diff --git a/PSTN-056.tex b/PSTN-056.tex index 0fe99ba..d0d6691 100644 --- a/PSTN-056.tex +++ b/PSTN-056.tex @@ -1,6 +1,6 @@ -\documentclass[PST,authoryear,toc, lsstdraft]{lsstdoc}%, lsstdraft] +\documentclass[PST,authoryear,toc]{lsstdoc}%, lsstdraft] \input{meta} \defcitealias{PSTN-055}{PSTN-055} @@ -56,7 +56,7 @@ % \setDocCurator{The Curator of this Document} \setDocAbstract{% -We present the final planned comprehensive recommendation for Rubin Observatory the Legacy Survey of Space and Time (LSST) survey strategy ahead of the start of LSST: this recommendation is the product of a many-years-long iterative process where community recommendations to maximize the scientific impact of LSST across domains of astrophysics were reviewed, synthesized, aggregate, and merged to define the overall plan for 10 years of LSST observations. The current recommendation builds on Phase 1 \citep{PSTN-053} and Phase 2 recommendations \citep{PSTN-055} and, together, they define a 10-year plan for observing. Here we answer questions left open in \citetalias{PSTN-055}, refine additional survey details, and describe the scope of future activities of the SCOC. +We present the final planned comprehensive recommendation for Rubin Observatory the Legacy Survey of Space and Time (LSST) survey strategy ahead of the start of LSST. This recommendation is the product of a many-years-long iterative process where community recommendations to maximize the scientific impact of LSST across domains of astrophysics were reviewed, synthesized, aggregated, and merged to define the overall plan for 10 years of LSST observations. The current recommendation builds on Phase 1 (\citetalias{PSTN-053}) and Phase 2 recommendations (\citetalias{PSTN-055}) and, together, they define a 10-year plan for observing. Here we answer questions left open in \citetalias{PSTN-055}, refine additional survey details, and describe the scope of future activities of the SCOC. } % Change history defined here. diff --git a/additional.tex b/additional.tex index ae82100..9e42bc4 100644 --- a/additional.tex +++ b/additional.tex @@ -10,7 +10,7 @@ \section{Additional Recommendations}\label{sec:additional} \caption{Small changes to the southern portion of the footprint improve the overlap with Euclid. \label{fig:euclid-overlap}} \end{figure} -{\bf The airmass limits for the Near-Sun Twilight microsurvey, introduced with baseline v3.0, were increased from $X=2.5$ to $X=3.0$} in \texttt{v3.2}, corresponding to decreasing the minimum solar elongation reached for this microsurvey from 40 degrees to 35 degrees (the range of solar elongations changed from 40 to 60 degrees in \texttt{v3.0} to 35 to 47 degrees in\texttt{v4.0}). This improves the likelihood of discovery of objects with interior-to-Earth orbits, increasing the survey sensitivity to this niche of discovery space. The recovered population of objects +{\bf The airmass limits for the Near-Sun Twilight microsurvey, introduced with \baseline{3.0}, were increased from $X=2.5$ to $X=3.0$} in \texttt{v3.2}, corresponding to decreasing the minimum solar elongation reached for this microsurvey from 40 degrees to 35 degrees (the range of solar elongations changed from 40 to 60 degrees in \texttt{v3.0} to 35 to 47 degrees in \texttt{v4.0}). This improves the likelihood of discovery of objects with interior-to-Earth orbits, increasing the survey sensitivity to this niche of discovery space. The recovered population of objects interior to Venus at magnitude $H\leq20$ goes from \mbox{$\sim$4\%} to \mbox{$\sim$40\%} in \texttt{v3.2} and later. The impacts outside the microsurvey are negligible. \clearpage @@ -20,7 +20,7 @@ \section{Additional changes introduced throughout the \texttt{v3.x} \opsim s } \ Some important assumptions underlying the simulations were updated in Phase 3 of the survey strategy recommendation process: \begin{itemize} -\item As of \baseline{3.6}, the downtime in year one was increased to reflect a more realistic transition into operations. This change adds approximately eight weeks of downtime reducing the number of visits by \mbox{$\sim$5\%}. The downtime in Y1 is simulated to be maximal early on and decreased to the level expected for the general LSST survey by the end of the first year (\autoref{fig:downtime}). Future simulations will aim to improve the unscheduled downtime model to better align with expectations from the Rubin Observatory Operations team. +\item As of \baseline{3.6}, the downtime in Y1 was increased to reflect a more realistic transition into operations. This change adds approximately eight weeks of downtime reducing the number of visits by \mbox{$\sim$5\%}. The downtime in Y1 is simulated to be maximal early on and decreased to the level expected for the general LSST survey by the end of the first year (\autoref{fig:downtime}). Future simulations will aim to improve the unscheduled downtime model to better align with expectations from the Rubin Observatory Operations team. \item As of \baseline{3.6}, the effect of jerk on slew time is included in the simulations, and thus included in scheduling choices. Functionally, this slightly increases the overhead and decreases survey efficiency (\autoref{fig:downtime}). @@ -29,7 +29,7 @@ \section{Additional changes introduced throughout the \texttt{v3.x} \opsim s } \ \includegraphics[width=0.37\linewidth]{figures/downtime_v3_5_year1.png} \includegraphics[width=0.37\linewidth]{figures/downtime_v4_0_year1.png} \includegraphics[width=0.24\linewidth]{figures/downtime_v4_0_year1_legend.png} - \caption{The time within each night of LSST (observing limited to hours darker than nautical twilight: when the sun is \mbox{$\leq-$12$\deg$} from the horizon) in Y1 divided into on-sky exposure time, overhead for those exposures (shutter and readout time), time spent slewing, and downtime due to weather, scheduled maintenance activities, or unscheduled engineering. Before \baseline{3.6} (left plot), simulations only included steady-state expected engineering downtime, modeled as full-night downtime blocks. The \baseline{4.0} simulation (right plot) includes additional unscheduled downtime time within the first 380 nights of the survey, including breaks as short as an hour to reflect the need for engineering early in the survey. } + \caption{The time within each night of LSST in Y1 divided into on-sky exposure time, overhead for those exposures (shutter and readout time), time spent slewing, and downtime due to weather, scheduled maintenance activities, or unscheduled engineering. Observing is limited to hours darker than nautical twilight (when the sun is \mbox{$\leq-$12$\deg$} from the horizon). Before \baseline{3.6} (left plot), simulations only included steady-state expected engineering downtime, modeled as full-night downtime blocks. The \baseline{4.0} simulation (right plot) includes additional unscheduled downtime time within the first 380 nights of the survey, including breaks as short as an hour, to reflect the need for engineering early in the survey. } \label{fig:downtime} \end{figure} diff --git a/answers.tex b/answers.tex index dbab935..8c03e72 100644 --- a/answers.tex +++ b/answers.tex @@ -9,7 +9,7 @@ \subsection{Swapping filters on the filter wheel}\label{sec:filterswap} Filters will be swapped in and out of the filter wheel based on sky brightness due to the lunar phase. In \citetalias{PSTN-055} \S4, the SCOC recommended further investigation of which filters to swap: \begin{quote} - {[\citetalias{PSTN-055} \S4] ``The SCOC recommends that the investigation of the filter swapping schemes on the filter wheel continue. After the November 2022 SCOC workshop, a few experiments in swapping $u$, $z$, and $y$ instead of $u$ and $z$ were implemented in \texttt{v2.99} simulations. More work is needed to understand the impacts of this decision on the DDFs as well as on the WFD. Filter pairing prescriptions for the observation pairs should also be explored in some more depth.''} + {[\citetalias{PSTN-055} \S4] ``The SCOC recommends that the investigation of the filter swapping schemes on the filter wheel continue. After the November 2022 SCOC workshop, a few experiments in swapping $u$, $z$, and $y$ instead of $u$ and $z$ were implemented in \texttt{v2.99} simulations. More work is needed to understand the impacts of this decision on the DDFs as well as on the WFD.''} \end{quote} Simulations prior to \texttt{v3.0} swapped $z$ with $u$ based on lunation, as scattered moonlight is blue and impacts observations in $u$-band most significantly. Simulations tagged \texttt{v3.2} experimented with swapping $u$ with $z$ or $y$, including putting all of $u$, $z$, and $y$ on rotation. Swapping a filter has two effects: it adds a gap for the period while it is unavailable, and it increases the cadence in that bandpass during the time it is mounted to achieve the final desired number of observations. Increasing the availability of $z$ on the filter wheel produced significant improvements in supernova (SN) cosmology, especially in the Deep Drilling Fields (DDFs), while swapping two filters instead of three improves coverage at short time scales in filters through $z$ with significant benefits for the study of rapid-evolving transients (\eg\ Kilonovae, KN, see \autoref{fig:swapping}). Keeping the $g$, $r$, $i$, and $z$ filters in the camera at all times also reduces the risk of damaging these critical filters during filter swaps. @@ -35,9 +35,9 @@ \subsection{Filter Balance}\label{sec:filterbalance} {[\citetalias{PSTN-055} \S4] ``While the SCOC recommends the filter balance as implemented starting in \baseline{2.0} should not be changed, it is possible that rebalancing the exposure time to compensate for performance and throughput in some filters as compared to others or shortening exposures in filters where the throughput exceeds expectations enabling the collection of more images in that filter (or overall) would lead to enhanced LSST science. The SCOC cannot finalize this recommendation at this time due to missing information about the characteristics of the system-as-built.''} \end{quote} -Simulations of the survey strategy up to and including \baseline{3.2} use throughput curves assuming mirror coating as Al-Ag-Al respectively for M1-M2-M3. The plan to coat the mirrors was updated in 2023 to Ag-Ag-Ag (or 3xAg), which leads to a \mbox{$\sim$15-20\%} increase in survey efficiency compared to Al-Ag-Al by increasing throughput in all bands redder than $u$, and bringing throughput closer to the design goals as stated in \citetalias{LPM-17}.\footnote{\url{https://community.lsst.org/t/rubin-sim-v1-3-released/7937} and \url{https://github.com/lsst-pst/syseng_throughputs/blob/main/notebooks/SilverVsAluminum.ipynb}.} \autoref{tab:dm5Agx3} shows the magnitude limit changes associated with the two different coatings for both detector types in the camera. Ag-Ag-Ag coating increases sensitivity in all LSST bands from $g$ to $y$, but it decreases the throughput in $u$. As of \baseline{3.3}, all \opsim\ simulations include the Ag-Ag-Ag expected throughput. +Simulations of the survey strategy up to and including \baseline{3.2} use throughput curves assuming mirror coating as Al-Ag-Al respectively for M1-M2-M3. The plan to coat the mirrors was updated in 2023 to Ag-Ag-Ag (or 3xAg), which leads to a \mbox{$\sim$15-20\%} increase in survey efficiency compared to Al-Ag-Al by increasing throughput in all bands redder than $u$, and bringing throughput closer to the design goals as stated in \citetalias{LPM-17}.\footnote{\url{https://community.lsst.org/t/rubin-sim-v1-3-released/7937} and \url{https://github.com/lsst-pst/syseng_throughputs/blob/main/notebooks/SilverVsAluminum.ipynb}.} However, while Ag-Ag-Ag coating increases sensitivity in $grizy$, it decreases the throughput in $u$. \autoref{tab:dm5Agx3} shows the magnitude limit changes associated with the two different coatings for both detector types in the camera. As of \baseline{3.3}, all \opsim\ simulations include the Ag-Ag-Ag expected throughput. -The SCOC reviewed the largely positive impact of the new throughput on science cases: nearly all MAFs responded positively to the increase in survey depth (see \autoref{fig:heatmap} and note the significant improvements between \baseline{3.2} and \baseline{3.3} when the new filter transmission curves were introduced). Some system metrics corresponding to \citetalias{LPM-17} requirements show improvements as large as 10\% (\eg\ Parallax uncertainty, see \autoref{fig:parallax}) and some time domain metrics improve by \mbox{$\sim$20\%} (Kilonovae and SN Ia metrics, see \autoref{fig:heatmap}). +The SCOC reviewed the largely positive impact of the new throughput on science cases: nearly all MAFs responded positively to the increase in survey depth (see \autoref{fig:heatmap} and note the significant improvements between \baseline{3.2} and \baseline{3.3}). Some system metrics corresponding to \citetalias{LPM-17} requirements show improvements as large as 10\% (\eg\ Parallax uncertainty, see \autoref{fig:parallax}) and some time domain metrics improve by \mbox{$\sim$20\%} (Kilonovae and SN Ia metrics, see \autoref{fig:heatmap}). \clearpage \begin{longtable}{lccccc} @@ -64,8 +64,8 @@ \subsection{Filter Balance}\label{sec:filterbalance} -However, the SCOC understands that the throughput loss in $u$-band (\mbox{$\sim$30\%} loss in coadded depth) would negatively impact science cases, including Photo-z, studies of the Milky Way halo, and Lyman Break Galaxies (LBGs, identified as $u$-band dropouts at redshift $z\sim3$). While the overwhelming majority of the MAF metrics available to the SCOC responded positively to the updated throughput, we are aware, as always, that these may not provide an exhaustive picture of the science outcomes. Therefore, guided by experts in the community, we explored ways to reduce the $u$-band magnitude decrease while preserving the benefit of increased throughput in redder bands. -We tracked the performance of Photo-z, as characterized in \citealt{Graham_2017}, by assessing the variance and bias in \pz\ at redshifts $z\lesssim{3}$ (\autoref{fig:pz}). \pz\ is sensitive to $u$-band depth at redshift $z\geq 2$ due to decreased power to identify Lyman break galaxies photometrically. We expect that recovering \pz\ performance is a good indicator of recovering performance for other science cases sensitive to $u$-band depth for which we do not have detailed metrics. \pz\ performance, along with a large set of MAFs, was run against a set of \opsim s that progressively changed the exposure time and the number of