Index: /trunk/doc/release.2015/systematics.20140411/Makefile =================================================================== --- /trunk/doc/release.2015/systematics.20140411/Makefile (revision 40305) +++ /trunk/doc/release.2015/systematics.20140411/Makefile (revision 40306) @@ -21,5 +21,6 @@ pics/radial_p1_r.pdf \ pics/radial_p2_r.pdf \ -pics/radial_p3_r.pdf +pics/radial_p3_r.pdf \ +pics/filter_trends.pdf OLD_PDFPICS = \ Index: /trunk/doc/release.2015/systematics.20140411/diffusion.tex =================================================================== --- /trunk/doc/release.2015/systematics.20140411/diffusion.tex (revision 40305) +++ /trunk/doc/release.2015/systematics.20140411/diffusion.tex (revision 40306) @@ -20,5 +20,6 @@ \definecolor{light-gray}{gray}{0.50} -\newcommand\oldtext[1]{\textbf{\color{light-gray}#1}} +% \newcommand\oldtext[1]{\textbf{\color{light-gray}#1}} +\newcommand\oldtext[1]{\ignorespaces} \newcommand\newtext[1]{\textbf{\color{blue}#1}} \newcommand\fixtext[1]{\textbf{\color{red}#1}} @@ -376,22 +377,22 @@ \label{sec:tree.rings} -\begin{table} -% \tiny -\begin{center} -\caption{Systematic Trends : Standard deviation by filter\label{table:sigmas.by.filter}} -\begin{tabular}{|l|rrrrr|} -\hline -{\bf Filter} & {\bf psf mags} & {\bf ap mags} & {\bf astrom} & {\bf smear} & {\bf flat} \\ - & mmags & mmags & mas & pixels$^2$ & mmags \\ -\hline -\gps & 11.8 & 13 & 8.0 & 0.169 & 3.0 \\ -\rps & 10.9 & 12 & 6.7 & 0.133 & 2.2 \\ -\ips & 8.5 & 10 & 6.0 & 0.069 & 1.7 \\ -\zps & 8.7 & 12 & 5.5 & 0.052 & 3.2 \\ -\yps & 16.5 & 26 & 6.8 & 0.059 & 15.3 \\ -\hline -\end{tabular} -\end{center} -\end{table} +%% \begin{table} +%% % \tiny +%% \begin{center} +%% \caption{Systematic Trends : Standard deviation by filter\label{table:sigmas.by.filter}} +%% \begin{tabular}{|l|rrrrr|} +%% \hline +%% {\bf Filter} & {\bf psf mags} & {\bf ap mags} & {\bf astrom} & {\bf smear} & {\bf flat} \\ +%% & mmags & mmags & mas & pixels$^2$ & mmags \\ +%% \hline +%% \gps & 11.8 & 13 & 8.0 & 0.169 & 3.0 \\ +%% \rps & 10.9 & 12 & 6.7 & 0.133 & 2.2 \\ +%% \ips & 8.5 & 10 & 6.0 & 0.069 & 1.7 \\ +%% \zps & 8.7 & 12 & 5.5 & 0.052 & 3.2 \\ +%% \yps & 16.5 & 26 & 6.8 & 0.059 & 15.3 \\ +%% \hline +%% \end{tabular} +%% \end{center} +%% \end{table} For many of the GPC1 OTA CCDs, we observe a spatial pattern in the @@ -422,9 +423,9 @@ illustrate the effects in detail, but a similar set of effects are seen in many, if not all, of the GPC1 detectors with varying -strengths. \fixtext{First, we show the residual PSF photometry. Second, we +strengths. First, we show the residual PSF photometry. Second, we show the residual aperture photometry. Third, we show the astrometric residual patterns. Fourth, we show the patterns observed in the flat-field images. Finally, we show measurements derived from the -second-moments of the stars.} +second-moments of the stars. For all effects discussed below, we are measuring the mean value of @@ -458,10 +459,12 @@ \hspace{\jumpleft} \parbox[b]{\capwidth}{ -\caption{PSF Magnitude residuals by filter (\grizy). White boxes are +\caption{PSF Magnitude residuals by filter (\grizy) for a single + example GPC1 device (XY40). White boxes are GPC1 cells which have been masked due to poor response. Superpixels representing regions of $10\times10$ pixels are used to determine the median deviation for measurements at the given chip pixel location compared with the average photometry for the given - object.} \label{fig:psfmags.by.filter}} + object. Fringing dominates the \yps-band signal, saturating the + color scale to black or white in areas.} \label{fig:psfmags.by.filter}} \end{center} \end{figure*} @@ -473,10 +476,12 @@ \hspace{\jumpleft} \parbox[b]{\capwidth}{ -\caption{Aperture Magnitude residuals by filter (\grizy). White boxes +\caption{Aperture Magnitude residuals by filter (\grizy) for a single + example GPC1 device (XY40). White boxes are GPC1 cells which have been masked due to poor response. Superpixels representing regions of $10\times10$ pixels are used to determine the median deviation for measurements at the given chip pixel location compared with the average photometry for the given - object. } \label{fig:apmags.by.filter}} + object. Fringing dominates the \yps-band signal, saturating the + color scale to black or white in areas.