These sets of examples describe how to retrieve data from the Spitzer archive using the Spitzer Heritage Archive (SHA) web-based interface. The SHA is housed at IRSA, the Infrared Science Archive. The URL is http://sha.ipac.caltech.edu/applications/Spitzer/SHA/
Detailed introductory examples of searching the SHA can be found in the first two recipes in the Spitzer Data Cookbook, available from the Spitzer documentation site.
12.2 Quick Hints, Tips, and Term Definitions:
An individual Spitzer observation sequence is an AOR, or Astronomical Observation Request. In certain cases (often calibration or sometimes science observations), you may also see an IER, or Instrument Engineering Request. Either one involves many individual frames. Both have a large integer number called an AORKEY that uniquely identifies that observation within the mission.
The individual data frames that emerge, calibrated, from the Spitzer pipeline are Level 1, or Basic Calibrated Data, or BCDs.
The products that come from combining these individual data frames (such as mosaics of individual pointings) are Level 2, or post-BCD, or PBCD data.
Spitzer observations can cover large areas or, by design, multiple targets. If you are interested in just portions of the larger observation, you can choose to have just individual data products returned -- e.g., just the observations that went into the portion of the sky for which you searched -- or you can return the products for the whole AOR.
The SHA website displays sub windows, called panes. Items within each pane often appear in tabs – click on different tabs to display different pane contents.
12.3 Quick Examples – Getting your feet wet
These all assume that you are new to the SHA, but that you are fluent in basic vocabulary (1.2).
12.3.1 I have detected an object at a non-Spitzer band, and I want to know if Spitzer saw anything there.
Start an SHA search by position. Enter your desired coordinates. At the very least, ask for Level 2 data to be returned. Enter a search radius of at least 500 arcsec.
When you get a search result, in the search results pane, click on one of the rows you wish to investigate. Go over to the Details Pane, and click on the “AOR Footprint” tab. This is the outline of the observations. Did the field of view or the slit actually hit the specific region you wanted?
If you need to investigate this region further, move your mouse over the image and hesitate until you see this icon in the corner: Click on that to bring up a set of image manipulation tools:
Let your mouse hesitate over any of the icons for a “tool tip” indicating what it does. Click on the “spider webby” icon to add coordinates to your image, for example. You can change the stretch, or change the background image.
If not, pick another observation and try this again until you find an observation you think would be useful to pursue further.
Having found that one observation that you think would be useful to pursue further, especially if you have more than one AOR returned, click on the checkbox on the far left of that row in the Results pane. Ensure that “restrict data in other tabs” is checked near the top of the Results pane. This makes sure that the data as shown in the other tabs only pertains to the one observation you have selected. (You can also select more than one, but let’s not for right now.) Click on the “Level 2 (PBCD)” tab.
The Level 2 tab shows you the highest-level products available for this observation, such as mosaics or extracted spectra. Look at the “file type” to discern whether they are images or tables (most often spectra). Click on a row corresponding to one of these products.
In the Details pane, click on the “Data” tab. These are the real data that are shown. Is the object you care about within the mosaic? Can you see it?
As before, move your mouse over the image and hesitate until you see this icon in the corner: Click on that to bring up a set of image manipulation tools. Click on the “spider webby” icon to add coordinates. You can change the color map and stretch; click on the icon that looks like a color cube to change the colors and the one that looks like a squashed color cube to change the stretch.
If it is a spectrum, you can examine the 2-D image prior to spectral extraction to see if there is something there; you can also make line plots of extracted spectra.
Convinced that you should pursue this? Go back to the AOR tab in the Results pane, and ensure that the checkbox on the far left of the line is checked. Click on “Prepare Download” (near the top of the Results pane) and ask for the Level 2 data, plus ancillary if you are feeling complete.
While the data are packaging up, you may wish to see if anyone has delivered reduced data on your region back to IRSA and the SSC. Click on “Spitzer searches” (blue tab in the top left) and repeat your position search, this time ensuring that the ‘Enhanced Products’ tabs are selected.
Enter your coordinates. This will search enhanced products from the SSC (in this case, IRS enhanced products), as well as enhanced products delivered to the SSC and IRSA from the community. (These are currently largely – but not entirely – deliveries from the Spitzer Legacy Teams.) Both of these kinds of enhanced products were generated from targets spread all over the sky, but none were all-sky surveys. You may get lucky and find some products, some from bands other than Spitzer bands. USE LARGE SEARCH RADII WITH CAUTION. Some of these catalogs are very large!
Additional tabs will be returned with your search results, though some may have “no data to display.”
If you do find Contributed Products, click on the first link provided (the name of the dataset) to bring up another tab with the search results. You can investigate the coverage to see if you want to download the data, or overlay the source list to see if you want that catalog. You can download the data here in the same fashion as for the rest of the SHA – click in the checkbox on the left, and ‘Prepare Download’ to package up the data. If you do download data or catalogs, please do obtain and read the delivery documentation (obtainable from the page you reach by clicking the ‘more’ link after the entry on the summary tab). There may be caveats about the images, or data quality flags in the catalogs.