exposures in $u$-band (see \autoref{fig:uband})\footnote{\url{https://community.lsst.org/t/release-of-v3-4-simulations/8548}}. +However, while the overwhelming majority of the MAF metrics available to the SCOC responded positively to the updated throughput, we are aware, as always, that these may not provide an exhaustive picture of the science outcomes. The SCOC understands that the throughput loss in $u$-band (\mbox{$\sim$30\%} loss in coadded depth) would negatively impact science cases, including Photo-z, studies of the Milky Way halo, and Lyman Break Galaxies (LBGs, identified as $u$-band dropouts at redshift $z\sim3$). Therefore, guided by experts in the community, we explored ways to reduce the $u$-band magnitude decrease while preserving the benefit of increased throughput in redder bands. +We tracked the performance of Photo-z, as characterized in \citealt{Graham_2017}, by assessing the variance and bias in \pz\ at redshifts $z\lesssim{3}$ (\autoref{fig:pz}). \pz\ is sensitive to $u$-band depth at redshift $z\geq 2$ due to decreased power to identify Lyman break galaxies photometrically. We expect that recovering \pz\ performance is a good indicator of recovering performance for other science cases sensitive to $u$-band depth for which we do not have detailed metrics. \pz\ performance, along with a large set of MAFs, was thus run against a set of \opsim s that progressively changed the exposure time and the number of exposures in $u$-band (see \autoref{fig:uband})\footnote{\url{https://community.lsst.org/t/release-of-v3-4-simulations/8548}}. {\bf The SCOC recommends:} @@ -76,9 +76,9 @@ \subsection{Filter Balance}\label{sec:filterbalance} \end{itemize} -This roughly restores the $u$-band depth of LSST in \baseline{3.0} with minimal impact on LSST science metrics. As a science case that is representative of those sensitive to $u$-band depth, these changes recover performance on \pz\ at redshift $z\sim2$, where the impact of the $u$-band throughput loss was most significant, while maintaining the performance improvement on \pz\ at low redshift afforded by the increased depth of LSST in all other bands. Furthermore, these changes minimally impact other science cases tracked by MAFs. +This roughly restores the $u$-band depth of LSST \baseline{3.0} with minimal impact on other LSST science. As a science case that is representative of those sensitive to $u$-band depth, these changes recover performance on \pz\ at redshift $z\sim2$, where the impact of the $u$-band throughput loss was most significant, while maintaining the performance improvement on \pz\ at low redshift afforded by the increased depth of LSST in all other bands (\autoref{fig:pz}). Furthermore, these changes minimally impact other science cases tracked by MAFs (\autoref{fig:uband}). -Because more science cases generally respond better to increasing the number of images, over increasing the exposure time to achieve the same depth, the added $u$-band time should be obtained by decreasing (minimally) the exposure time in other bands rather than decreasing the number of images. +Because more science cases generally respond better to increasing the number of images, over increasing the exposure time to achieve the same depth, the added $u$-band time should be obtained by decreasing (minimally) the exposure time in other bands rather than decreasing the number of visits. Simulations show a decrease in exposure of 0.8 seconds per visit in all other bands compensates for the added $u$-band time. \textbf{ We note that this recommendation is subject to ongoing feasibility studies by the Rubin Data Management team.} @@ -87,7 +87,7 @@ \subsection{Filter Balance}\label{sec:filterbalance} \centering \includegraphics[height=0.6\textwidth]{figures/photo-z.png} -\caption{The effect of changes in $u$-band 10-year depth on \pz\ Robust Standard deviation as a function of redshift $z$, as measured in \cite{Graham_2017}. The dashed line represents the \citetalias{LPM-17} requirements on \pz. The black solid curve is the \baseline{3.2}, the latest baseline before the filter transmission curves were updated in the Rubin simulation system, the colored curves represent the \pz\ Robust Standard Deviation varying the $u$-band exposure time between $30\leq u_{expt} \leq45$~seconds and the number of exposures in $u$-band between $1.0\times ns \leq N_u \leq 1.1\times ns$, where $ns$ is the number of $u$-band exposures in \baseline{3.2}. Note that, with the caveat that sampling uncertainties are large at $z>1.5$, with the Ag-Ag-Ag transmission curves we note an improvement in \pz\ at low $z$ associated with increased depth in bands redder than $u$, but a degradation at $z>1.5$. The current filter balance recommendation (closely reflected by the red line in this plot) more than recovers performance at high $z$ while preserving the low $z$ gains. A similar impact is seen in \pz\ bias.} +\caption{The effect of changes in $u$-band 10-year depth on \pz\ Robust Standard deviation as a function of redshift $z$, as measured in \cite{Graham_2017}. The dashed line represents the \citetalias{LPM-17} requirements on \pz. The black solid curve is the \baseline{3.2}, the latest baseline before the filter transmission curves were updated in the Rubin simulation system, the colored curves represent the \pz\ Robust Standard Deviation varying the $u$-band exposure time between $30\leq u_{expt} \leq45$~seconds and the number of exposures in $u$-band between $1.0\times ns \leq N_u \leq 1.1\times ns$, where $ns$ is the number of $u$-band exposures in \baseline{3.2}. Note that, with the caveat that sampling uncertainties are large at $z>1.5$, with the Ag-Ag-Ag transmission curves we note an improvement in \pz\ at low $z$ associated with increased depth in bands redder than $u$, but a degradation at $z>1.5$. The current filter balance recommendation (closely reflected by the red line in this plot) more than recovers performance at high $z$ while preserving the low $z$ gains. A similar effect is seen in \pz\ bias.} \label{fig:pz} \end{figure} @@ -100,7 +100,7 @@ \subsection{Filter Balance}\label{sec:filterbalance} \FloatBarrier \subsection{Rolling}\label{sec:rolling} -In a rolling strategy, instead of distributing visits uniformly on the WFD footprint, the sky is split into regions that alternate between high- and low-intensity monitoring. In \citetalias{PSTN-055}, the SCOC recommended the implementation of a rolling strategy for the LSST WFD at a strength of 0.9\footnote{This number represents the fraction of the visits that the scheduler attempts to place in the high-activity rolling region. However, the resulting visit distribution is more uniform (75--80\% in high-activity regions, 25--20\% in low ones) due to competing requirements (\eg , filter balance, minimum number of observations per pointing per year in each filter to produce templates, weather, etc...)} +In a rolling strategy, instead of distributing visits uniformly on the WFD footprint, the sky is split into regions that alternate between high- and low-intensity monitoring. In \citetalias{PSTN-055}, the SCOC recommended the implementation of a rolling strategy for the LSST WFD at a strength of 0.9\footnote{This number represents the fraction of the visits that the scheduler attempts to place in the high-activity rolling region. However, the resulting visit distribution is more uniform (75--80\% in high-activity regions, 25--20\% in low ones) due to competing requirements (\eg , filter balance, minimum number of observations per pointing per year in each filter to produce templates, weather, etc...).} with the sky split into two rolling regions constituted by four longitudinal stripes. The primary drivers for this recommendation are time-domain science, including the exploration of the transient and variable sky and SN cosmology. Rolling as described decreases the median time gaps compared to a no-rolling implementation of LSST: distributing the \mbox{$\sim$800} visits per pointing evenly into 10 seasons results in a median revisit time per pointing of about 4.5 nights, while rolling can increase the cadence on the areas of sky closer to 2.5 nights.\footnote{The reader is reminded that each pointing receives two or three visits per night. The time gaps reported here are for inter-night observations.} @@ -117,9 +117,43 @@ \subsection{Rolling}\label{sec:rolling} With the goal of quantifying the necessary uniformity to enable cosmological results at certain key data releases,\footnote{These intermediate releases were selected because they enable equally spaced time intervals between new datasets for comprehensive static science analysis: years 1, 4, 7, and 10 corresponding to DR2, DR5, DR8, DR11.} DR5 and DR8, as well as identifying solutions that enable rolling (at the strength recommended in \citetalias{PSTN-055}) while increasing the uniformity of key data releases, a Uniformity Task Force developed alternative rolling implementations. - {\it Uniform rolling} implements interruptions of rolling before specific data releases to increase the uniformity of those releases and recover an acceptable level of uniformity at key years---see Figs.~\ref{fig:stripiness} and~\ref{fig:uniform-rolling}. It was found that uniform rolling permits the full survey area to be used for cosmological analysis at years 4 and 7, whereas in previous rolling versions, approximately 35\% of the cosmological constraining power\footnote{Here we quantify cosmological constraining power through emulated forecasts of combined constraints from cosmological weak lensing and large-scale structure measurements \citep{2022ApJS..259...58L}. The constraints assume a $w_0 w_a$CDM cosmological model, with $w_0$ and $w_a$ entering as two parameters in the dark energy equation of state. The constraining power is quantified through the area of the uncertainty contours in the $(w_0, w_a)$ part of parameter space, marginalizing over other cosmological parameters and systematic uncertainties -- then taking the inverse of that area (so that higher values mean lower uncertainty, \ie, tighter cosmological constraints). However, this can be considered more generally as a proxy for how well we are measuring cosmological structure growth, translating into tighter constraints on the amplitude of matter fluctuations if a $\Lambda$CDM cosmological model is assumed.} was lost at those years due to the need for area cuts. -We note that over the 10-year LSST, the envisioned two-region rolling strategy can be implemented with at most four rolling cycles (that is, starting rolling in Y2 and ending rolling in Y10) where a cycle is defined as a pair of two years, where the high- and low-intensity regions are swapped. Uniform rolling requires limiting rolling to three cycles. Rolling primarily benefits science sensitive to timescales of \mbox{$\sim$24-48} hours.\footnote{Shorter time scales are primarily covered by observations in triplets, as discussed in \citetalias{PSTN-055}.} + + + + +\begin{figure} + \centering + %\begin{overpic}[width=0.8\textwidth]{figures/Rolling.png} + % \put(50,30){\color{lsstblue}\huge DRAFT} + %\end{overpic} + \includegraphics[width=0.32\linewidth]{figures/baseline_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} + %\begin{overpic}[width=0.8\textwidth]{figures/RollingCompare.png} + % \put(50,30){\color{lsstblue}\huge DRAFT} + %\end{overpic} + \includegraphics[width=0.32\linewidth]{figures/noroll_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} + %\begin{overpic}[width=0.8\textwidth]{figures/RollingUniform.png} + % \put(50,30){\color{lsstblue}\huge DRAFT} + %\end{overpic} + \includegraphics[width=0.32\linewidth]{figures/roll_uniform_early_half_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} + + \includegraphics[width=0.6\linewidth]{figures/roll_colorbar.png} + \caption{Comparison of the depth of LSST at the end of Y4 (DR5) under different rolling strategies. The left plot shows the LSST number of visits map for a standard implementation of rolling at strength 0.9 in two sky regions designed as four longitudinal stripes. The center plot represents an implementation of LSST without rolling, for comparison, and provides an upper limit to expected uniformity. The right plot shows the implementation of Uniform Rolling described in this section and implemented in \baseline{4.0}.} + \label{fig:uniform-rolling} +\end{figure} + + {\it Uniform rolling} implements interruptions of rolling before specific data releases to increase the uniformity of those releases and recover an acceptable level of uniformity at key years, see \autoref{fig:uniform-rolling} and \autoref{fig:stripiness}. It was found that uniform rolling permits the full survey area to be used for cosmological analysis at years 4 and 7, whereas in previous rolling versions, approximately 35\% of the cosmological constraining power\footnote{Here we quantify cosmological constraining power through emulated forecasts of combined constraints from cosmological weak lensing and large-scale structure measurements \citep{2022ApJS..259...58L}. The constraints assume a $w_0 w_a$CDM cosmological model, with $w_0$ and $w_a$ entering as two parameters in the dark energy equation of state. The constraining power is quantified through the area of the uncertainty contours in the $(w_0, w_a)$ part of parameter space, marginalizing over other cosmological parameters and systematic uncertainties -- then taking the inverse of that area (so that higher values mean lower uncertainty, \ie, tighter cosmological constraints). However, this can be considered more generally as a proxy for how well we are measuring cosmological structure growth, translating into tighter constraints on the amplitude of matter fluctuations if a $\Lambda$CDM cosmological model is assumed.} was lost at those years due to the need for area cuts. + +\begin{figure} + \centering + \includegraphics[width=0.75\linewidth]{figures/stripiness_metric.png} + \caption{ + A quantitative assessment of the non-uniform exposure time variation vs.