} \label{fig:apmags.by.filter}} \end{center} \end{figure*} @@ -495,10 +500,10 @@ The tree-ring pattern is clearly visible for the four blue filters, -but finging dominates the pattern for \yps. Small offsets of +but fringing dominates the pattern for \yps. Small offsets of individual cells are also apparent for \zps. While the patterns are clear across the image, the signal-to-noise of the structures per -pixel is not very strong in these images. The per-pixel standard +pixel is not very strong in these images. \oldtext{The per-pixel standard deviations of these plots are listed in -Table~\ref{table:sigmas.by.filter}. The per-pixel standard deviation +Table~1.} The \oldtext{per-pixel} standard deviation \newtext{of the pixel values in the images (a measure of the noise in the absence of any systematic signal)} is comparable to the amplitude of the correlated structures, so we need to integrate along the radial structures to make stronger @@ -506,11 +511,11 @@ Figure~\ref{fig:apmags.by.filter} shows the equivalent measurement for -aperture photometry instead of PSF photometry. The finging -pattern again dominates the plot for \yps, but the tree rings are not -seen in any of the filters. A diagonal pattern is visible in \gps\ -which is not observed in the PSF magnitudes. While the per-pixel -scatter is somewhat (10\% to 20\%) higher for these aperture -magnitudes than for the PSF magnitudes -(Table~\ref{table:sigmas.by.filter}), a structure with the amplitude +aperture photometry instead of PSF photometry. The fringing pattern +again dominates the plot for \yps, but the tree rings are not seen in +any of the filters. A diagonal pattern is visible in \gps\ which is +not observed in the PSF magnitudes. While the \newtext{standard + deviation of the pixel values} \oldtext{per-pixel scatter} is +somewhat (10\% to 20\%) higher for these aperture magnitudes than for +the PSF magnitudes\oldtext{ (Table~1)}, a structure with the amplitude of the PSF magnitude tree-rings would certainly have been obvious. @@ -541,9 +546,9 @@ %% \end{figure*} -\oldtext{Figure~3} \newtext{Figure~\ref{fig:all.effects.rband} (middle-left)} -shows a similar type of measurement -for astrometric residuals. To generate this plot, we use the same -selection of measurements for astrometry as for photometry. In this -case, we extract the residual in both the RA and DEC directions +\oldtext{Figure~3} \newtext{Figure~\ref{fig:all.effects.rband} + (middle-left)} shows a similar type of measurement for astrometric +residuals \newtext{in \rps-band}. To generate this plot, we use the +same selection of measurements for astrometry as for photometry. In +this case, we extract the residual in both the RA and DEC directions ($\delta RA = \overline{RA} - RA_i$, $\delta DEC = \overline{DEC} - DEC_i$) and rotate these values to the chip coordinate system ($\delta @@ -557,18 +562,36 @@ offsets into $\delta R,\delta \theta$ measurements ($\delta R$ : radial component away from the center of the pattern, $\delta \theta$ -: tangential component). - -\oldtext{Figure~\ref{fig:astrom.by.filter} shows the 2D patterns of $\delta R$ -for each filter (\grizy). The dynamic range of the color scale is -from -20 to +20 milliarcseconds for all 5 plots.} A tree-ring -pattern is visible for all five filters, with systematic structures -following a circular pattern centered on the chip corner; the finging -pattern is not apparent in the \yps\ astrometry. \oldtext{The per-pixel -standard deviations of these plots are listed in -Table~\ref{table:sigmas.by.filter}.} The signal-to-noise of these -structures is again somewhat weak, but the pattern is clearly visible -in these figures. +: tangential component). \newtext{The dynamic range of the color scale + is from -20 to +20 milliarcseconds for this plot.} + +\oldtext{Figure~\ref{fig:astrom.by.filter} shows the 2D patterns of + $\delta R$ for each filter (\grizy). The dynamic range of the color + scale is from -20 to +20 milliarcseconds for all 5 plots.