By now, your original Spitzer data packaging should be done and ready for download. Click on the “Background Monitor” link in the upper right to bring up a pop-up with the packages.
Click on “Download Now” to download the files; if there are many files, try out the scripts to download them all automatically. Consult the IRAC, IRS, or MIPS Instrument Handbook for additional information about the files, descriptions of artifacts, data reduction examples, etc., and you’re on your way!
12.3.2 I study {this subject, as opposed to this target}, and I want to know if there were any Spitzer programs specifically studying {this subject}.
From the main search page, pick “Abstract” to do an abstract search. If you study planetary nebulae, then a search on “planetary nebula” ought to pull up both the singular and plural forms of the noun. Note that big surveys will cover huge swaths of sky that will contain objects of many different types (e.g., many planetary nebulae in the galactic plane covered by both GLIMPSE and MIPSGAL), and this sort of search will only return projects which mentioned your search words in the abstract. However, it should give you some good leads to follow.
Click on the title of any project to pull up all of the observations associated with that program into the Search Results window pane, and proceed from there as with the other examples earlier above.
12.3.3 I want to search the SHA for a whole list of objects all at once.
You can do a “batch search.” Search by position then click on “from file.”
A viable input file can either be an IPAC table file or a relatively simple text file described below. For an IPAC table file, you may find the IPAC table file verification service helpful (http://irsa.ipac.caltech.edu/applications/TblCheck/ ). If doing a regular non-IPAC table file search, the file format is:
COORD_SYSTEM: Equatorial
# Equatorial, Galactic, or Ecliptic - default is Equatorial
EQUINOX: J2000
# B1950, J2000, or blank for Galactic - default is J2000
NAME-RESOLVER: NED
# NED or Simbad - default is Simbad
#Name RA/LON DEC/LAT PM-RA PM-DEC EPOCH
"NGC 001" 12h34m23.45s 34d23m56.2s 2.3 3.4 1950.3
NGC2222 23.56d 34.456d 2.3 3.4 1950.3
NGC3333 23.56h 34.456d 2.3 3.4 1950.3
NGC4444 "12 34 12.23" "34 23 45.45"
m31
legacy "17 18 00" "59 30 00"
m32
m33 simbad
NGC6946
NGC5194
ngc2992
Note in the example above that in some cases, I have specified the epoch of the observations (e.g., “NGC3333”), or I have allowed NED to resolve the name (“m31”), or I have asked for SIMBAD to resolve the name (“m33”). Note that the name resolver is not case-sensitive (“NGC5194” vs. “ngc2992”).
The SHA parses on spaces, so a space is the delimiter between fields. Therefore, if there is a space in your object name (e.g., "NGC 1001" versus "NGC1001") or position ("34 23 45.45" versus 34d23m45.45s"), you need to put quotes around the target name or its position.
When you are ready to search, go back to the SHA, find the file on your disk and upload it to the SHA. The search results should be returned like any other search, though you cannot tag batch searches for later repeat searches. If the same AOR is returned for more than one of your targets, it appears in the list more than once.
Having problems making this work? Double-check your formatting -- that's the most common error. If name resolution fails for some of the targets, the rest of the search may fail -- provide coordinates for the troublesome names, or remove them.
12.4 More Quick Case Studies (Examples).
These all assume that you have some experience with the SHA, at the very least the examples above, and you need to be given a quick indication of how to do some more sophisticated things with it. No screen shots are given.
12.4.1 Just getting one program’s data out of a recent campaign.
Why? Many programs in the Warm Era are time series observations, and some investigators may want all the data from a given program that have just been released.
How: Search by campaign for the most recent campaign. Many AORs will be returned, but only the first n rows will appear (where n defaults to 50). Once the results appear, apply a filter to the results. Click on "add filters" in the upper right of the "Campaign Search Results" pane. Click on "Program ID" from the first drop-down menu. Leave the operator as the default "=". Enter the program ID you want. Apply the filter and close the filters window. Just the AORs from that campaign and that program will appear. Select all of them by clicking the checkbox on the top of the column of checkboxes. Click ‘Prepare Download’ and select which data you want packaged. Once packaged, download the data from the background monitor.
NOTE THAT the filters as imposed this way only apply to your current tab, even if "restrict data in other tabs" is selected!
12.4.2 Just getting calibration data.
Why? Although your data arrive calibrated after having gone through the pipelines, some investigators may wish to find the calibration data that are tied to a particular campaign.
