7.8 IR Sensitivity

7.8.1 Limiting Magnitudes

Table 7.10 presents the predicted limiting-magnitude performance of the WFC3 IR channel and compares it with that of the camera 3 on NICMOS (NIC3). The calculations are based on an optimal extraction of a point source. The limiting ABMAG at an SNR of 10 was calculated for a 1-hour and a 10-hour exposure. The throughput curves for the WFC3 filters listed in column 2 were used; for NIC3, the most similar wide-band filter was used, and its name is given in column 3.

An online Exposure Time Calculator (ETC) is available (see Section 9.2).

Table 7.10: Limiting-magnitude performance of WFC3 compared with that of the NICMOS NIC3, based on on-orbit sensitivity from SMOV. 

7.8.2 Sensitivity

The WFC3/IR detector exhibits a low level sensitivity loss at rates of ~0.1% per year, with greater losses at shorter wavelengths (WFC3 ISR 2024-06). The total sensitivity loss since WFC3 installation in 2009 is ~ 1-2%, comparable to the IR detector repeatability (WFC3 ISR 2024-01; WFC3 ISR 2021-05; WFC3 ISR 2020-10; WFC3 ISR 2019-07).

To account for the time-dependent sensitivity of the IR detector, inverse sensitivities for all 15 WFC3/IR filters were updated in December 2024 (Calamida et al., in prep). These new time-dependent inverse sensitivities were derived by using ~ 14 years of observations of five CALSPEC standards, corrected for the recommended loss rates from WFC3 ISR 2024-06 (see Table 7.11). In addition, a new version of the calibration pipeline, calwf3 v3.7.2, implements the time-dependent flux calibration and populates the image file header with time-dependent photometric keywords. The new inverse sensitivities provide an internal photometric precision better than 0.5% for all wide--, medium--, and narrow-band filters.

The new inverse sensitivities can now be computed for specific observation dates by using the Python package stsynphot (Calamida et al., in prep).

Table 7.11: Sensitivity loss rates and pivot wavelengths for F098M and all five wide-band filters. Remaining medium-band and all narrow-band filters should use the nearest wavelength solution. 

Monitoring of the IR channel sensitivity is carried out via several ongoing calibration programs. Staring mode observations of stellar clusters over a ~14-year baseline suggests small sensitivity loss rates of ~0.1% per year (WFC3 ISR 2024-06; WFC3 ISR 2022-07). Studies of WFC3/IR grism observations of CALSPEC standards also show small declines in sensitivity, about 0.1-0.3% per year, depending on the length of the observation baseline and the extraction wavelength range (WFC3 ISR 2024-06; WFC3 ISR 2024-01; Bohlin and Deustua, 2019). WFC3/IR scan observations of stars in the open cluster M35 over ~ 7 years suggest losses of about 0.16% per year in F098M and 0.06% per year in F140W. Recent analysis of 13 cycles of WFC3/IR internal flat field calibration programs indicated greater count rate losses at bluer wavelengths, with an average of ~0.3% per year averaged across all filters, suggesting a combination of effects from both the changing overall sensitivity as well as previously observed lamp reddening (WFC3 ISR 2024-10).

In contrast to the UVIS detector, staring mode monitoring of CALSPEC standards in the IR filters have large systematic uncertainties which limit their 1-sigma photometric repeatability to +/- 1.0% (WFC3 ISR 2024-06), and therefore cannot be used for accurately measuring the sensitivity loss rates, despite their significant time baseline (~14 years). However, these data can be (and were) used to test the appropriateness of calculated sensitivity loss rates, such as those previously listed; this approach was leveraged in WFC3 ISR 2024-06 in order to optimize sensitivity loss rates for each WFC3/IR filter. 

For a more detailed discussion of WFC3/IR sources of errors, see Section 7.11 and Section 9.1 of the WFC3 Data Handbook.