} A +tree-ring pattern is visible for all five filters, with systematic +structures following a circular pattern centered on the chip corner; +the fringing pattern is not apparent in the \yps\ astrometry. +\oldtext{The per-pixel standard deviations of these plots are listed + in Table~1.} The signal-to-noise of these structures is again +somewhat weak, but the pattern is clearly visible in \oldtext{these figures} \newtext{Figure~\ref{fig:all.effects.rband} (middle-left)}. \subsection{Flat-field Structures} + +% All Effects in r-band +\begin{figure*}[htbp] +\begin{center} +\parbox[b]{\figwidth}{\includegraphics[width=5.0in]{\picdir/all_effects_r.\plotext}} +\caption{All 6 measured effects for \rps for a single + example GPC1 device (XY40). This figure illustrates the + different spatial structure observed for each of the 6 patterns + measured in this work. The PSF magnitude (upper-left) and smear + residuals (lower-left) have a very clear common tree-ring structure, + while the astrometric residual (middle-left) and flat-field + residuals (middle-right) have their own common tree-ring pattern with + much higher frequencies than the previous two effects. Aperture + magnitude (upper-right) and shear residuals (lower-right) do not + show a strong signal consistent with either of the two patterns.} +\label{fig:all.effects.rband} +\end{center} +\end{figure*} % flat-field residual @@ -590,5 +613,5 @@ \oldtext{Figure~4} \newtext{Figure~\ref{fig:all.effects.rband} (middle-right)} shows the high-spatial-frequency -structures in the flat-field images. For this measurement, we have +structures in the \newtext{\rps-band} flat-field\oldtext{ images}. For this measurement, we have used a set of monochromatic flat-field images obtained with a tunable laser. The laser is used to illuminate our flat-field screen which is @@ -607,18 +630,18 @@ pixels associated with each superpixel. -\fixtext{Figure~\ref{fig:flats.by.filter} shows the superpixel images for the -flat-fields in the five filters.} These flat-field images are -displayed as fractional deviations relative to the median flat-field +\oldtext{Figure~\ref{fig:flats.by.filter} shows the superpixel images for the +flat-fields in the five filters. These flat-field images are} \newtext{The flat-field image is} +displayed as fractional deviations relative to the median of the flat-field image and can thus be compared to the magnitude residuals. When -flattening an image, these flat-fields would be divided into the flux +flattening an image, \oldtext{these flat-fields} \newtext{the flat-field image} would be divided into the flux of the raw image. The residuals are thus defined in the sense that a -positive feature in these flats which does {\em not} represent a real +positive feature in \oldtext{these flats} \newtext{the flat} which does {\em not} represent a real quantum efficiency deviation would induce a {\em reduction} in the measured flux in those pixels, and thus a {\em negative} deviation in $\delta m_{psf}$ as defined above. The dynamic range of the color -scale in these plots is -0.01 to +0.01. The tree-ring pattern is +scale in \oldtext{these plots} \newtext{this plot} is -0.01 to +0.01. The tree-ring pattern is strong in the (\gps,\rps,\ips) images, but nearly swamped by fringing -in \zps, and completely lost to finging in \yps. A diagonal banding -pattern is also seen in \gps: this feature is thought to be due to +in \zps, and completely lost to fringing in \yps. A diagonal banding +pattern is also seen in \gps and \rps: this feature is thought to be due to the lithography process used to generate the CCD. A blob can also been seen covering 4 cells near the center of this chip; this is @@ -633,5 +656,9 @@ analysis of the tree rings, we high-pass filter the superpixel image by subtracting a copy smoothed with a Gaussian of $\sigma = 3.0$ -superpixels. +superpixels. \newtext{This smoothing kernel is large enough compared + to the tree ring structures that they are not suppressed + significantly. Without this smoothing, features from the diagonal + banding pattern remain in the \rps-band image and contaminate the + tree-ring signal.