Merge pull request #591 from rhiannonlynne/master

Update PHA section of white paper.

Author: ivezic

whitepaper/SolarSystem/SolarSystem_PHA.tex

@@ -49,8 +49,8 @@ \subsection{Target measurements and discoveries}
 Using the same range of discovery criteria as in the previous section,
 \ref{sec:solarsystem:discovery}, we can look at the differential and
 cumulative completeness for a population of PHAs. For this sample of
-PHAs, we simply pulled the orbits of the brightest (D$>$1~km)
-$\sim1500$ PHAs from the Minor Planet Center record. These orbits were
+PHAs, we pulled the orbits of 2,000 objects with MOID~$<= 0.05$~AU from 
+the Grav S3M model \citep{2011PASP..123..423G}. These orbits were
 then cloned over a range of $H$ values to evaluate the chances of
 discovery for that orbit at each of those $H$ values. The differential
 completeness as a function of $H$ is then simply the fraction of
@@ -81,95 +81,41 @@ \subsection{OpSim Analysis}
 
 The differential and cumulative completeness for the baseline survey,
 \opsimdbref{db:baseCadence}, at a range of years is shown in
-\autoref{fig:baselinePHA}. The baseline cadence achieves a cumulative completeness of 73\% for
-H$\le$22 PHAs. The differential completeness at $H$=22 for the same
-survey is 58\%, 15\% lower due to increasing completeness toward
+\autoref{fig:baselinePHA}. The baseline cadence achieves a cumulative completeness of 66\% for
+H$\le$22 PHAs when requiring pairs of visits on 3 separate nights within 15 days. 
+The differential completeness at $H$=22 for the same
+survey is 49\%, 17\% lower due to increasing completeness toward
 smaller $H$ (larger objects).
 
 %%%%%%%%%%%%%%%%%%%%%%%%%%%
 \begin{figure}[th]
 %\vskip -1.1in
-\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_Completeness_2_10_8_6_4_pha_year_3_pairs_in_15_nights_MOOB_ComboMetricVsH}
-\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_CumulativeCompleteness_2_10_8_6_4_pha_year_3_pairs_in_15_nights_MOOB_ComboMetricVsH}
+\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_DifferentialCompleteness_PHA_3_pairs_in_15_nights_Years_1_to_10_MOOB_ComboMetricVsH}
+\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_CumulativeCompleteness_PHA_3_pairs_in_15_nights_Years_1_to_10_MOOB_ComboMetricVsH}
 %\vskip -1.2in
 \caption{The PHA completeness for \opsimdbref{db:baseCadence}, as a function of the object's absolute
 visual magnitude H on the horizontal axes (left: differential completeness at a given H;
-right: cumulative completeness for all objects brighter than a given H).
+right: cumulative completeness for all objects brighter than a given H), as it increases year over year.
 The cumulative completeness for H$\le$22 NEOs (those with diameters larger than 140m)  for this
-simulation is 73\% after 10 years.}
+simulation is 66\% after 10 years.}
 \label{fig:baselinePHA}
 \end{figure}
 %%%%%%%%%%%%%%%%%%%%%%%%%%%
 
+We find that the PHA and NEO completeness are very similar for a given simulated survey and set of discovery criteria, as shown in \autoref{fig:neopha}.  The analysis of the various observing run strategies (singles, pairs, triples or quads of visits) described in the previous section thus applies to PHAs as well; while changing the discovery metric to triplets or quads significantly decreases completeness, simply changing the survey strategy has a softer effect, most likely due to current limitations of the simulated surveys.
+
 %%%%%%%%%%%%%%%%%%%%%%%%%%%
 \begin{figure}[bh]
 %\vskip -1.2in
-\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_Completeness_3_15_pairs_3_30_pairs_quads_3_30_3_30_triplets_pairs_20_4_nights_in_pha_year_10_MOOB_ComboMetricVsH}
-\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/enigma_1282_Completeness_3_15_pairs_3_30_pairs_quads_3_30_3_30_triplets_pairs_20_4_nights_in_pha_year_10_MOOB_ComboMetricVsH}
+\includegraphics[angle=0,width=0.49\hsize:,clip]{figs/solarsystem/minion_1016_CumulativeCompleteness_NEO_and_PHA_Cumulative_Completeness}
 %\vskip -1.3in
 \caption{%
-Comparison of the differential PHA completeness for the baseline cadence
-\opsimdbref{db:baseCadence}, requesting two detections per night (left), and
-\opsimdbref{db:NEOwithVisitQuads}, requesting four detections per
-night (right). With a discovery criteria of 3 pairs within 15 nights,
-both surveys perform roughly similarly; with a discovery criteria of 3
-sets of quad visits within 30 nights,
-\opsimdbref{db:NEOwithVisitQuads} performs better (as expected),
-although still at a lower completeness level than
-\opsimdbref{db:baseCadence} did with the pairs criteria.}
-\label{fig:strategiesPHA}
+Comparison of the cumulative NEO and PHA completeness for the baseline cadence
+\opsimdbref{db:baseCadence}.}
+\label{fig:neopha}
 \end{figure}
 %%%%%%%%%%%%%%%%%%%%%%%%%%%
 
