create directory name where dqe*filter files exist from #defines DATADIR and QEFILTSUBDIR (in config.h) obtain counts per sec in a pixel from thermal radiation from HST and NICMOS optics get_nic_physical(...relative dqe at pivot wavelength, detector T) scale scale scale scale synphot source counts synphot background counts sky counts thermal background counts (*= (*= (*= (*= rdqe) rdqe) rdqe) rdqe)

For point source request: obtain camera- and pivot-wavelength-dependent aperture correction (1", 1", 2" dia apertures) scale source counts by aperture correction (for 'counts in std aperture' info only) obtain camera- and filter-dependent central-pixel-fraction For extended source request: calculate target area from input target diameter calculate counts per pixel from synphot source counts, target area and pixel scale Calculate SNR or Exposure time (as requested in input)

A5: Filter Summary
The following tables show the three cameras' filters' properties. The definitions of the all columns except the last (Fraction) are those of the STSDAS.SYNPHOT.BANDPAR task. These definitions are summarized here. For further details please see the on-line help on BANDPAR. Here `INT' refers to an integral over wavelength, and `THRU' is the throughput of the filter at that wavelength. `LAM' is wavelength. Central Wavelength = bandpar.pivwv defined as PIVWV = SQRT(INT(THRU * LAM) / INT (THRU / LAM)) Mean Wavelength = bandpar.avgwv defined as AVGWV = INT(THRU * LAM) / INT(THRU) Peak Wavelength = bandpar.wpeak

16


defined as wavelength at peak throughput FWHM = bandpar.fwhm defined as FWHM = SQRT(8 * LOG(2)) * BANDW Max Transmission = maximum transmission OF THE FILTER ONLY (no other optical components) as measured pre-launch in the lab Max Throughput = bandpar.tpeak defined as peak bandpass throughput The right most column, Fraction, is the fraction of light falling on the central pixel (assuming the point source PSF is centered on that pixel as estimated by Tiny Tim PSF simulations. All data in these tables (except for the PSF-related `Fraction') are generated from the STScI CDBS tables, which are updated as more accurate information is obtained. The observer is strongly advised to regenerate these tables from the most up-to-date data in CDBS if the exact values are critical to planning observations. The SYNPHOT user's guide (Bushouse, 1998) is the best guide on how to do this. Table 6: NIC 1 Filters
Filter F090M F095N F097N F108N F110M F110W F113N F140W F145M F160W F164N F165M F166N F170M F187N F190N Central 0.9041 0.9537 0.9716 1.0816 1.1026 1.1292 1.1298 1.4399 1.4557 1.6071 1.6460 1.6489 1.6607 1.7067 1.8748 1.8986 Mean 0.9058 0.9537 0.9716 1.0816 1.1042 1.1412 1.1298 1.4593 1.4569 1.6114 1.6461 1.6500 1.6607 1.7078 1.8748 1.8986 Peak 0.9176 0.9564 0.9739 1.0790 1.1294 1.3430 1.1312 1.7775 1.5115 1.7730 1.6476 1.7122 1.6624 1.7888 1.8756 1.8942 FWHM 0.1318 0.0088 0.0104 0.0090 0.1390 0.3894 0.0111 0.5592 0.1384 0.2755 0.0142 0.1396 0.0145 0.1410 0.0158 0.0169 Range 0.8 - 1.0 1% 1% 1% 1.0-1.2 0.8-1.35 1% 0.8-1.8 1.35-1.55 1.35-1.75 1% 1.55-1.75 1% 1.6-1.8 1% 1% Trans 79.64 66.31 73.51 81.94 90.49 95.11 86.24 95.50 94.03 96.37 93.41 94.78 90.12 95.55 88.93 93.05 Thru 4.6174 3.8254 4.5612 5.6077 7.2642 9.6072 6.6160 16.0600 12.0850 16.1670 13.5890 15.6350 13.8580 16.3380 16.2170 16.6510 Fraction 0.150 0.136 0.131 0.111 0.108 0.113 0.102 0.0766 0.0640 0.0562 0.0534 0.0521 0.0505 0.0481 0.0411 0.0401