} \subsection{Second Moments} @@ -711,5 +738,5 @@ \oldtext{Figure~5} \newtext{Figure~\ref{fig:all.effects.rband} (lower-left)} shows the spatial trend of the smear, -$e_0$. The dynamic range of these images is -0.3 to +0.3 pixel$^2$. A +$e_0$. The dynamic range of \oldtext{these images} \newtext{this image} is -0.3 to +0.3 pixel$^2$. A tree-ring pattern is visible for all 5 filters, though \yps\ is dominated by the fringing pattern. Structures with relatively low @@ -724,44 +751,38 @@ ellipse orientation as a function of postion. The length of the vectors corresponds to the value of $e_2$. The tree-ring structure is -{\em not} apparent in this figure for any filter. The spatial +{\em not} apparent in \oldtext{this figure} \newtext{the shear} for any filter. The spatial variations are low-frequency and unrelated to the radial trend from the upper-left corner. - -\subsection{Correlations Between Tree-Ring Patterns} % All Effects in r-band \begin{figure*}[htbp] \begin{center} -\parbox[b]{\figwidth}{\includegraphics[width=5.0in]{\picdir/all_effects_r.\plotext}} -\caption{All 6 measured effects for \rps. This figure illustrates the - different spatial structure observed for each of the 6 patterns - measured in this work. The PSF magnitude (upper-left) and smear - residuals (lower-left) have a very clear common tree-ring structure, - while the astrometric residual (middle-left) and flat-field - residuals (middle-right) have their own common tree-ring pattern with - much higher frequencies than the previous two effects. Aperture - magnitude (upper-right) and shear residuals (lower-right) do not - show a strong signal consistent with either of the two patterns.} -\label{fig:all.effects.rband} +\parbox[b]{\figwidth}{\includegraphics[width=5.0in]{\picdir/filter_trends.\plotext}} +\caption{Amplitude of the 4 effects which follow the tree-rings as a + function of filter, relative to the amplitude in the \gps-band.} +\label{fig:filter.trend} \end{center} \end{figure*} -\begin{table} -% \tiny -\begin{center} -\caption{Systematic Trends : Correlations by filter\label{table:correlation.by.filter}} -\begin{tabular}{|l|rrrr|} -\hline -{\bf Filter} & {\bf smear} & {\bf psf mags} & {\bf astrom} & {\bf flat} \\ -\hline -\gps & 1.00 & 1.00 & 1.00 & 1.00 \\ -\rps & 0.78 & 0.84 & 0.84 & 0.76 \\ -\ips & 0.40 & 0.50 & 0.66 & 0.64 \\ -\zps & 0.16 & 0.26 & 0.37 & 0.33 \\ -\yps & 0.10 & 0.10 & 0.25 & 0.30 \\ -\hline -\end{tabular} -\end{center} -\end{table} +\subsection{Correlations Between Tree-Ring Patterns} + +%% \begin{table} +%% % \tiny +%% \begin{center} +%% \caption{\newtext{Amplitude of the four systematic trends in each filter +%% relative to \gps.} \oldtext{Systematic Trends : Correlations by filter}\label{table:correlation.by.filter}} +%% \begin{tabular}{|l|rrrr|} +%% \hline +%% {\bf Filter} & {\bf smear} & {\bf psf mags} & {\bf astrom} & {\bf flat} \\ +%% \hline +%% \gps & 1.00 & 1.00 & 1.00 & 1.00 \\ +%% \rps & 0.78 & 0.84 & 0.84 & 0.76 \\ +%% \ips & 0.40 & 0.50 & 0.66 & 0.64 \\ +%% \zps & 0.16 & 0.26 & 0.37 & 0.33 \\ +%% \yps & 0.10 & 0.10 & 0.25 & 0.30 \\ +%% \hline +%% \end{tabular} +%% \end{center} +%% \end{table} Tree-ring patterns are clearly seen in 4 of the measurement types @@ -807,6 +828,6 @@ For all four types of measurements, the \oldtext{slope of the fitted - lines} \newtext{amplitudes relative to \gps} are listed in -Table~\ref{table:correlation.by.filter}. There is a consistency in + lines} \newtext{amplitudes relative to \gps} are \oldtext{listed in +Table~2} \newtext{plotted in Figure~\ref{fig:filter.trend}}. There is a consistency in the trend from \gps, with the strongest systematic tree-ring effects to \yps, with the weakest effects. Note that the second moment smear @@ -882,9 +903,8 @@ \begin{center} \includegraphics[width=\figwidth]{\picdir/radial_p1_r.\plotext} -\caption{Correlation of the PSF magnitude residuals ($\delta m_{PSF}$) - with the smear ($\sigma^2_{\mbox{major}} + \sigma^2_{\mbox{minor}}$) - signal for \gps\ (upper-left), \rps\ (upper-right), \ips\ (lower-left), - \zps\ (lower-right). -} \label{fig:effects.vs.radius} +\caption{Radial run of the four tree-ring trends for \rps: smear + ($\sigma^2_{\mbox{major}} + \sigma^2_{\mbox{minor}}$), PSF magnitude + residuals ($\delta m_{PSF}$), flat-field, and astrometric residuals + ($\delta R$). } \label{fig:effects.vs.radius} \end{center} \end{figure*} @@ -895,8 +915,6 @@ \begin{center} \includegraphics[width=\figwidth]{\picdir/radial_p2_r.