\ year under different observing strategies. The test statistic plotted on the vertical axis effectively measures the fractional difference between the variations in depth between the northern and southern Galactic regions, with a value of 0 indicating that the two are the same, as expected for a perfectly uniform survey. The light green shaded envelope between the dashed black lines indicates the region for which we consider the stripe features to be negligible (meaning manageable within the limits of existing analysis algorithms). The narrower dark green shaded envelope shows the expected statistical fluctuations for a survey without rolling, as estimated using the \texttt{noroll\_v3.4} strategy simulation. As shown, at the highlighted years (1, 4, 7, 10), the uniform rolling strategy (\texttt{roll\_uniform\_early\_half\_mjdp0\_v3.4\_10yrs}) is very close to uniform within the level of statistical fluctuations at Y4 and Y7 (DR5 and DR8), while the \baseline{3.4} strategy is highly non-uniform, especially in those years. +} + \label{fig:stripiness} +\end{figure} + +We note that over the 10-year LSST, the envisioned two-region rolling strategy can be implemented with at most four rolling cycles (that is, starting rolling in Y2 and ending rolling in Y10) where a cycle is defined as a pair of two years where the high- and low-intensity regions are swapped. Uniform rolling requires limiting rolling to three cycles. Rolling primarily benefits science sensitive to timescales of \mbox{$\sim$24-48} hours.\footnote{Shorter time scales are primarily covered by observations in triplets, as discussed in \citetalias{PSTN-055}.} We note that these time scales had been identified as sensitive and requiring additional improvements in \citetalias{PSTN-055} within the recommendation on rolling (\citetalias{PSTN-055} \S2.4.1): \begin{quote} @@ -129,7 +163,6 @@ \subsection{Rolling}\label{sec:rolling} %While not designed for this purpose, t The implementation of the 0.9 strength, two-sky-areas rolling with four cycles (\baseline{3.2}-\baseline{3.5}) improved coverage at 24-48 hours (see figure \autoref{fig:rolling_sampling}) over the \baseline{3.0} that accompanied \citetalias{PSTN-055}. - \begin{figure} \centering %\begin{overpic}[width=0.8\textwidth]{figures/rolling_sampling.png}{figures/rolling_nsamples.png} @@ -142,21 +175,14 @@ \subsection{Rolling}\label{sec:rolling} \label{fig:rolling_sampling} \end{figure} -Considering the above input for multiple science cases, the SCOC recognizes the positive impact that this rolling implementation (3-cycle {\it uniform rolling}) has on static cosmological and extragalactic probes and considers this a promising solution for the uniformity concerns raised in \citetalias{PSTN-055}, with limited detrimental impact on time-domain probes. However, this implementation of rolling is a significant and new departure from earlier implementations, and the number of cycles of rolling had not been previously explicitly discussed as a parameter in the survey strategy. Furthermore, the current Uniform Rolling implementation requires rolling to start early in Y2 (in the current implementations, rolling starts on survey day $\leq 400$) but, as discussed in \citetalias{PSTN-055}, rolling shall not start until sufficient sky coverage has been achieved to enable proper photometric calibration.%\footnote{\tbd{A discussion and references for "good enough" calibration should be added, see Eli 2023 PCW - @fed}} -\begin{figure} - \centering - \includegraphics[width=0.75\linewidth]{figures/stripiness_metric.png} - \caption{ - A quantitative assessment of the non-uniform exposure time variation vs.\ year under different observing strategies. The test statistic plotted on the vertical axis effectively measures the fractional difference between the variations in depth between the northern and southern Galactic regions, with a value of 0 indicating that the two are the same, as expected for a perfectly uniform survey. The light green shaded envelope between the dashed black lines indicates the region for which we consider the stripe features to be negligible (meaning manageable within the limits of existing analysis algorithms). The narrower dark green shaded envelope shows the expected statistical fluctuations for a survey without rolling, as estimated using the \texttt{noroll\_v3.4} strategy simulation. As shown, at the highlighted years (1, 4, 7, 10), the uniform rolling strategy (\texttt{roll\_uniform\_early\_half\_mjdp0\_v3.4\_10yrs}) is very close to uniform within the level of statistical fluctuations at Y4 and Y7 (DR5 and DR8), while the \baseline{3.4} strategy is highly non-uniform, especially in those years. -} - \label{fig:stripiness} -\end{figure} +Considering the above input for multiple science cases, the SCOC recognizes the positive impact that this rolling implementation (3-cycle Uniform Rolling) has on static cosmological and extragalactic probes and considers this a promising solution for the uniformity concerns raised in \citetalias{PSTN-055}, with limited detrimental impact on time-domain probes. However, this implementation of rolling is a significant and new departure from earlier implementations, and the number of cycles of rolling had not been previously explicitly discussed as a parameter in the survey strategy. Furthermore, the current Uniform Rolling implementation requires rolling to start early in Y2 (in the current implementations, rolling starts on survey day $\leq 400$) but, as discussed in \citetalias{PSTN-055}, rolling shall not start until sufficient sky coverage has been achieved to enable proper photometric calibration.%\footnote{\tbd{A discussion and references for "good enough" calibration should be added, see Eli 2023 PCW - @fed}} + For these reasons, the SCOC is not committing at this time to recommend any specific implementation of rolling, beyond confirming the strength of 0.9 and two-region strategy. Since in all current implementations, rolling does not begin until Y2, the SCOC intends to continue investigating rolling implementations and their impact throughout Y1, with the support of the community, and release a recommendation of how to implement rolling as part of its first annual recommendation ahead of Y2 of Operations. In particular, we intend to (1) investigate sensitivity to the outcomes of Y1, (2) ensure the community has time to evaluate the potential impacts of these changes that are not currently highlighted by our metrics, and (3) refine the uniform rolling implementation details. {\bf The SCOC recommends that the time domain community, particularly those interested in phenomena that have evolutionary timescales of hours-to-days, urgently quantify the impact of the proposed uniform rolling compared to rolling in four cycles. For this purpose, while the baseline is implemented with 3-cycle uniform rolling, the Survey Strategy team has prepared \texttt{v3.6} \opsim s with different rolling implementations.} -{\bf Further, the SCOC restates its recommendation that Data Management scopes a plan for producing uniform data releases in DR5 and DR8, in addition to the standard data releases. The cost of the development and storage of these additional data and the timing of their release should be scoped and shared with the scientific community.} Even if produced by the Rubin DM, uniform data releases will require the input of the DESC and extragalactic science community at large to develop the algorithm that will achieve sufficient uniformity and depth. Understanding the cost of producing two additional {\it uniform} data releases is necessary to compare this cost to the scientific cost of three vs. four cycles of rolling, to be measured by the community (see previous paragraph). In addition, if rolling cannot start early enough to interrupt rolling ahead of DR5 and DR8, this remains the only alternative solution currently identified to achieve sufficient uniformity. Sharing information on the cost of additional data releases will place the community in a position to, if needed, advocate for and secure funding for this purpose. +{\bf Further, the SCOC restates its recommendation that Data Management scopes a plan for producing uniform data releases in DR5 and DR8, in addition to the standard data releases. The cost of the development and storage of these additional data and the timing of their release should be scoped and shared with the scientific community.} Even if produced by the Rubin DM, uniform data releases will still require the input of the DESC and extragalactic science community at large to develop the algorithm that will achieve sufficient uniformity and depth. Understanding the cost of producing two additional {\it uniform} data releases is necessary to compare this cost to the scientific cost of three vs. four cycles of rolling, to be measured by the community (see previous paragraph). In addition, if rolling cannot start early enough to interrupt rolling ahead of DR5 and DR8, this remains the only alternative solution currently identified to achieve sufficient uniformity. Sharing information on the cost of additional data releases will place the community in a position to, if needed, advocate for and secure funding for this purpose. The SCOC is thankful to the Uniformity Task Force, chaired by Rachel Mandelbaum, which provided invaluable contributions and analysis that led us to this recommendation. @@ -164,31 +190,12 @@ \subsection{Rolling}\label{sec:rolling} -\begin{figure} - \centering - %\begin{overpic}[width=0.8\textwidth]{figures/Rolling.png} - % \put(50,30){\color{lsstblue}\huge DRAFT} - %\end{overpic} - \includegraphics[width=0.32\linewidth]{figures/baseline_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} - %\begin{overpic}[width=0.8\textwidth]{figures/RollingCompare.png} - % \put(50,30){\color{lsstblue}\huge DRAFT} - %\end{overpic} - \includegraphics[width=0.32\linewidth]{figures/noroll_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} - %\begin{overpic}[width=0.8\textwidth]{figures/RollingUniform.png} - % \put(50,30){\color{lsstblue}\huge DRAFT} - %\end{overpic} - \includegraphics[width=0.32\linewidth]{figures/roll_uniform_early_half_v3_4_Nvisits_Year_4_HEAL_SkyMap.png} - - \includegraphics[width=0.6\linewidth]{figures/roll_colorbar.png} - \caption{Comparison of the depth of LSST at the end of Y4 (DR5) under different rolling strategies. The left plot shows the LSST number of visits map for a standard implementation of rolling at strength 0.9 in two sky regions designed as four longitudinal stripes. The center plot represents an implementation of LSST without rolling, for comparison, and provides an upper limit to expected uniformity. The right plot shows the implementation of Uniform Rolling described in this section and implemented in \baseline{4.0}.} - \label{fig:uniform-rolling} -\end{figure} \FloatBarrier \subsection{Galaxy}\label{sec:galaxy} -In \citetalias{PSTN-055} the SCOC identified areas of work needed to finalize the WFD survey strategy on the Galactic sky and special regions of interest to Galaxy science, including the Clouds and South Celestial Pole, which can be observed within the WFD but with different observing choices than the low-dust footprint, of primary interest for extragalactic science. +In \citetalias{PSTN-055} the SCOC identified areas of work needed to finalize the WFD survey strategy on the Galactic sky and special regions of interest to Galaxy science, including the LMC, SMC, and South Celestial Pole, which can be observed within the WFD but with different observing choices than the low-dust footprint, of primary interest for extragalactic science. \begin{quote} {[\citetalias{PSTN-055} \S4] ``The SCOC is not ready to finalize a recommendation for the filter balance in the Galactic Plane, or for a final Galactic Plane/Bulge footprint, or the rolling scheme to be implemented on the Galactic Plane. The SCOC will work with the SMWLV and TVS SCs to ascertain the best solutions for Galactic science regarding filter balance and footprint. These decisions should, however, not impact decisions relating to the WFD and the time spent collectively on Galactic regions should not change. Galactic Plane pencil-beam surveys need to be defined more clearly to assess if they would ultimately result in ``nano-surveys'', which will require a fraction of time too small to be optimized at this stage, or to evaluate the possibility of incorporating them into a final Galactic Footprint recommendation''.} @@ -201,7 +208,7 @@ \subsubsection{Footprint and Time Distribution of Visits}\label{sec:subG:footpri Extensive work has already led to the present division of the dense regions of the Galaxy into a high-visit region that encompasses both a large area around the Bulge and a long, thick strip of the Plane, surrounded by a larger area in the Plane with fewer visits. %This division was needed because the overall survey constraints prevent covering this entire area at WFD depth. The subsequent efforts of the SCOC and scientific community have been focused on refining these choices. -One feature of the \baseline{3.0} (\citetalias{PSTN-055}) survey in the Plane is that it left a high-visit ``blob'' centered around a Galactic longitude of $l=+45$ surrounded by a lower visit area. This resulted from a previous candidate survey design that included high-visit pencil beams\footnote{In a subset of previous candidate survey designs, ``pencil beams" were a series of 20 high-visit single pointings distributed in galactic longitude with the goal of ensuring the survey sampled a range of stellar environments.} at varying Galactic longitudes along the Plane and considered stellar density, but it was not due to any other specific science goal in this region. Visits centered around this high-declination blob would necessarily have to occur at high airmass, and would additionally be separated from other high-visit areas, reducing survey efficiency. +One feature of the \baseline{3.0} survey (\citetalias{PSTN-055}) is that it left a high-visits ``blob'' in the Plane centered around a Galactic longitude of $l=+45$ surrounded by a lower visits area. This resulted from a previous candidate survey design that included high-visit pencil beams\footnote{In a subset of previous candidate survey designs, ``pencil beams" were a series of 20 high-visit single pointings distributed in galactic longitude with the goal of ensuring the survey sampled a range of stellar environments.} at varying Galactic longitudes along the Plane and considered stellar density, but it was not due to any other specific science goal in this region. Visits centered around this high-declination blob would necessarily have to occur at high airmass, and would additionally be separated from other high-visit areas, reducing survey efficiency. {\bf The SCOC recommends redistributing the visits concentrated in the ``blob'' centered around a Galactic longitude of $l=+45$ to cover a low-visit ``barrier'' at $l=+335$ in the Plane and at the border of the Plane and Bulge. This change would give continuous longitude coverage along the Plane from a longitude of $l=+30$ down through $l=+280$ and boost metrics for time-domain science in the Bulge/Plane.