17


Table 6: NIC 1 Filters
Filter POL0S Central 1.0602 Mean 1.0684 Peak 1.1435 FWHM 0.3102 Range 0.8-1.3 Trans 77.60 Thru 3.4483 Fraction 0.048

Table 7: NIC 2 Filters
Filter F110W F160W F165M F171M F180M F187N F187W F190N F204M F205W F207M F212N F215N F216N F222M F237M POL0L Central 1.1285 1.6060 1.6516 1.7212 1.7971 1.8740 1.8718 1.9003 2.0355 2.0714 2.0824 2.1213 2.1487 2.1641 2.2182 2.3696 2.0007 Mean 1.1402 1.6103 1.6527 1.7214 1.7973 1.8740 1.8732 1.9004 2.0357 2.0795 2.0829 2.1213 2.1487 2.1641 2.2188 2.3701 2.0017 Peak 1.3402 1.7722 1.7136 1.7259 1.8131 1.8746 1.8906 1.9005 2.0634 2.3171 2.0683 2.1228 2.1562 2.1668 2.1866 2.4338 2.0750 FWHM 0.3844 0.2772 0.1395 0.0584 0.0563 0.0159 0.1707 0.0147 0.0799 0.4314 0.1069 0.0206 0.0187 0.0186 0.1205 0.1115 0.1490 Range 0.8-1.4 1.4-1.8 1.55-1.75 1.68-1.75 1.765-1.835 1% 1.75-2.0 1% 1.99-2.09 1.75-2.35 2.0-2.15 1% 1% 1% 2.15-2.3 2.3-2.45 1.9-2.1 Trans 94.90 96.59 95.36 95.15 93.40 88.91 87.97 93.22 91.77 98.14 96.40 90.90 85.91 90.07 89.90 92.38 96.67 Thru 11.7710 18.5650 18.4150 17.7640 18.4280 18.3620 18.0140 19.2570 21.9010 29.5350 20.7120 21.4380 21.0440 21.8620 22.2510 28.6910 12.0740 Fraction 0.288 0.159 0.149 0.138 0.128 0.119 0.117 0.116 0.104 0.107 0.0976 0.0953 0.0933 0.0915 0.0881 0.0797 0.33

Table 8: NIC 3 Filters
Filter F108N F110W F113N F150W F160W Central 1.0800 1.1264 1.1284 1.5484 1.6078 Mean 1.0800 1.1388 1.1284 1.5658 1.6120 Peak 1.0777 1.3402 1.1315 1.7685 1.7730 FWHM 0.0112 0.3964 0.0106 0.5522 0.2761 Range 1% 0.8-1.4 1% 1.1-1.9 1.4-1.8 Trans 75.82 94.90 84.81 97.67 96.59 Thru 6.1366 10.3110 7.3788 18.6900 17.8900 Fraction 0.582 0.590 0.579 0.534 0.524

18


Table 8: NIC 3 Filters
Filter F164N F166N F175W F187N F190N F196N F200N F212N F215N F222M F240M Central 1.6460 1.6583 1.8360 1.8748 1.9004 1.9639 1.9975 2.1213 2.1487 2.2184 2.3969 Mean 1.6461 1.6583 1.8635 1.8748 1.9004 1.9639 1.9975 2.1213 2.1487 2.2190 2.3976 Peak 1.6476 1.6602 2.2843 1.8756 1.9005 1.9697 2.0022 2.1228 2.1562 2.1866 2.4162 FWHM 0.0142 0.0139 0.7605 0.0158 0.0147 0.0177 0.0196 0.0206 0.0187 0.1204 0.1384 Range 1% 1% 1.2-2.3 1% 1% 1% 1% 1% 1% 2.15-2.3 2.3-2.5 Trans 86.27 87.24 96.58 88.91 93.22 93.74 91.75 90.90 85.91 89.9 92.43 Thru 14.9580 15.1480 28.7950 17.8240 19.2900 20.2760 19.7980 21.1640 20.7940 22.0990 28.5600 Fraction 0.523 0.513 0.486 0.471 0.466 0.450 0.445 0.418 0.413 0.397 0.363

19