\plotext} -\caption{Correlation of the PSF magnitude residuals ($\delta m_{PSF}$) - with the smear ($\sigma^2_{\mbox{major}} + \sigma^2_{\mbox{minor}}$) - signal for \gps\ (upper-left), \rps\ (upper-right), \ips\ (lower-left), - \zps\ (lower-right). +\caption{Radial run of the derivative of the smear ($\frac{\partial (\sigma^2_{major} + \sigma^2_{minor})}{\partial radius}$) + and astrometric residuals ($\delta R$) for \rps. } \label{fig:dsmear.and.astrom} \end{center} @@ -908,9 +926,7 @@ \begin{center} \includegraphics[width=\figwidth]{\picdir/radial_p3_r.\plotext} -\caption{Correlation of the PSF magnitude residuals ($\delta m_{PSF}$) - with the smear ($\sigma^2_{\mbox{major}} + \sigma^2_{\mbox{minor}}$) - signal for \gps\ (upper-left), \rps\ (upper-right), \ips\ (lower-left), - \zps\ (lower-right). -} \label{fig:dastrom.and.flat} +\caption{Radial run of + the derivative of the astrometric residuals ($\frac{\partial \delta + R}{\partial radius}$) and the flat-field for \rps.} \label{fig:dastrom.and.flat} \end{center} \end{figure*} @@ -931,9 +947,9 @@ Finally, the radial derivative of the radial component of the astrometric residual is correlated with the flat-field residual -errors. -\newtext{Figure~\ref{fig:dastrom.and.flat} shows the radial run of - $\frac{\partial \delta R}{\partial radius}$ and $\delta flat$ together - to illustrate this relationship.} -\oldtext{: $\frac{\partial \delta R}{\partial radius} \sim \delta flat$ (see Figure~14).} +errors. \newtext{Figure~\ref{fig:dastrom.and.flat} shows the radial + run of $\frac{\partial \delta R}{\partial radius}$ and the + flat-field together to illustrate this relationship.} \oldtext{: + $\frac{\partial \delta R}{\partial radius} \sim \delta flat$ (see + Figure~14).} This last relationship is somewhat weakly measured. Because of the @@ -953,21 +969,21 @@ image.} -\begin{table} -% \tiny -\begin{center} -\caption{Systematic Trends : Correlations between trends\label{table:correlation.by.trend}} -\begin{tabular}{|l|rrr|} -\hline -{\bf Filter} & {\bf psf mags} & {\bf $\grad$ smear} & {\bf $\grad$ astrom} \\ - & {\bf vs smear} & {\bf vs astrom} & {\bf vs flat} \\ -\hline -\gps & -0.056 & -0.060 & -0.47 \\ -\rps & -0.071 & -0.073 & -0.45 \\ -\ips & -0.077 & -0.095 & -0.45 \\ -\zps & -0.082 & -0.078 & -0.17 \\ -\hline -\end{tabular} -\end{center} -\end{table} +%% \begin{table} +%% % \tiny +%% \begin{center} +%% \caption{Systematic Trends : Correlations between trends\label{table:correlation.by.trend}} +%% \begin{tabular}{|l|rrr|} +%% \hline +%% {\bf Filter} & {\bf psf mags} & {\bf $\grad$ smear} & {\bf $\grad$ astrom} \\ +%% & {\bf vs smear} & {\bf vs astrom} & {\bf vs flat} \\ +%% \hline +%% \gps & -0.056 & -0.060 & -0.47 \\ +%% \rps & -0.071 & -0.073 & -0.45 \\ +%% \ips & -0.077 & -0.095 & -0.45 \\ +%% \zps & -0.082 & -0.078 & -0.17 \\ +%% \hline +%% \end{tabular} +%% \end{center} +%% \end{table} % smear vs psfmag @@ -1059,5 +1075,5 @@ \oldtext{(Figure~14)}\newtext{(Figure~\ref{fig:dastrom.and.flat})} is consistent with radial variations in the plate-scale. The -tree-rings observed by DES are completely attributed to effective +tree-rings observed in DECam are completely attributed to effective plate scale changes. Effective plate scale changes result in flat-field deviations because the flat-field illumination is a source @@ -1066,13 +1082,15 @@ affects the astrometry since these variations occur on spatial scales much smaller than the astrometric model. In this description of the -tree rings, the flat-field deviations are $-1 \times \frac{\partial +tree rings, the flat-field deviations are \newtext{proportional to $\frac{\partial +\delta R}{\partial r}$, as observed in Figure~\ref{fig:dastrom.and.flat}.} +\oldtext{$-1 \times \frac{\partial \delta R}{\partial r}$. The best-fit slopes of our correlations are \approx 0.5, but the signal-to-noise is rather low. A slope of -1 -appears to be consistent with our measurements. +appears to be consistent with our measurements.