} @@ -232,24 +239,24 @@ \subsubsection{Footprint and Time Distribution of Visits}\label{sec:subG:footpri \subsubsection{Galactic Filter Balance}\label{sec:subG:filterbalance} -The filter balance in the Bulge in the current \baseline{3.4} already differs from that used for WFD: the primary difference is fewer visits in $y$, which are redistributed to bluer filters to better optimize Galactic science since $u$ and $g$ are vital for stellar characterization even in the presence of foreground dust. In the WFD, $y$ receives a large number of visits, comparable to the number in $z$ and only slightly less than $r$ or $i$, while $g$ and especially $u$ receive fewer visits. Hence $y$, with its relatively low sensitivity, is the optimal choice for redistribution to bluer bands. +The filter balance in the Bulge in \baseline{3.4} and later \opsim s differs from that used for WFD: the primary difference is fewer visits in $y$, which are redistributed to bluer filters to better optimize Galactic science since $u$ and $g$ are vital for stellar characterization even in the presence of foreground dust. In the WFD, $y$ receives a large number of visits, comparable to the number in $z$ and only slightly less than $r$ or $i$, while $g$ and especially $u$ receive fewer visits. Hence $y$, with its relatively low sensitivity, is the optimal choice for redistribution to bluer bands. -Noting the relatively low sensitivity of the $y$-band, and its resulting negligible reddening advantage over $z$ even in dusty regions, the SCOC considered several simulations that redistributed additional visits in the dense regions of the Galaxy from $y$ to a combination of $z$, $g$, and $u$, while still recognizing the fundamental discovery potential of a multi-filter survey over a broad contiguous area. The main finding from these new simulations was that most existing metrics showed mixed or marginal changes, even where the relative number of visits in $u$ and $g$ substantially increased. The metrics considered included Galactic transients, young stars, detection of several classes of periodic variables, light curve gaps, as well as solar system metrics (since the ecliptic passes through this region). +Noting the relatively low sensitivity of the $y$-band, and its resulting negligible reddening advantage over $z$ even in dusty regions, the SCOC considered several simulations that redistributed {\it additional} visits in the dense regions of the Galaxy from $y$ to a combination of $z$, $g$, and $u$, while still recognizing the fundamental discovery potential of a multi-filter survey over a broad contiguous area. The main finding from these new simulations was that most existing metrics showed mixed or marginal changes, even where the relative number of visits in $u$ and $g$ substantially increased. The metrics considered included Galactic transients, young stars, detection of several classes of periodic variables, light curve gaps, as well as Solar System metrics (since the ecliptic passes through this region). -{\bf The SCOC finds that the adoption of a revised filter balance in the Bulge and Plane with less $y$ and more $z$, $g$, and $u$ compared to the present baseline is potentially beneficial on the net, but that existing metrics are not adequately sensitive to the explored filter balance changes for some expected science cases. The SCOC concludes that a survey using the currently implemented filter balance in the Bulge and Plane in \baseline{3.4} will produce excellent science and the LSST can start with this implementation.} +{\bf The SCOC finds that the adoption of a revised filter balance in the Bulge and Plane with less $y$ and more $z$, $g$, and $u$ compared to the present baseline is potentially beneficial on the net, but that existing metrics are not adequately sensitive to the explored filter balance changes for some expected science cases. The SCOC concludes that a survey using the filter balance implemented in the Bulge and Plane in \baseline{3.4} will produce excellent science and the LSST can start with this implementation.} However, the SCOC also welcomes input from the community whose science is affected by the details of filter balance in the dense regions of the Galaxy to help define improved metrics that could lead to further optimization in future years. \subsubsection{The LMC/SMC and South Celestial Pole}\label{sec:subG:specialregions} The scientific goals of the survey in the region of the LMC and SMC (together MCs) \footnote{There is an effort underway avoid using the current full name of the MCs, as reasoned in \url{https://physics.aps.org/articles/v16/152}. We adopt the acronyms LMC/SMC without expanding them into the full name here to reflect the broad and inclusive reach of Rubin LSST.} -and South Celestial Pole (SCP) differ somewhat from WFD. In particular, the major areas of focus of the survey in the main bodies of the LMC and SMC are microlensing and other variable/transient science. In the peripheries of the MCs, including the South Celestial Pole region, the central goal is to detect dwarf satellites and other low-surface-brightness stellar substructures. +and South Celestial Pole (SCP) differ somewhat from WFD. In particular, the major areas of focus of the survey in the main bodies of the LMC and SMC are microlensing and other variable/transient science. In the peripheries of the MCs, including the SCP region, the central goal is to detect dwarf satellites and other low-surface-brightness stellar substructures. These goals are supported in the current baseline, as the MCs are covered with the same number of visits as the WFD, while the SCP region, only observable at relatively high airmass, has a low number of total visits, but sufficient to detect many potential dwarf satellites and substructures. However, the current baseline also adopts the WFD filter baseline in the MC and SCP regions, which may not be ideal for the stated goals. -A number of simulations were considered that used an alternate filter balance for both the Clouds and the South Celestial Pole, moving visits out of $z$/$y$ and toward $u$/$g$ in both regions. These simulations show large improvements in metrics relevant to the detection of low surface-brightness dwarfs as well as some improvements in microlensing and variable star/transient metrics. +A number of simulations were considered that used an alternate filter balance for both the MCs and the SCP, moving visits out of $z$/$y$ and toward $u$/$g$ in both regions. These simulations show large improvements in metrics relevant to the detection of low surface-brightness dwarfs as well as some improvements in microlensing and variable star/transient metrics. -{\bf The SCOC recommends a bluer filter mix in these regions, bounded by the requirement that the increased number of dark-time visits in a relatively narrow range of right ascension does not affect other areas of the survey.} +{\bf The SCOC recommends a bluer filter mix in these regions, bounded by the requirement that the increased number of dark-time visits in a relatively narrow range of right ascension does not affect other areas of the LSST survey.} \begin{figure} \centering @@ -282,14 +289,16 @@ \subsection{Targets of Opportunity (ToO)}\label{sec:ToO} {[\citetalias{PSTN-055} \S2.8] ``[...] be contained to $\leq$3\% of the LSST time. The SCOC recommends that Rubin organizes a workshop in 2023 to bring together members of the scientific community, members of Rubin Observatory (including observing and scheduler specialists, and Data Management specialists), and members of the SCOC to define the details of the implementation of the Rubin ToO program. This workshop should produce a document detailing recommendations for implementation, including suggestions for the questions outlined above, that the experts agree would accomplish the scientific goals of the program.''} \end{quote} -A meeting was organized in March 2024 (Rubin ToO 2024\footnote{\url{https://lssttooworkshop.github.io/images/Rubin_2024_ToO_workshop_final_report.pdf}}) with the explicit purpose of making a community recommendation for a Rubin ToO program within bounds previously established by the SCOC. After evaluating this community consensus report and considering simulations of its implementation, the SCOC finds that the impact on WFD science is generally small and that the proposed ToO programs have the potential to lead to important scientific results. +A meeting was organized in March 2024 (Rubin ToO 2024\footnote{\url{https://lssttooworkshop.github.io/images/Rubin_2024_ToO_workshop_final_report.pdf}}) with the explicit purpose of making a community recommendation for a Rubin ToO program within the bounds previously established by the SCOC. After evaluating this community consensus report and considering simulations of its implementation, the SCOC finds that the impact on WFD science is generally small and that the proposed ToO programs have the potential to lead to important scientific results. {\bf The SCOC recommends the implementation of a LSST ToO program as detailed in the community report Rubin ToO 2024: Envisioning the Vera C. Rubin Observatory LSST Target of Opportunity program\footnote{\url{https://docs.google.com/document/d/1WE4NGl3dFOVGo7lzpyG1fe_JiX9m-kLl5JYQkhu9iso/edit?usp=sharing}} (hereafter RubinToO2024) by the scientific community at large}. +RubinToO24 identified several different classes of ToOs for which Rubin's observations are well-justified. The vast majority of ToOs will be to follow up gravitational wave (GW) events, while a much smaller number of neutrino and Solar System ToOs are expected. +The report includes ToO follow-up plans for GW requiring \mbox{$\sim$85\%} of the ToO time, neutrino counterparts taking \mbox{$\sim$5\%}, and small Potentially Hazardous Asteroids (PHAs) taking \mbox{$\sim$10\%} of the ToO time. -This report includes ToO follow-up plans for Gravitational Wave counterparts (GW, \mbox{$\sim$85\%} of the ToO time), neutrino counterparts (\mbox{$\sim$5\%} of the ToO time, and small Potentially Hazardous Asteroids (PHAs, \mbox{$\sim$10\%} of the ToO time). The impact of including a ToO program as recommended in RubinToO2024 is shown in \autoref{fig:too}. In the current implementation, the program takes between 3\% and 4\% of the survey time. While this is slightly in excess of the recommendation in \citetalias{PSTN-055}, we are still improving the efficiency of the program's implementation, and the current implementation likely represents an upper limit as no triggered sequence is terminated due to reclassification of the event and/or as the transient is identified. -We note that metrics that are very sensitive to the number of WFD observations collected, like SNIa cosmology and Kilonova discovery MAFs suffer a \mbox{$\sim$5\%} impact. However, the SCOC holds that the potential for discovery of KN counterparts of MMA triggers, and by the promise of KN counterparts of gravitational waves as cosmological probes \citep[\eg ,][]{PhysRevResearch.2.022006, gianfagna2024potential} compensate for this loss. We further note that the data collected within the ToO program, with a denser cadence and deeper images, can result in an effective dataset for the study of fast transients alternative to the WFD data. A negative impact is also seen in some Solar System metrics in \autoref{fig:too}. However, the core Solar System metrics do not suffer from the introduction of the ToO program which, as a reminder, while dominated by GW follow up will be in part used for Solar System objects. +The impact of including a ToO program as recommended in RubinToO2024 on science and system metrics is shown in \autoref{fig:too}. In the current implementation, the program takes between 3\% and 4\% of the survey time. While this is slightly in excess of the recommendation in \citetalias{PSTN-055}, we are still improving the efficiency of the program's implementation, and the current implementation likely represents an upper limit as no triggered sequence is terminated due to reclassification of the event and/or as the transient is identified. +We note that metrics that are very sensitive to the number of WFD observations collected, like SNIa cosmology and Kilonova discovery MAFs suffer a few \% impact. However, the SCOC holds that the potential for discovery of KN counterparts of MMA triggers, and by the promise of KN counterparts of gravitational waves as cosmological probes \citep[\eg ,][]{PhysRevResearch.2.022006, gianfagna2024potential} compensate for this loss. We further note that the data collected within the ToO program, with a denser cadence and deeper images, can result in an effective dataset for the study of fast transients alternative to the WFD data. A negative impact is also seen in some Solar System metrics in \autoref{fig:too}. However, the core Solar System metrics do not suffer from the introduction of the ToO program which, as a reminder, while dominated by GW follow up will be in part used for Solar System objects. \begin{figure} \centering @@ -297,27 +306,25 @@ \subsection{Targets of Opportunity (ToO)}\label{sec:ToO} \caption{The impact of the inclusion of the ToO program on static (left) and transient and variable (right) LSST science. Note that the marginal negative impact on the number of well-characterized SNIa (\texttt{SNIa N} MAF on the right) and identifiable Kilonovae (\texttt{KNe-} MAFs in the same plot) in the WFD is compensated, respectively, by the potential for the discovery of KN counterparts of MMA triggers, and by the promise of KN counterparts of gravitational waves as cosmological probes \citep[\eg ,][]{PhysRevResearch.2.022006, gianfagna2024potential}.