TIPS, Feb 15, Tomas Dahlen1 Update on the NICMOS flat –field monitoring Tomas Dahlen & NICMOS team
TIPS, Feb 15, Tomas Dahlen2 Update on the NICMOS flat –field monitoring The current pipe-line flats are the “Post NCS On-Orbit Flat-fields” from 2002 Flat-fields have been monitored every Post NCS cycle - most frequently in F110W, F160W and F222M - at least some flats have been taken in all available filters/grisms Q: have the flat-fields changed over time since 2002 in a way that would require an updated set to be constructed/distributed? To investigate this we create 1) “Single-date-flats” from each observing run 2) “Epoch-flats” from combining flats over time-spans ~1 yr
TIPS, Feb 15, Tomas Dahlen3 Update on the NICMOS flat –field monitoring Dates when flats have been taken ( ) Red=cam 1 Green=cam 2 Blue=cam 3 Wave-length
TIPS, Feb 15, Tomas Dahlen4 Update on the NICMOS flat –field monitoring Changes in flat-fields: 1) Overall DQE changes 2) Spatial changes in normalized flats
TIPS, Feb 15, Tomas Dahlen5 Update on the NICMOS flat –field monitoring Overall DQE changes Figures show change in counts/s (with mean normalized to unity). Total change is typically a decrease by 1-2%
TIPS, Feb 15, Tomas Dahlen6 Update on the NICMOS flat –field monitoring Could the decrease in DQE be due to change in detector temperature? Decrease in detector temperature would lower DQE: ~2% per K at F110 ~1.5% per K at F160 ~1% per K at F222 Post NCS T=77K Pre NCS T=62K Gold points: Mounting cup temperature Rest of points: Temp-from-bias (Bergeron 05) ( Temperature variationsPre and Post NCS DQE
TIPS, Feb 15, Tomas Dahlen7 Update on the NICMOS flat –field monitoring Overall DQE changes CONCLUSION: Direction of DQE change is consistent with decrease in temp, but seems smaller than suggested by the change of temp-from-bias. If mounting cup temp reflects detector temp, then another explanation is needed to explain the decrease in DQE Gray lines: predictions from temperature dependence of DQE
TIPS, Feb 15, Tomas Dahlen8 Update on the NICMOS flat –field monitoring DQE spatial changes As a measurement of spatial changes we define a “spatial ratio”: R(sp)=mean(abs[ratio-image – 1]), where: ratio-image=date-flat/pipe-line flat. Strongest variations at short wave-lengths F110W: increase from to F160W: increase from to Figures show change in R(sp) with time
TIPS, Feb 15, Tomas Dahlen9 Update on the NICMOS flat –field monitoring Ratio images for cam 1+F110W R(sp) increases from (2002) to (2007) (Stretch = +/- 2%)
TIPS, Feb 15, Tomas Dahlen10 Update on the NICMOS flat –field monitoring Ratio images for cam 2+F110W R(sp) increases from (2002) to (2005)
TIPS, Feb 15, Tomas Dahlen11 Update on the NICMOS flat –field monitoring Ratio images for cam 3+F110W R(sp) increases from (2002) to (2007)
TIPS, Feb 15, Tomas Dahlen12 Update on the NICMOS flat –field monitoring What is the photometric uncertainty due to the spatial variations? To investigate this, we measure the ratios between the new flats and the pipe-line flats at random positions over the CCD. Figure plots the distribution for F110W. Cam1+F110W (2007) 34% probability of error >0.01 mag 1% probability of error >0.02 mag Cam 2+F1110W (2005) 3% probability of error >0.01 mag 0% probability of error >0.02 mag Cam3+F110W (2007) 14 % probability that error >0.01 mag 2 % probability that error >0.02 mag Cam 1 Cam 2 Cam 3
TIPS, Feb 15, Tomas Dahlen13 Update on the NICMOS flat –field monitoring Combining flats 1)Flats are divided into 5 epoch 2) Each flat has a ‘use after’ date 3) And an associated temperature 4) Typical S/N = IMAGE S/N t(bias) t(cup) N(image) date cam2_F160W_pipeline 596 8? cam2_F160W_ep ~2002 cam2_F160W_ep ~2003 cam2_F160W_ep ~2004 cam2_F160W_ep ~2005 cam2_F160W_ep ~2006 Example: cam1 + F160W
TIPS, Feb 15, Tomas Dahlen14 Update on the NICMOS flat –field monitoring Conclusions (prel.) DQE has decreased 1-2% since 2002 Spatial changes up to 1% In total: ~0.03 mag photometric error (“worst case” F110W) “Epoch” flats soon ready for delivery