} The fact that the PSF ellipticity changes are {\em not} correlated with the tree-ring structure \oldtext{(Figure~6)}\newtext{(Figure~\ref{fig:all.effects.rband})} -tells us that, unlike the case for DES, the effective plate-scale +tells us that, unlike the case for DECam, the effective plate-scale changes seen in the flat-field and astrometry signals are not the dominant cause of the PSF photometry errors. Also, the fact that we @@ -1099,5 +1117,5 @@ tree-ring effects is the pattern of the doping variations in the silicon. As discussed by \cite{2014PASP..126..750P}, the tree-ring -patterns seen by the DES team are caused by lateral electic fields in +patterns seen by the DECam team are caused by lateral electic fields in the detector silicon (in the plane of the CCD wafer) generated by variations in the space charges embedded in the silicon, in turn @@ -1138,7 +1156,9 @@ by \cite{Holland.2003}, the charge diffusion is related to the space charge density by $\sigma \sim \rho^{-\frac{1}{2}}$ (their equation -6). Regions with high space charge densities increase the migration -speed of the photoelectrons and reduce the amount of charge diffusion -smearing; and vice versa for regions of low space-charge densities. +6). Regions with high space charge densities increase the electric +field in the depletion region for a fixed voltage, and thus increase +the migration speed of the photoelectrons, reducing the amount of +charge diffusion smearing; and vice versa for regions of low +space-charge densities. In summary, the variations in the space-charge density caused by @@ -1149,5 +1169,5 @@ photoelectrons, resulting in astrometric and flat-field deviations. -The DES team did not detect these charge diffusion variations. In +The DECam team did not detect these charge diffusion variations. In that case, the amplitude of the photometric effects due to the lateral field are dominant; these include both the modification of the @@ -1208,8 +1228,8 @@ diffusion. Unlike the non-uniform pixel-size effects, correction of the PSF photometry cannot simply be performed as an average flat-field -correction on the measurements after they have been processed. -The additional smearing acts as a convolution with a Gaussian kernel -of fixed size for a given filter. The photometry bias is a function -of the fractional change of the PSF size. Thus, the introduced error +correction on the measurements after they have been processed. The +additional smearing acts as a convolution with a Gaussian kernel of +fixed size for a given filter. The photometry bias is a function of +the fractional change of the PSF size. Thus, the introduced error depends on the average PSF for the image in question: an image with good image quality will suffer larger PSF model errors than an image @@ -1218,4 +1238,15 @@ modify the model PSFs as a function of position before they are used for the image analysis. + +The PV3 analysis of the Pan-STARRS $3\pi$ dataset applied an average +correction to the photometry and astrometry for each exposure as a +function of camera position with fine-enough resolution to follow +these tree-ring effects. However, since the photometry was only +corrected with an average flat-field-like correction, the full impact +of the smearing on the PSF photometry is not corrected. The remaining +systematic structure will tend to average out with many observations +in which the stars land on different portions of the detector. A +future re-processing will be required to completely correct for this +effect. The charge diffusion variations may also have an impact on Index: /trunk/doc/release.2015/systematics.20140411/response.v1.tex =================================================================== --- /trunk/doc/release.2015/systematics.20140411/response.v1.tex (revision 40306) +++ /trunk/doc/release.2015/systematics.20140411/response.v1.tex (revision 40306) @@ -0,0 +1,213 @@ +\documentstyle[gletter,epsfig]{article} +\begin{document} +\letterhead + +\begin{verbatim} +Figures: there are substantially more than are needed; I don't think you need +to show every band of every effect in Figs 3,4,5, and 6 - Fig 7 is by far the +most informative, allowing comparison of your