} \label{fig:too} \end{figure} -This report identified several different classes of ToOs for which Rubin's observations are well-justified. The vast majority of ToOs will be to follow up gravitational wave (GW) events, while a much smaller number of neutrino and solar system ToOs are expected. -The current LIGO-Virgo-KAGRA (LVK) GW observing run (Observing Run 4 or O4) will end before the start of LSST. Hence, GW ToOs will not commence until the start of the Observing Run 5 (O5) of the LVK detectors. We note that the start time of O5 has no expected impacts on the LSST WFD or the ToO program. Improved system performance, primarily afforded by the consistent working of three detectors (with similar sensitivity), will maximize the scientific productivity of the Rubin ToO program while reducing the impact on other programs. Two working LIGO detectors at their design sensitivity, combined with a third detector working at 30-50\% that of LIGO, will reduce the skymaps to tractable sizes for rapid Rubin coverage. We encourage the LVK science collaboration and the International Gravitational Wave Network (IGWN), to prioritize a high-performing system with three working detectors over an early start of the O5 run. +The current LIGO-Virgo-KAGRA (LVK) GW observing run (Observing Run 4 or O4) will end before the start of LSST. Hence, GW ToOs will not commence until the start of the Observing Run 5 (O5) of the LVK detectors. We note that the start time of O5 has no expected impacts on the LSST WFD or the ToO program. Improved system performance, primarily afforded by the consistent working of three detectors (with similar sensitivity), will maximize the scientific productivity of the Rubin ToO program while reducing the impact on other programs. Two working LIGO detectors at their design sensitivity, combined with a third detector working at 30-50\% that of LIGO, will reduce the skymaps to tractable sizes for rapid Rubin coverage. We encourage the LVK science collaboration and the International Gravitational Wave Network (IGWN), to prioritize a high-performing system with three working detectors over an early start of the O5 run. As the GW component of the ToO program takes the largest amount of time and has the most impact on WFD, to enable optimal use of Rubin resources, {\bf the SCOC recommends that a meeting to follow Rubin ToO 2024 be organized closer to the start of O5 to refine the GW follow-up survey strategy with improved knowledge of the expected performance of the GW detector networks and systems in O5 and of the performance of the full Rubin system.} % -There is no comparable time restriction for the solar system ToO program (to follow up potentially hazardous asteroids) or the neutrino ToO program (to follow-up high-energy neutrinos or those from a Galactic supernova). Hence, -%{\bf the SCOC recommends that the solar system and neutrino ToOs should start as soon as possible: as soon as suitable templates are available.} -%KV suggests replacing line above with: -\textbf{ the SCOC recommends that the solar system and neutrino ToOs should start as soon as possible.} This would be as soon as suitable templates are available for neutrino ToOs and after enough time to assess both the PHA impactor false positive and event rate with the influx of Rubin discoveries (which RubinToO2024 estimated will take \mbox{$\sim$3} months). +There is no comparable time restriction for the Solar System ToO program (to follow up PHAs) or the neutrino ToO program (to follow-up high-energy neutrinos or those from a Galactic supernova). Hence, +\textbf{ the SCOC recommends that the Solar System and neutrino ToOs should start as soon as possible.} This would be as soon as suitable templates are available for neutrino ToOs and after enough time to assess both the PHA impactor false positive and event rate with the influx of Rubin discoveries (which RubinToO2024 estimated will take \mbox{$\sim$3} months). For all ToOs, to enable ToO response from the Rubin system, a high level of automation is required. For each potential ToO, a response shall be predetermined algorithmically, including which targets Rubin responds to and the sequence of observations, based on the transient’s characteristics. Informal systems can easily lead to mistakes. For this reason, {\bf the SCOC recommends that Rubin only consider potential ToOs that emanate from vetted discovery and distribution systems that produce and dispatch fully machine-readable alerts.\footnote{At the time of writing, the SCOC understands that full automation is not currently in place for all IceCube neutrino triggers.}} The SCOC considers the current list of vetted systems to be: LIGO-Virgo-KAGRA (gravitational waves); IceCube (neutrinos); SNEWS (neutrinos); JPL Scout or Sentry for potentially hazardous asteroids. The SCOC will evaluate future systems for inclusion in this list (\eg , a new neutrino observatory) on formal request. Human input may still be required to evaluate in real time the value of a ToO trigger and the specific response. One (or more) Rubin members %(assigned on any given night from a group of Rubin Scientists) -should review triggers and be allowed to, if desired, overwrite the algorithmic decision to pursue/not pursue a ToO or interrupt the ToO observing sequence. Further, to ensure that appropriate expertise is available, an Advisory Committee of community members can interact and advise the observer in real-time, with communication initiated either by the committee or by the observer. -This Advisory Committee should be composed of community members and have a nomination-selection process (including self-nomination) to be outlined in detail before the start of survey operations, ensuring broad coverage of scientific competence in all areas relevant to the ToO program and diversity along all relevant axes. -The committee, observers, and Rubin leadership will review the ToO outcomes post factum to advise on program changes. -{\bf The SCOC recommends real-time human review of potential ToO triggers and the establishment of a Rubin ToO Advisory Committee as described above.} The Advisory Committee should be composed of community members and, collectively, have relevant expertise on all ToO science cases (Solar System, Neutrino, Gravitational Waves, and any science case that may be added to the program in the future). The Advisory Committee should be empowered to propose changes to the observing strategy based on the outcomes of the program and scientific developments. The SCOC will also solicit and consider feedback on the implementation of these ToO programs as necessary to ensure they meet the science goals outlined in the community ToO report. +should review triggers and be allowed to, if desired, overwrite the algorithmic decision to pursue/not pursue a ToO or interrupt the ToO observing sequence. Further, to ensure that appropriate expertise are available, the program should be supported by the establishment of an Advisory Committee that can interact and advise the observer in real-time, with communication initiated either by the committee or by the observer. +This Advisory Committee should be composed of community members, collectively have relevant expertise on all ToO science cases (Solar System, neutrino, GW, and any science case that may be added to the program in the future) and have a nomination-selection process (including self-nomination) to be outlined in detail before the start of survey operations, ensuring broad coverage of scientific competence in all areas relevant to the ToO program and diversity along all relevant axes. +{\bf The SCOC recommends real-time human review of potential ToO triggers and the establishment of a Rubin ToO Advisory Committee as described above.} + +The committee, observers, and Rubin leadership will review the ToO outcomes post factum to advise on program changes. The Advisory Committee should be empowered to propose changes to the observing strategy based on the outcomes of the program and scientific developments at any time. The SCOC will also solicit and consider feedback on the implementation of these ToO programs as necessary to ensure they meet the science goals outlined in the community ToO report. @@ -330,7 +337,7 @@ \subsection{Snaps}\label{sec:snaps} Saturation limits will be slightly higher but this will only impact a small number of objects compared to the large volume of sources in the LSST universe. Other surveys are better equipped to work with those targets that are too bright for LSST. -Some science cases (\eg\ Cataclysmic Variables and Flares, very fast moving solar system objects) could benefit from the separate exposures, but the planned data processing for the individual snap images is more limited than that applied to the combined visit, so these science cases would need to rely on pipelines contributed by the community and user-generated data products. Additionally, for these cases too, other surveys are better equipped to work within those time scales. +Some science cases (\eg\ Cataclysmic Variables and flares, or very fast-moving Solar System objects) could benefit from the separate exposures, but the planned data processing for the individual snap images is more limited than that applied to the combined visit, so these science cases would need to rely on pipelines contributed by the community and user-generated data products. Additionally, for these cases too, other surveys are better equipped to work within those time scales. Thus, the SCOC does not see scientific opportunities associated with retaining the two 15s snaps that can compete with the 7-9\% gain in survey efficiency. @@ -382,12 +389,12 @@ \subsection{Deep Drilling Fields (DDF)}\label{sec:DDF} \FloatBarrier -\subsection{Early Science}\label{sec:early} +\subsection{Early Survey}\label{sec:early} % \tbd{The SCOC recommends implementing a detailed coordination plan with the Early Science Rubin team to reach a final recommendation on the strategy to be implemented in the first year of the survey, including a scheme for the construction of templates.} The SCOC emphasizes that the priority in Y1 of operations should be obtaining a dataset that supports and facilitates science throughout the survey. This includes a dataset sufficient for calibration across the \mbox{$\sim$20,000} square degrees of the WFD, including images at different airmasses, illuminations, field crowdedness, etc. - The SCOC supports Rubin's commitment to acquiring incremental templates throughout Y1 to begin dispatching alerts (via the Alert Brokers) and encourages the Observatory to release alerts as early as possible. The SCOC reviewed the Alert Production team's proposal to prioritize timeliness over the quality of templates and build templates from fewer images ($\geq 3$) in Y1 than in subsequent years. Releasing some alerts in Y1 is an important goal to enable the time domain and solar system science communities to prepare for the full-volume, full-fidelity alert streams to come in subsequent years, as well as increasing the discovery potential of LSST in early operations. Earlier template generation is particularly important for testing solar system alert streams that require post-discovery re-detections of solar system objects. + The SCOC supports Rubin's commitment to acquiring incremental templates throughout Y1 to begin dispatching alerts (via the Alert Brokers) and encourages the Observatory to release alerts as early as possible. The SCOC reviewed the Alert Production team's proposal to prioritize timeliness over the quality of templates and build templates from fewer images ($\geq 3$) in Y1 than in subsequent years. Releasing some alerts in Y1 is an important goal to enable the time domain and Solar System science communities to prepare for the full-volume, full-fidelity alert streams to come in subsequent years, as well as increasing the discovery potential of LSST in early operations. Earlier template generation is particularly important for testing Solar System alert streams that require post-discovery re-detections of Solar System objects. However, this goal should not overwrite the priority of obtaining a fully calibrated system by the end of Y1. {\bf The SCOC recommends that the filter balance is adjusted as needed in Y1 to acquire a sufficient number of $u$-band images for calibration (and template construction). } diff --git a/baseline4.tex b/baseline4.tex index 4b02826..802ca2d 100644 --- a/baseline4.tex +++ b/baseline4.tex @@ -13,17 +13,17 @@ \section{The \texttt{v4.0} simulations} \label{sec:baseline4_0} In all of these regions on the sky, visits within a night are attempted in pairs, with each first visit to a pointing paired with a return visit typically 33 minutes later in a different filter. This provides the opportunity to measure colors for transient or variable objects. Visits are paired as follows: $u+g$ or $u+r$, $g+u$ or $g+r$, $r+u$ or $r+i$, $i+r$ or $i+z$, $z+i$ or $z+y$, $y+z$ or $y+y$ (\citetalias{PSTN-053} \S3). During twilight, the interval between these pairs may be shorter or visits may be scheduled as singles instead of paired. Approximately 4\% of all visits are part of a triplet of visits, instead of just a pair. In these cases, the first pair is followed several hours later by a third visit, in a filter matching the earlier pair. This enables probing short timescale variability, although at the cost of increased slew time and obtaining observations at higher airmass (\citetalias{PSTN-055} \S2.4). -Within the low-dust WFD footprint, a rolling cadence is followed. This rolling cadence distributes visits unevenly across seasons; in some seasons, a portion of the footprint will receive more than the typical number of visits while, in other seasons, the same portion of the footprint will receive fewer than the typical number of visits. The exact details of how many more or less visits are received in alternating seasons depends on the `strength' of the rolling cadence as well as the typically number of visits in a season and the minimum number of visits in any season needed in order to continue creating templates for difference imaging (\citetalias{PSTN-055} \S2.5). In \baseline{4.0}, the high and low seasons correspond to 125 and 25 visits, with a typical season of 75 visits. The low-dust WFD region is split into four declination bands, two of which are active in a high season, while two are in a low season at any point when rolling cadence is active at that point in the sky (\citetalias{PSTN-055} \S2.5). A `cycle' of rolling consists of two seasons, so that both a high and low cadence season can occur. In \baseline{4.0}, there are three cycles of rolling cadence at each point on the sky, with a uniform (non-rolling) season in between each of these seasons. This `uniform rolling' cadence provides the opportunity for maximum uniformity at intermediate data releases in year 4 and 7 (\autoref{sec:rolling}). +Within the low-dust WFD footprint, a rolling cadence is followed. This rolling cadence distributes visits unevenly across seasons; in some seasons, a portion of the footprint will receive more than the typical number of visits while, in other seasons, the same portion of the footprint will receive fewer than the typical number of visits. The exact details of how many more or less visits are received in alternating seasons depends on the `strength' of the rolling cadence as well as the typically number of visits in a season and the minimum number of visits in any season needed in order to continue creating templates for difference imaging (\citetalias{PSTN-055} \S2.5). In \baseline{4.0}, the high and low seasons correspond to 125 and 25 visits, with a typical season of 75 visits. The low-dust WFD region is split into four declination bands, two of which are active in a high season, while two are in a low season at any point when rolling cadence is active at that point in the sky (\citetalias{PSTN-055} \S2.5). A `cycle' of rolling consists of two seasons, so that both a high and low cadence season can occur. In \baseline{4.0}, there are three cycles of rolling cadence at each point on the sky, with a uniform (non-rolling) season in between each of these seasons. This `Uniform Rolling' cadence provides the opportunity for higher uniformity at intermediate data releases in year 4 and 7 (\autoref{sec:rolling}). -The filter balance within different areas of the footprint varies. Per pointing, the low-dust WFD obtains a median of 54 visits in $u$ band, 66 visits in $g$ band, 174 in $r$ and 176 in $i$ bands, and 158 in $z$ and 155 in $y$ bands (\autoref{sec:filterbalance}). The Galactic Plane WFD region tilts the balance toward fewer $u$ and $y$ band visits and more $g$ band (\autoref{sec:subG:filterbalance}), while the LMC/SMC region gets fewer $z$ and $y$ band visits to obtain more $u$ and $g$ band visits (\autoref{sec:subG:specialregions}). The dusty plane region focuses on $g$, $r$, $i$, and $z$, with fewer visits in $u$ and $y$. The NES only obtains visits in $g$, $r$, $i$, and $z$ band as these are the most useful for Solar System Objects (\citetalias{PSTN-053} \S3). The SCP region uses its fewer visits per pointing with a higher fraction of $g$, $r$ and $i$ band visits. +The filter balance within different areas of the footprint varies. Per pointing, the low-dust WFD obtains a median of 54 visits in $u$ band, 66 visits in $g$ band, 174 in $r$ and 176 in $i$ bands, and 158 in $z$ and 155 in $y$ bands (\autoref{sec:filterbalance}). The Galactic Plane WFD region tilts the balance toward fewer $u$ and $y$ band visits and more $g$ band (\autoref{sec:subG:filterbalance}), while the LMC/SMC region gets fewer $z$ and $y$ band visits to obtain more $u$ and $g$ band visits (\autoref{sec:subG:specialregions}). The dusty plane region focuses on $g$, $r$, $i$, and $z$, with fewer visits in $u$ and $y$. The NES only obtains visits in $g$, $r$, $i$, and $z$ band as these are the most useful for Solar System objects (\citetalias{PSTN-053} \S3). The SCP region uses its fewer visits per pointing with a higher fraction of $g$, $r$ and $i$ band visits. -The Near-Sun Twilight Microsurvey is implemented in Y1 because it is at risk of interference by satellite constellations in later years (\citetalias{PSTN-055} \S2.7.2). It runs every fourth night at both evening and morning twilight, obtaining quads of short (15~seconds) visits in $g$, $r$, $i$ and $z$ bands at high airmass towards the Sun. These low solar elongation visits permit detection of interior-to-earth asteroids. %Additional microsurveys will be considered by the SCOC, to be added as survey observing time permits. +The Near-Sun Twilight microsurvey is implemented in Y1 because it is at risk of interference by satellite constellations in later years (\citetalias{PSTN-055} \S2.7.2). It runs every fourth night at both evening and morning twilight, obtaining quads of short (15~seconds) visits in $g$, $r$, $i$ and $z$ bands at high airmass towards the Sun. These low solar elongation visits permit detection of interior-to-earth asteroids. %Additional microsurveys will be considered by the SCOC, to be added as survey observing time permits. -Target of Opportunity observations are simulated in \baseline{4.0}, matching the 2024 Rubin ToO Workshop outcomes and SCOC recommendations (\autoref{sec:ToO}). The bulk of these ToO visits correspond to followup of multi-messenger events during the estimated dates of LKV O5. Additional ToO visits are dedicated to following Solar System and neutrino triggers. The fraction of on-sky time spent in ToOs in \baseline{4.0} is between 3 and 4\%. +Target of Opportunity observations are simulated in \baseline{4.0}, matching the 2024 Rubin ToO Workshop outcomes and SCOC recommendations (\autoref{sec:ToO}). The bulk of these ToO visits correspond to followup of GW events during the estimated dates of LKV O5. Additional ToO visits are dedicated to following Solar System and neutrino triggers. The fraction of on-sky time spent in ToOs in \baseline{4.0} is between 3 and 4\%. The simulations provided in \texttt{v4.0} include \begin{itemize} \item \texttt{baseline\_v4.0\_10yrs} - the baseline simulation as described above \item \texttt{four\_cycle\_v4.0\_10yrs} - a simulation similar to the one described above, except using four cycles of rolling cadence in the low-dust WFD instead of three cycles interspersed with uniform seasons. - \item \texttt{one\_snap\_v4.0\_10yrs} - a simulation similar to the baseline, but using single exposures for all visits instead of two exposures per visit in $g$, $r$, $i$, $z$ and $y$ bands. This improves the duty cycle of observing and increases the efficiency of the survey. + \item \texttt{one\_snap\_v4.0\_10yrs} - a simulation similar to the baseline, but using single exposures for all visits instead of two exposures per visit in $g$, $r$, $i$, $z$ and $y$ bands. %This improves the duty cycle of observing and increases the efficiency of the survey. \end{itemize} diff --git a/intro.tex b/intro.tex index e8c86e5..4724cf3 100644 --- a/intro.tex +++ b/intro.tex @@ -1,12 +1,13 @@ \section{Introduction} With an unprecedented engagement of the scientific community at large, the Vera C. Rubin Observatory (hereafter Rubin) has designed a process of incremental improvements to the survey strategy to maximize the overall scientific throughput of the Legacy Survey of Space and Time (LSST). The high-level requirements for the LSST are set by four science pillars: probing dark energy and dark matter, building an unprecedented inventory of the Solar System, mapping the Milky Way and Local Volume, and exploring the transient universe. - These requirements are described in \cite{LPM-17} ---hereafter the Science Requirements Document, or \citetalias{LPM-17}---, but significant flexibility remains in survey cadence within these requirements. The optimization of the survey strategy process is aimed at maximizing science for the four science pillars and increasing the portfolio of LSST science by tuning the survey strategy and cadence within the SRD requirements (\citetalias{LPM-17}). -As part of this process, the Survey Cadence Optimization Committee (SCOC) was set up by Rubin's Science Advisory Committee in 2018 to solicit, review, and integrate community feedback and make recommendations for the implementation of the LSST survey strategy to the Director of Operations. This document constitutes the third SCOC recommendation, resulting from the Phase 3 process of survey design which started in January 2023, after the delivery of the Phase 2 recommendation (\citealt{PSTN-055} ---hereafter \citetalias{PSTN-055}--- and the baseline simulation \texttt{baseline\_v3.0}. \cite{PSTN-056} (this document) is planned to be the last recommendation for the LSST as a whole before the start of LSST. However, the SCOC will refine the plan for Y1 in particular and the LSST in general in light of commissioning outcomes; reviews of the survey strategy will continue throughout the 10-year survey, renewing its recommendation on an annual basis (\autoref{sec:next}). +As part of this process, the Survey Cadence Optimization Committee (SCOC) was set up by Rubin's Science Advisory Committee in 2018 to solicit, review, and integrate community feedback and make recommendations for the implementation of the LSST survey strategy to the Director of Operations. This document constitutes the third SCOC recommendation, resulting from the Phase 3 process of survey design which started in January 2023, after the delivery of the Phase 2 recommendation (\citealt{PSTN-055} ---hereafter \citetalias{PSTN-055}) and the baseline simulation \texttt{baseline\_v3.0}. + +\cite{PSTN-056} (this document) is planned to be the last recommendation for the LSST as a whole before the start of LSST. However, the SCOC will refine the plan for Y1 in particular and the LSST in general in light of commissioning outcomes; SCOC reviews of the survey strategy will continue throughout the 10-year survey, renewing its recommendation on an annual basis (\autoref{sec:next}). -The Phase 3 recommendation (this document) responds directly to the questions left open in Phase 2 (\citetalias{PSTN-055}) and updates and refines previous recommendations (\citetalias{PSTN-055} and \citetalias{PSTN-053}) The present document generally does not reiterate previous recommendations that have not changed. +The Phase 3 recommendation (this document) responds directly to the questions left open in Phase 2 (\citetalias{PSTN-055}) and updates and refines previous recommendations (\citetalias{PSTN-055} and \citetalias{PSTN-053}). The present document generally does not reiterate previous recommendations that have not changed. @@ -38,7 +39,7 @@ \section{Executive Summary of the Phase 3 recommendation} \subsection{Note on how to read the SCOC plots} -The SCOC typically reviews the outcome of metrics measuring system performance and science outcomes built within the Metric Analysis Framework \citep{2014SPIE.9149E..0BJ}, hereafter MAFs. MAFs are reviewed across multiple simulations of the 10-year sequence of LSST observations, each simulation referred to as an \opsim\, to compare scientific performance. More details on the SCOC process are available in \citetalias{PSTN-055} and \citealt{Bianco_2022}. In this document, you will see sky maps measuring quantities by healpixels (\eg, number of visits, or any MAF). The typical sky pixelization that underlies the metric calculations the SCOC reviews is 128 sides healpixels (covering an area \mbox{$\sim$0.2$\mathrm{deg}^2$}), although for particularly computationally intense MAFs this can be turned down to 64 or 32. You will see metric plots in the following styles: +The SCOC typically reviews the outcome of metrics measuring system performance and science outcomes built within the Metric Analysis Framework \citep{2014SPIE.9149E..0BJ}, hereafter MAFs\footnote{see \url{https://zenodo.org/records/10215451}}. MAFs are reviewed across multiple simulations of the 10-year sequence of LSST observations\footnote{produced via the Operations Simulator \url{https://zenodo.org/records/13835841}}, each simulation referred to as an \opsim, to compare scientific performance. More details on the SCOC process are available in \citetalias{PSTN-055} and \citealt{Bianco_2022}. In this document, you will see sky maps measuring quantities by healpixels (\eg, number of visits, or any MAF). The typical sky pixelization that underlies the metric calculations the SCOC reviews is 128 sides healpixels (covering an area \mbox{$\sim$0.2$\mathrm{deg}^2$}), although for particularly computationally intense MAFs this can be turned down to 64 or 32. You will see metric plots in the following styles: diff --git a/next.tex b/next.tex index eed1e50..5fe195b 100644 --- a/next.tex +++ b/next.tex @@ -1,9 +1,8 @@ \section{The future activities of the SCOC and areas of focus leading to and through Year 1}\label{sec:next} -The SCOC is a standing committee of Rubin Observatory, and it will continue its activities through the 10-year duration of LSST. The SCOC will review the survey and issue annual recommendations to the Observatory Director for modifications of the survey strategy in light of scientific outcomes, technical challenges and upgrades, and the evolving scientific landscape. The SCOC will continue to liaise with the scientific community with {\bf annual workshops, liaisons to the SCOC, office hours, and by making its activities public via posts on the LSST Community forum in the Survey Strategy topic} -\footnote{\url{https://community.lsst.org/c/sci/survey-strategy/37}}. As discussed in \autoref{sec:ToO}, the ToO program will be further supported by an Advisory Committee composed of community members that will evaluate the ToO program on an ongoing basis. +The SCOC is a standing committee of Rubin Observatory, and it will continue its activities through the 10-year duration of LSST. The SCOC will review the survey and issue annual recommendations to the Observatory Director for modifications of the survey strategy in light of scientific outcomes, technical challenges and upgrades, and the evolving scientific landscape. The SCOC will continue to liaise with the scientific community with {\bf annual workshops, liaisons to the Science Collaborations, office hours, and by making its activities public via posts on the LSST Community forum in the Survey Strategy topic}\footnote{\url{https://community.lsst.org/c/sci/survey-strategy/37}}. As discussed in \autoref{sec:ToO}, the ToO program will be further supported by an Advisory Committee composed of community members that will evaluate the program on an ongoing basis. -The focus of the SCOC leading up to and into Year 1 will be to review the recommendations shared in \citetalias{PSTN-053}, \citetalias{PSTN-055}, and \citetalias{PSTN-056} in the light of commissioning outcomes and to strategize effective plans for Y1 data collection, including templates, and integrating the data that will be collected before the start of LSST into its recommendations. The deliberations on this topic will necessarily be fluid and evolve rapidly as the commissioning and science verification phases of LSST progress. The SCOC continues to solicit recommendations from the community about scientific prioritization in the collection of templates while restating that the priority in Y1 of operations should be obtaining a dataset that supports and facilitates science throughout the survey. In practice, this means collecting a dataset sufficient for calibration across the \mbox{$\sim$20,000} square degrees of the WFD, including images at different airmasses, illuminations, field crowdedness, etc (\autoref{sec:additional}). +The focus of the SCOC leading up to and into Y1 will be to review the recommendations shared in \citetalias{PSTN-053}, \citetalias{PSTN-055}, and \citetalias{PSTN-056} in the light of commissioning outcomes and to strategize effective plans for Y1 data collection, including templates acquisition and integrating the data that will be collected before the start of LSST into its recommendations. The deliberations on this topic will necessarily be fluid and evolve rapidly as the commissioning and science verification phases of LSST progress. The SCOC continues to solicit recommendations from the community about scientific prioritization in the collection of templates while restating that the priority in Y1 of operations should be obtaining a dataset that supports and facilitates science throughout the survey. In practice, this means collecting a dataset sufficient for calibration across the \mbox{$\sim$20,000} square degrees of the WFD, including images at different airmasses, illuminations, field crowdedness, etc (\autoref{sec:additional}). As discussed in \autoref{sec:rolling}, the SCOC will continue to study the impact of the new rolling implementations and of the number of rolling cycles on time-domain science, uniformity of coadds for cosmology and extragalactic science, and all LSST science in general. These investigations will include considerations on the outcomes of Y1 itself as the first year is underway and on the results of the Data Management assessments of feasibility and cost of adding uniform data releases at key years for cosmological analysis (DR5 and DR8). We expect to release a recommendation in the second half of Y1 (likely \mbox{$\sim$2} months before the start of Y2) including a recommendation on rolling implementations. @@ -12,11 +11,15 @@ \section{The future activities of the SCOC and areas of focus leading to and thr {[\citetalias{PSTN-055} \S4] The SCOC recommends that two microsurveys be scheduled in Y1: the near-sun NEO twilight survey and, if time is available, the Northern Strip survey. Additional microsurveys should be added in the future, when the system characteristics and survey efficiency are better assessed, and a process is recommended to receive and review refined and additional microsurvey proposals after the beginning of LSST.