-The differential completeness for a range of discovery criteria, for
-both the baseline survey and \opsimdbref{db:NEOwithVisitQuads}, is
-shown in \autoref{fig:strategiesPHA}. When the discovery algorithm
-requires pairs of visits, the runs have fairly similar PHA
-completeness, with \opsimdbref{db:NEOwithVisitQuads} having a
-differential completeness about 6\% lower than
-\opsimdbref{db:baseCadence}.  When the discovery algorithm requires 4
-detections per night, the simulation with quads achieves a
-differential completeness of about 15\% higher than the baseline
-cadence (as some quads are unintentionally produced by chance, see
-\autoref{fig:NvisitStats}).
-
-\begin{table}[h]
-\centering
-\caption{Differential PHA completeness at $H$=22}
-\label{phacompleteness}
-\begin{tabular}{l|c|c|c|c}
- & \opsimdbref{db:baseCadence} & \opsimdbref{db:NoVisitPairs} &
-                                                                \opsimdbref{db:NEOswithVisitTriplets}
-  & \opsimdbref{db:NEOwithVisitQuads} \\
-
-3 pairs in 15 nights & 58 & 51 & 56 & 52 \\
-3 pairs in 30 nights & 61 & 56 & 59 & 57 \\
-4 pairs in 20 nights & 50 &  41 & 46 & 42 \\
-3 triplets in 30 nights & 35 & 33 & 50 & 48 \\
-3 quads in 30 nights & 22 & 18 & 19 & 37 \\
-
-\end{tabular}
-\end{table}
-
-
-\begin{table}[h]
-\centering
-\caption{Cumulative PHA completeness at $H$=22}
-\label{phacompleteness}
-\begin{tabular}{l|c|c|c|c}
- & \opsimdbref{db:baseCadence} & \opsimdbref{db:NoVisitPairs} &
-                                                                \opsimdbref{db:NEOswithVisitTriplets}
-  & \opsimdbref{db:NEOwithVisitQuads} \\
-
-3 pairs in 15 nights & 73 & 69 & 71 & 68 \\
-3 pairs in 30 nights & 76 & 73 & 74 & 73 \\
-4 pairs in 20 nights & 68 & 62 & 64 & 61 \\
-3 triplets in 30 nights & 57 & 55 & 66 & 65 \\
-3 quads in 30 nights & 42 & 37 & 37 & 55 \\
-
-\end{tabular}
-\end{table}
-
 % ====================================================================
 %
 % \subsection{Conclusions}

whitepaper/figs/solarsystem/enigma_1281_Completeness_3_15_pairs_3_30_pairs_quads_3_30_3_30_triplets_pairs_20_4_nights_in_pha_year_10_MOOB_ComboMetricVsH.pdf

@@ -5277,3 +5277,18 @@ @ARTICLE{2013APh....42...52G
    adsurl = {http://adsabs.harvard.edu/abs/2013APh....42...52G},
   adsnote = {Provided by the SAO/NASA Astrophysics Data System}
 }
+
+@ARTICLE{2011PASP..123..423G,
+   author = {{Grav}, T. and {Jedicke}, R. and {Denneau}, L. and {Chesley}, S. and 
+	{Holman}, M.~J. and {Spahr}, T.~B.},
+    title = "{The Pan-STARRS Synthetic Solar System Model: A Tool for Testing and Efficiency Determination of the Moving Object Processing System}",
+  journal = {\pasp},
+     year = 2011,
+    month = apr,
+   volume = 123,
+    pages = {423},
+      doi = {10.1086/659833},
+   adsurl = {http://adsabs.harvard.edu/abs/2011PASP..123..423G},
+  adsnote = {Provided by the SAO/NASA Astrophysics Data System}
+}
+