diagnostics. Similarly, showing +all the scatter plots for all bands in Figs 8,9,10,11,12 is overkill, in my +view. I'm not sure any of these x vs y scatter plots are really needed; for me +it suffices to give the amplitudes of each effect relative to g band in +tabular or graphical form. The most important figure that I would want to see +is, in fact, missing: a plot of the radial run of each effect (a 1d plot of +Fig 7, basically), so that I can e.g. visually see whether the flat field is +the derivative of the astrometric displacements, as the lateral-field model +predicts. +\end{verbatim} + +This is a good point. We have greatly reduced the number of plots, +removing the 5-filter images in Figs 3,4,5,6 (astrometry, flat-field, +smear, shear) and the scatter plots for figures 8-14. We have +replaced these with 3 new figures: one showing the radial run of the 4 +effects with significant tree-ring patters (smear, psf mags, +flat-field, and astrometry) as suggested; a second plot showing the derivative +of the smear and the astrometry; a third plot showing the derivative +of the astrometry with the flat-field. We feel these new plots make +the relationships between the different trends much clearer than the +scatter plots. + +\begin{verbatim} +Also the spacing of the arrows in the ellipticity figures is too coarse to see +whether there are radial shear variations that are coherent with any of the +ring patterns. Here a 1d radial plot would be essential. +\end{verbatim} + +We disagree : the arrows are only showing the direction; the image in +the background shows the amplitude of the ellipticity variations, +which are completely dominated by the large-scale (optical) effects. +The resolution of the image is sufficient to follow the tree rings, +but there is no signal matching that pattern. + +\begin{verbatim} +page 4, Line 18: "non-pixel uniformity" ?? +\end{verbatim} + +Changed the wording to be clearer. + +\begin{verbatim} +4/22: "relatively coarse grid" - can you give the grid size? +\end{verbatim} + +Added the grid size. + +\begin{verbatim} +4/36: Can you say whether there are any differences between the "PV2 analysis +discussed here" and the public data releases, i.e. corrections that this work +implies need to be made to the public data? +\end{verbatim} + +Added a sentence about this. + +\begin{verbatim} +5/12: "diffusive effects of the lateral fields" - the lateral fields effects I +would not consider "diffusive" since they simply are conducting the charge +packets in a tranverse direction. The effect you find dominating the PS1 +devices differs in being primarily diffusive. +\end{verbatim} + +True, we mean charge motions or migrations caused by the lateral +fields. We've reworded this to be more accurate. + +\begin{verbatim} +Table 1, 8/13, other places: these values are referred to as the "per-pixel +standard deviations" which is unclear to me. Is this the SD of different +stellar measurements or flat-field pixels contributing to the mean of a given +super-pixel in the plots? Or is it the SD of the super-pixel averages, i.e. +the SD of the plotted values? Or something else? The most relevant statistic +is the SD after the shot noise of the individual measurements has been +suppressed, since we don't care about measurement noise in this paper. +\end{verbatim} + +This referred to the standard deviation of the pixel values in the +superpixel images. In retrospect, this table did not add much useful, +so we've removed it and changed the wording in the text to be clearer. + +\begin{verbatim} +Fig 1: In the caption it would be good to state that this is for one device in +the camera (even though this is stated in the text). +\end{verbatim} + +Done. + +\begin{verbatim} +Figs 1,2, 4: Why does Y band have so many missing (white) pixels? +impossible to look for patterns. +\end{verbatim} + +The effect from y-band fringing is much larger than the relevant +effects, saturating the dynamic range of the color scale. Added +wording to the captions to explain. + +\begin{verbatim} +Fig 3: what are the units of the plot? +\end{verbatim} + +This figure has now been removed; the text gives the dynamic range for +all 6 sub-figures