} \end{quote} -The community is best placed to write effective proposals for nano- and micro-microsurveys (<0.3\% and between 3\% and 0.3\% of the LSST observing time respectively), and the SCOC best placed to evaluate them, when the capabilities of the system-as-built are estimated with on-sky data. That is, we expect proposals for nano- and micro-survey will be more compelling and realistic after DR1. As a reminder, DR1 will include the first six-months of LSST data, and it will be released \mbox{$\sim$1} year after the start of the survey. An opportunity will be provided to propose timely nano- and micro-surveys before the end of Y1 for science cases of justifiable urgency that require data collected in Y2. This proposal call is expected to be issued no earlier than six months after the start of LSST (when the data for DR1 are collected) with a likely deadline of nine months from the start of LSST; this will allow the SCOC to review and possibly recommend proposals for Y2 implementation. Proposals for nano- and micro-surveys will continue to be solicited and reviewed through the LSST on a regular basis. +The community is best placed to write effective proposals for nano- and micro-microsurveys (<0.3\% and between 0.3\% and 3\% of the LSST observing time respectively), and the SCOC best placed to evaluate them, when the capabilities of the system-as-built are estimated with on-sky data. That is, we expect proposals for nano- and micro-survey will be more compelling and realistic after DR1. As a reminder, DR1 will include the first six-months of LSST data, and it will be released \mbox{$\sim$1} year after the start of the survey. An opportunity will be provided to propose timely nano- and micro-surveys before the end of Y1 for science cases of justifiable urgency that require data collected in Y2. This proposal call is expected to be issued no earlier than six months after the start of LSST (when the data for DR1 are collected) with a likely deadline of nine months from the start of LSST; this will allow the SCOC to review and possibly recommend proposals for Y2 implementation. Proposals for nano- and micro-surveys will continue to be solicited and reviewed through the LSST on a regular basis. In the past several years, the SCOC has received and considered feedback in the form of community and Science Collaboration reports, communications with the liaisons to the Science Collaborations, discussions held in the SCOC Office Hours, and at the annual SCOC workshops. Feedback from the community will always be welcome and encouraged throughout LSST. The modalities of feedback may evolve in Operations, but communication with the SCOC via the channels mentioned above is planned to continue. Rubin Observatory and the SCOC are infinitely grateful for the continuing contributions of the community to the design of the LSST Survey Strategy. The progress made in the past 10 years has led to important expected science gains across all science areas, as demonstrated by the significant improvements in science metrics built by the community. The unprecedented involvement of the scientific community at large in the refinement of the LSST survey strategy has been and continues to be a formidable success and a shining example of constructive collaborative practices in the scientific community. -Rubin Observatory is grateful for the work of the SCOC members who, over the last four years volunteered their service to the Observatory and to the scientific community at large. +The SCOC acknowledges with gratitude the Survey Strategy team who supported the SCOC work and the community contribution to the optimization of the Rubin LSST survey strategy: Dr. Peter Yoachim, Dr. Eric Neilson, and Dr. Lynne Jones. + +Rubin Observatory is grateful for the work of the SCOC members who, over the last four years volunteered their service to the Observatory and to the scientific community at large and in particular express their gratitude to the outgoing members of the committee: Dr. Franz Baur, Dr. Knut Olsen, Dr. Colin Slater, and Dr. Jay Strader. + + \ No newline at end of file diff --git a/summaryrecommendation.tex b/summaryrecommendation.tex index e67a492..e11565c 100644 --- a/summaryrecommendation.tex +++ b/summaryrecommendation.tex @@ -6,69 +6,72 @@ \section{Summary of SCOC Phase 3 recommendations}\label{sec:summary} \renewcommand{\labelenumi}{\roman{enumi}.} %[i/] -\item The SCOC recommends swapping $u$- and $y$-band according to the moon phase. Having the $z$ filter always available produces benefits for SN cosmology while preserving coverage on short timescales. This recommendation is implemented starting in \baseline{3.2} (\autoref{sec:filterswap}). +\item FILTER SWAPS: The SCOC recommends swapping $u$- and $y$-band according to the moon phase. Having the $z$ filter always available produces benefits for SN cosmology while preserving coverage on short timescales. This recommendation is implemented starting in \baseline{3.2} (\autoref{sec:filterswap}). -\item All three mirrors in the system will now be coated in silver, which increases throughput in all bands bluer than $u$ but decreases $u$ throughput (by $\lesssim30\%$). Hence, the SCOC recommends an increase of the exposure time in $u$-band to 38 seconds per visit and an increase of the number of $u$-band visits of 10\% compared to \baseline{3.0}. To compensate for the additional time dedicated to $u$-band the SCOC recommends decreasing the exposure time in the other bands. Simulations show that this corresponds to a decrease of 0.8 second exposure time per visit in all other bands (\autoref{sec:filterbalance}). +\item FILTER BALANCE: All three mirrors in the system will now be coated in silver, which increases throughput in all bands bluer than $u$ but decreases $u$ throughput (by $\lesssim30\%$). Hence, the SCOC recommends an increase of the exposure time in $u$-band to 38 seconds per visit and an increase of the number of $u$-band visits of 10\% compared to \baseline{3.0}. To compensate for the additional time dedicated to $u$-band the SCOC recommends decreasing the exposure time in the other bands. Simulations show that this corresponds to a decrease of 0.8 second exposure time per visit in all other bands (\autoref{sec:filterbalance}). -\item {The SCOC confirms the recommendation for rolling in two sky areas at 0.9 strength on the WFD low-dust footprint (\citetalias{PSTN-055} \S2.5). We are adopting the rolling uniform strategy designed by the Uniformity Task Force in \baseline{3.6} simulations, which implements three cycles of rolling, but we will continue to investigate implementations of rolling until our Y1 recommendation because rolling will not begin before the start of Y2 with any of the implementations under consideration. The SCOC recommends that the time domain community, particularly those interested in phenomena that have evolutionary time scales of hours-to-days, urgently quantify the impact of the proposed uniform rolling compared to rolling in four cycles. Further, the SCOC restates the recommendation that Data Management scopes a plan for producing uniform data releases in DR5 and DR8, in addition to the standard data releases, and that the cost of the development and storage of these additional data releases is scoped and shared with the scientific community (\autoref{sec:rolling}).} +\item ROLLING: The SCOC confirms the recommendation for rolling in two sky areas at 0.9 strength on the WFD low-dust footprint (\citetalias{PSTN-055} \S2.5). We are adopting the Uniform Rolling strategy designed by the Uniformity Task Force in \baseline{3.6} simulations, which implements three cycles of rolling, but, because rolling will not begin before the start of Y2 with any of the implementations under consideration, we will continue to investigate three- and four-cycles implementations of rolling until our Y1 recommendation. +The SCOC recommends that the time domain community, particularly those interested in phenomena that have evolutionary time scales of hours-to-days, urgently quantify the impact of the proposed uniform rolling compared to rolling in four cycles. -\item The SCOC concludes that rolling on the Galactic footprint would have a net negative effect on the survey as a whole, and recommends no rolling in the Plane or Bulge (\autoref{sec:subG:footprint}). +\item ROLLING: The SCOC restates its recommendation that Data Management scopes a plan for producing uniform data releases in DR5 and DR8, in addition to the standard data releases. The cost of the development and storage of these additional data and the timing of their release should be scoped and shared with the scientific community \autoref{sec:rolling}). +\item GALAXY: The SCOC concludes that rolling on the Galactic footprint would have a net negative effect on the survey as a whole, and recommends no rolling in the Plane or Bulge (\autoref{sec:subG:footprint}). -\item The SCOC recommends redistributing the visits concentrated in the ``blob'' centered around a Galactic longitude of $l=+45$ (see \autoref{fig:gpfootprint}, top) to cover a low-visit ``barrier'' at $l=+335$ in the Plane and at the border of the Plane and Bulge. This change would give continuous longitude coverage along the Plane from a longitude of $l=+30$ through $l=+280$ and boost metrics for time-domain science in the Bulge/Plane (\autoref{sec:subG:footprint}). -\item The SCOC recommends a visit plan consistent with the \texttt{roman\_v3.3} simulation, capping the number of visits redistributed to the Roman Bulge field at the value (\mbox{$\sim$1,600}) used in this simulation. However, the timing of the implementation of this augmented observing campaign should remain flexible to respond to the final Roman launch date and survey scheduling (\autoref{sec:subG:footprint}). +\item GALAXY: The SCOC recommends redistributing the visits concentrated in the ``blob'' centered around a Galactic longitude of $l=+45$ (see \autoref{fig:gpfootprint}, top) to cover a low-visit ``barrier'' at $l=+335$ in the Plane and at the border of the Plane and Bulge. This change would give continuous longitude coverage along the Plane from a longitude of $l=+30$ through $l=+280$ and boost metrics for time-domain science in the Bulge/Plane (\autoref{sec:subG:footprint}). -\item The SCOC finds that the adoption of a revised filter balance in the Bulge and Plane with less $y$ and more $z$, $g$, and $u$ compared to the present baseline is potentially beneficial for a broad range of science, but that existing metrics are not adequately sensitive to the explored filter balance changes for some expected science cases. The SCOC concludes that a survey using the currently implemented filter balance in the Bulge and Plane in \baseline{3.4} will produce excellent science and the LSST can start with this implementation (\autoref{sec:subG:filterbalance}). +\item GALAXY: The SCOC recommends a visit plan consistent with the \texttt{roman\_v3.3} simulation, capping the number of visits redistributed to the Roman Bulge field at the value used in this simulation (\mbox{$\sim$1,600}). However, the timing of the implementation of this augmented observing campaign should remain flexible to respond to the final Roman launch date and survey scheduling (\autoref{sec:subG:footprint}). -\item The SCOC recommends a bluer filter mix in the SMC, LMC, and SCP regions, bounded by the requirement that the increased number of dark-time visits in a relatively narrow range of right ascension does not affect other parts of the survey (\autoref{sec:subG:specialregions}). +\item GALAXY: The SCOC concludes that a survey using the filter balance bluer than on the WFD in the Bulge and Plane, as implemented in \baseline{3.4}, will produce excellent science and that LSST can start with this implementation. The adoption of a further revised filter balance in the Bulge and Plane with less $y$ and more $z$, $g$, and $u$ is potentially beneficial on the net, but existing metrics are not adequately sensitive to the explored filter balance changes for some expected science cases. -\item The SCOC recommends the implementation of an LSST ToO program as detailed in \emph{Rubin ToO 2024: +\item GALAXY: The SCOC recommends a bluer filter mix in the SMC, LMC, and SCP regions, bounded by the requirement that the increased number of dark-time visits in a relatively narrow range of right ascension does not affect other parts of the LSST survey (\autoref{sec:subG:specialregions}). + +\item ToO: The SCOC recommends the implementation