in the r-band figure showing all effects. + +\begin{verbatim} +Table 2 (and in text in many places): you talk about "correlations" between +filters or different effects, but my guess is that what you are really +plotting is the amplitude of variations relative to those in g band. The +correlation coefficient would be something different - and less interesting, +since the correlation coefficient will depend heavily on size of irrelevant +sources of variation like measurement noise and the g-band striping. A graph +of these values (grizY along the x axis, different symbols for each effect) +might be more effective than this table. +\end{verbatim} + +This is correct. The table shows the relative amplitudes of the +effects. We have replaced this table with a plot of the relative +amplitudes of the 4 effects which show the tree-ring patterns. + +\begin{verbatim} +Table 3: same comment, what does "correlation" mean here? +\end{verbatim} + +Again, the is the relative amplitude of the effects compared with +g-band. Along with Table 1, we have decided to remove this table. +The conclusions of the paper are not dependent on the relative +amplitude of the effects, except to the extent that the blue filters +show stronger effects than the red filters. + +\begin{verbatim} +11/9: "deviations relative to the median flat-field image" - do you mean that +you are plotting the deviations relative to the median *of* the flat-field +image? Otherwise I don't understand. +\end{verbatim} + +Correct. The wording is fixed. + +\begin{verbatim} +11/25: A Gaussian smoothing of 3 superpixels - is this well above the max +"wavelength" of the rings? If not it can suppress a lot of them. Seems small +compared to the features visible in some of the figures. +\end{verbatim} + +This smoothing scale is large enough that it does not suppress +significantly the tree-ring features; without the smoothing, some of +the brick-wall pattern remains. The wording has been updated to clarify. + +\begin{verbatim} +Fig 14: what are the red and blue lines? +\end{verbatim} + +This figure has been removed and the information replaced with the new +Figure 6. + +\begin{verbatim} +24/19-21: These statements seem contradictory: the slope should be -1; the +data give a slope of 0.5; they are consistent. +\end{verbatim} + +The best-fit slope was actually $\sim -0.5$, with errors large enough +that it was consistent with -1.0: the missing minus sign was a typo. +However, we had a sign error in the calculation of the +proportionality; with the correct definition, the proprotionality is +positive. However, the noisy measurement of the slope was not very +informative. Referring to the new Figure 6, it is clear that that the +derivative of the astrometric variations is roughly proportional to +the flat-field variations. The point of this portion of this +discussion is simply that the astrometry and flat-field variations +behave in the way we would expect if they were caused by the lateral +charge migration effect, and thus this is a plausible explanation for +those observed effects. + +We have reworded the text to be clearer about this. + +\begin{verbatim} +24/23, 25/28, 25/30: Here you should probably be referring to DECam (the +camera) rather than DES (the survey). +\end{verbatim} + +Fixed. + +\begin{verbatim} +25/17: Is there any concise way to explain why adding dopants (space charge) +will *decrease* the transit time and charge diffusion at fixed voltage and +distance across the device? +\end{verbatim} + +The increased space charge increase the electric field in the device +for a fixed applied voltage, thus increasing the migration speed. +The description has been somewhat clarified. + +\begin{verbatim} +Conclusions: you should probably say something about which published +quantities of the PS1 catalogs are affected by this effect, by how much, and +whether any corrections have been applied, whether the effects are reduced by +averaging over multiple exposures. +\end{verbatim} + +Done. + +\epsfig{figure=signature1.ps,width=2.5in,angle=0} + +Eugene Magnier \\ +Specialist / Astronomer \\ +Pan-STARRS IPP Lead \\ +IfA / Pan-STARRS \\ +University of Hawaii + +\end{document}