of an LSST ToO program as detailed in \emph{Rubin ToO 2024: Envisioning the Vera C. Rubin Observatory LSST Target of Opportunity program }\footnote{\url{https://docs.google.com/document/d/1WE4NGl3dFOVGo7lzpyG1fe_JiX9m-kLl5JYQkhu9iso/edit?usp=sharing}} by the scientific community at large (\autoref{sec:ToO}). -\item The SCOC recommends that a meeting to follow Rubin ToO 2024 is organized closer to the start of the LKV O5 run (expected 2027) to refine the GW follow-up survey strategy with improved knowledge of the expected performance of the GW detector networks and systems in O5 and of the performance of the full Rubin system. +\item ToO: The SCOC recommends that a meeting to follow Rubin ToO 2024 is organized closer to the start of the LKV O5 run (expected 2027) to refine the GW follow-up survey strategy with improved knowledge of the expected performance of the GW detector networks and systems in O5 and of the performance of the full Rubin system. -\item The SCOC recommends that Solar System and neutrino ToOs start as soon as possible: when templates are available and event rates and false positive rates are appropriately assessed. (\autoref{sec:ToO}). +\item ToO: The SCOC recommends that Solar System and neutrino ToOs start as soon as possible: when templates are available and event rates and false positive rates are appropriately assessed. (\autoref{sec:ToO}). -\item The SCOC recommends that Rubin only consider potential ToOs that emanate from vetted discovery and distribution systems that produce and dispatch fully machine-readable alerts (\autoref{sec:ToO}). +\item ToO: The SCOC recommends that Rubin only consider potential ToOs that emanate from vetted discovery and distribution systems that produce and dispatch fully machine-readable alerts (\autoref{sec:ToO}). -\item The SCOC recommends real-time human review of potential ToO triggers and the establishment of a Rubin ToO Advisory Committee as described above (\autoref{sec:ToO}). +\item ToO: The SCOC recommends real-time human review of potential ToO triggers and the establishment of a Rubin ToO Advisory Committee as described in \autoref{sec:ToO}. -\item The SCOC recommends that, if the technical feasibility is confirmed in commissioning, the survey is conducted with single exposures. With our recommendation of modifying the exposure time for $u$-band to 38 seconds, and compensating for this extra $u$-band survey time by short decrease in exposure across all other bands, the single visits would be \mbox{$\sim$1$\times$29} seconds (\autoref{sec:snaps}). +\item SNAPS: The SCOC recommends that, if the technical feasibility is confirmed in commissioning, the survey is conducted with single exposures. With our recommendation of modifying the exposure time for $u$-band to 38 seconds, and compensating for this extra $u$-band survey time by a small decrease in exposure time across all other bands, the single visits would be \mbox{$\sim$1$\times$29} seconds (\autoref{sec:snaps}). -\item The SCOC recommends that DDF observations should be sequences of multiple WFD-like visits (as opposed to increased exposure times) to allow rapid alert generation (\autoref{sec:DDF}). +\item DDF: The SCOC recommends that DDF observations should be sequences of multiple WFD-like visits (as opposed to increased exposure times) to allow rapid alert generation (\autoref{sec:DDF}). -\item The SCOC recommends that the baseline translational dithering scale of DDF observations should be reduced from 0.7 degrees to 0.2 degrees (with exploration of even smaller translational dithers compatible with instrumental signature removal and calibration needs). (\autoref{sec:DDF}). +\item DDF: The SCOC recommends that the baseline translational dithering scale of DDF observations should be reduced from 0.7 degrees to 0.2 degrees (with exploration of even smaller translational dithers compatible with instrumental signature removal and calibration needs). (\autoref{sec:DDF}). -\item The SCOC recommends that the baseline survey strategy should accommodate varying the nightly depth, filters, or cadence of different DDFs throughout the course of LSST, while maintaining the Phase 2 (\citetalias{PSTN-055} \S2.6) recommendations for the 10-year depth of each field (including the enhanced COSMOS observations to reach 10-year depth in the first 3 years) (\autoref{sec:DDF}). +\item DDF: The SCOC recommends that the baseline survey strategy should accommodate varying the nightly depth, filters, or cadence of different DDFs throughout the course of LSST, while maintaining the Phase 2 (\citetalias{PSTN-055} \S2.6) recommendations for the 10-year depth of each field (including the enhanced COSMOS observations to reach 10-year depth in the first 3 years) (\autoref{sec:DDF}). -\item The SCOC urges the Data Management and Alert Production teams to assess the feasibility of, and resources needed for, enabling nightly co-adds of sequential DDF visits and recommends that a path is developed to enable the creation of these co-adds, subtraction with deep templates, and faint alert generation (with higher latency as needed, \eg , after sunrise) (\autoref{sec:DDF}). +\item DDF: The SCOC urges the Data Management and Alert Production teams to assess the feasibility of, and resources needed for, enabling nightly co-adds of sequential DDF visits and recommends that a path is developed to enable the creation of these co-adds, subtraction with deep templates, and faint alert generation (with higher latency as needed, \eg , after sunrise) (\autoref{sec:DDF}). -\item The SCOC recommends that the filter balance is adjusted as needed in Y1 to acquire a sufficient number of $u$-band images for calibration (and template construction) (\autoref{sec:early}). +\item EARLY SURVEY: The SCOC recommends that the filter balance is adjusted as needed in Y1 to acquire a sufficient number of $u$-band images for calibration (and template construction) (\autoref{sec:early}). -\item The SCOC does not recommend beginning rolling before the end of Y1 to ensure sufficiently uniform sky coverage for cosmological analysis (the DESC expects its first data analysis to be based on DR2), acquire sufficiently good data for sky calibration, and collect a +\item EARLY SURVEY: The SCOC does not recommend beginning rolling before the end of Y1 to ensure sufficiently uniform sky coverage for cosmological analysis (the DESC expects its first data analysis to be based on DR2), acquire sufficiently good data for sky calibration, and collect a sufficiently complete set of templates across the sky (\autoref{sec:early}). -\item The SCOC recommends that the airmass limit for the Near-Sun Twilight microsurvey is increased to $X=3.0$ (\autoref{sec:additional}). +\item TWILIGHT SURVEY: The SCOC recommends that the airmass limit for the Near-Sun Twilight microsurvey is increased to $X=3.0$ (\autoref{sec:additional}). -\item The SCOC recommends a slight modification of the \baseline{3.0} footprint to improve overlap with the Euclid footprint (\autoref{sec:additional}). +\item EUCLID OVERLAP: The SCOC recommends a slight modification of the \baseline{3.0} footprint to improve overlap with the Euclid footprint (\autoref{sec:additional}). \end{enumerate} -These recommendations are implemented in the \baseline{4.0} simulations (\autoref{sec:baseline4_0}). A set of simulations tagged \texttt{v3.6} was made available for the community in early September to assess the impact of different aspects of the recommendation. Note that all of these simulations include the updated, more realistic downtime and effects of slew jerk. In what follows we describe only key simulations; for a more comprehensive description of all \opsim\ s released by the Observing Strategy team, please see posts within the Survey Strategy topic on the LSST Community forum.\footnote{\url{https://community.lsst.org/c/sci/survey-strategy/37}.} +These recommendations are implemented in the \baseline{4.0} simulations (further described in \autoref{sec:baseline4_0}). A set of simulations tagged \texttt{v3.6} was made available for the community in early September to assess the impact of different aspects of the recommendation. Note that all of these simulations include the updated, more realistic downtime and effects of slew jerk. In what follows we describe only key simulations; for a more comprehensive description of all \opsim\ s released by the Survey Strategy team, please see posts within the Survey Strategy topic on the LSST Community forum.\footnote{\url{https://community.lsst.org/c/sci/survey-strategy/37}.} \autoref{fig:nvisits} shows the number of visits to the WFD survey across simulations, starting with the early vision for LSST (2018) through today's recommendation. \autoref{fig:summary} shows the performance of the survey strategy on a set of core LSST science (top 19 rows) and system metrics (bottom 3 rows) over the same \opsim s. - First, note that nearly all science cases have seen improvements over time, with some science cases improving by over 50\% (the stretch of the color scale) demonstrating the success of the community-driven approach to survey strategy design that Rubin Observatory has committed to for the past decade, which makes Rubin LSST a much more complete and comprehensively transformational survey. Significant improvements were obtained on most metrics through v3.0 (\citetalias{PSTN-055}). Those are to be attributed to changes of the survey strategy through community input and SCOC recommendations. + First, note that nearly all science cases have seen improvements over time, with some science cases improving by over 50\% (the stretch of the color scale) demonstrating the success of the community-driven approach to survey strategy design that Rubin Observatory has committed to for the past decade, which makes Rubin LSST a much more complete and comprehensively transformational survey. Significant improvements were obtained on most metrics through \texttt{v3.0} (\citetalias{PSTN-055}). Those are to be attributed to changes of the survey strategy through community input and SCOC recommendations. The visible improvement on nearly all metrics between \baseline{3.2} and \baseline{3.3} is attributed to the updated filter transmission curves. The survey strategy is largely unchanged between \baseline{3.3} and \baseline{3.4}; the small changes in performance are to be attributed to \texttt{rubin\_scheduler} code updates\footnote{See \url{https://survey-strategy.lsst.io/baseline/changes.html}} (also remember that metrics have different degrees of stochasticity in their design). - The \baseline{3.5} \opsim\ (labeled \texttt{v3.5} in \autoref{fig:summary}) represents an early implementation of the current SCOC recommendations: it includes the new filter balance ($u_{exp}=38$ seconds, $N_u \leq1.1\times ns$, \autoref{sec:filterbalance}), updated Galactic Plane footprint, a slightly extended fraction of time spent on DDFs (still within 7\% as recommended in \citetalias{PSTN-055} and \autoref{sec:DDF} in this document), 3-cycle uniform rolling (but note that, while implemented, the SCOC has not committed to this recommendation, \autoref{sec:rolling}), but does not include snaps or ToOs. Most metrics are stable or improved, except for some time domain metrics (\eg, KNe metrics and SNIa) due to the rolling in three, instead of four cycles and to the new filter balance. + The \baseline{3.5} \opsim\ (labeled \texttt{v3.5} in \autoref{fig:summary}) represents an early implementation of the current SCOC recommendations: it includes the new filter balance ($u_{exp}=38$ seconds, $N_u \leq1.1\times ns$, \autoref{sec:filterbalance}), updated Galactic Plane footprint, a slightly extended fraction of time spent on DDFs (still within 7\% as recommended in \citetalias{PSTN-055}), 3-cycle uniform rolling (but note that, while implemented, the SCOC has not committed to this recommendation, see \autoref{sec:rolling}), but does not include snaps or ToOs. Most metrics are stable or improved, except for some time domain metrics (\eg, KNe metrics and SNIa) due to the rolling in three, instead of four cycles and to the new filter balance. The overall apparent drop in performance between \baseline{3.5} and \baseline{3.6} is primarily due to the inclusion of slew time jerk effects and more realistic estimates of downtime in Y1 (\autoref{sec:opsimchanges}). \autoref{fig:nvisits} shows the associated drop in number of visits. The \baseline{3.6} has three rolling cycles and includes the ToO program. We provide an \opsim\ consistent with \baseline{3.6}, but without the ToO program to allow the community to investigate the effects that the introduction of ToOs has on LSST. - The \baseline{4.0} represents the current recommendation outlined in this document. An implementation of this recommendation with four cycles of rolling is offered in \texttt{four cycles v4.0} to enable the investigations of different rolling implementations. Finally, we provide an implementation of \baseline{4.0} with single exposure visits (instead of 2x15 second snaps, \autoref{sec:snaps}) which, pending commissioning outcomes, is the expected observing mode. In this \opsim, the survey time gained by dropping snaps (decreased readtime per visit) is allocated evenly across all observing modes: this includes the WFD, NES, SCP, and Galactic Plane. In the future, with better knowledge of the system as built, the SCOC will consider how the additional time may be allocated, including allocations to special programs, to compensate for unexpected performance loss or to increase science throughput. + The \baseline{4.0} represents the current recommendation as outlined in this document. An implementation of this recommendation with four cycles of rolling is offered (\texttt{four cycles v4.0}) to enable the investigations of different rolling implementations. Finally, we provide an implementation of \baseline{4.0} with single exposure visits (instead of 2x15 second snaps, \autoref{sec:snaps}) which, pending commissioning outcomes, is the expected observing mode. In this \opsim, the survey time gained by dropping snaps (decreased readtime per visit) is allocated evenly across all observing modes: this includes the WFD, NES, SCP, and Galactic Plane. In the future, with better knowledge of the system as built, the SCOC will consider how the additional time may be allocated, including allocations to special programs, to compensate for unexpected performance loss or to increase science throughput. \begin{figure} \centering %\begin{overpic}[width=0.8\textwidth]{figures/total_nvisits.png}