How: Search by campaign for the one you want. Many AORs will be returned, but only the first n rows will appear (where n defaults to 50). Calibration observations are likely to be among the first in the campaign. Look at one of these rows by selecting the row (clicking anywhere on it) and noting the details that appear in the Details pane. Calibration programs will have a program ID, title, and PIs just like a regular science observing program, but the titles will look like “Calibration Program” and the PI will look like “IRAC Calibration”. Each campaign should have its own calibration program ID, though some may have more than one. You can explore these and note the program ID (PID) and the PI of a calibration observation in your campaign. Apply a filter to the search results to leave only programs like that calibration program. Click on "add filters" in the upper right of the "Campaign Search Results" pane. Click on “PI” and enter the PI of calibration program, being careful to match the text exactly, OR click on “PID” and enter the program ID of the calibration program. Apply the filter and close the filters window. Just the AORs from that campaign and that program will appear. Select all of them by clicking the checkbox on the top of the column of checkboxes. Click ‘Prepare Download’ and select which data you want packaged. Once packaged, download the data from the background monitor.
NOTE THAT the filters as imposed this way only apply to your current tab, even if "restrict data in other tabs" is selected!
12.4.3 Getting many zip files to automatically download
Why? If you have asked the SHA to prepare a large number of files (big programs, whole campaigns, etc.) for you, it will break it up into "manageable" pieces, where "manageable" is defined as "not larger than common computers and software can handle."
How:
If you don't want to click to download each piece, use the download script provided by the Background Monitor (when you have more than 1 zipfile to download), available either from the Background Monitor itself or from email sent to you when the packaging is complete. There are several download scripts provided
Simplest incarnation:
First, ensure that you have "wget" installed by typing "which wget" at a terminal prompt. If you do not have it, download and install it from gnu.org. Save the wget script from the SHA to a plain text file, and invoke the wget lines from this plain text file either by copying and pasting those lines individually into your terminal window, or by typing "csh [yourtextfile]" at the prompt. The files will be automatically and sequentially downloaded to your disk. The files stay on disk here for at least 72 hours, so you have a window of time to download them.
More sophisticated options:
Automatic unzipping: Another wget (and curl; see below) script that is provided will enable your machine to automatically unzip the packages that are downloaded. Download that script and invoke it, as above.
Curl: Curl (also spelled cURL) is another interface, like wget, but is optimized for Macs. Curl probably came installed on your system (unlike wget), but to check, type “which curl” at a terminal prompt. If you do not have it, Google it, download it, and install it. Save the curl script from the SHA to a plain text file and invoke the script from this plain text file either by copying and pasting those lines individually into your terminal window, or by typing "csh [yourtextfile]" at the prompt. A curl script that automatically unzips the files is also provided by the SHA.
Your own script: a plain text file listing the URLs for the zipfile downloads is also provided, should you want to write your own program for downloading these zipfiles.
Windows machines:
Google, download, and install wget for Windows. Google, download, and install unzip.exe for the command-line for Windows. See below for some suggestions. Download the plain text file listing the URLs for the zipfile downloads, being sure to save it as plain text. Then, to download using wget and the file of URLs downloaded from SHA:
- double-click .exe will extract a bunch of files. Only unzip.exe is needed. For convenience, copy unzip.exe into the <wget_install_dir>\bin
Add <wget_install_dir>\bin into your path.
12.4.4 Finding other data that are similar to yours.
Why? Here is a real-life example of a recent request: I have spent xxx months slaving over my time-series data of yyy taken on zzz, and I don't have the scatter that I expect to have in my non-variable stars. I would like to know if it's my analysis or this particular data set which is the problem, so I need to find another observation of something (anything, doesn't much matter what) taken with staring mode IRAC, 30 second frame time, so that I can push the new data set through my analysis and see if I am getting the same scatter.
How: There is no easy way to find observations in a particular very specific mode without a priori knowing which targets were observed with those modes. You'd have to search the whole sky for observations matching a particular set of criteria. The best way to do this is to contact us through the Helpdesk, and we will get into the database itself and help you find similar observations.
12.4.5 Finding Contributed Enhanced Product deliveries in any given region
Why? The Spitzer Legacy Programs were large, coherent science investigations. The teams agreed to deliver enhanced products back to the Spitzer Science Center and IRSA. Other teams, not necessarily formally Legacy teams, also delivered products. Some Exploration Science programs (large programs in the Warm Era) have also expressed a desire to deliver products. Each of the deliveries currently at IRSA contain something different -- some just delivered images, some delivered images, catalogs, and spectra, some delivered multi-wavelength resources, extending well beyond the Spitzer bands, some delivered models. All of the products that have been ingested into IRSA are available through this SHA interface. Note that as of this writing, some teams still have pending deliveries, and that IRSA ingestion is not immediate. When a Legacy team delivers enhanced data products to IRSA, the data must go through a quality assurance (QA) and ingestion process before being available through IRSA search engines, and this process does not happen instantaneously.
At any rate, it may very well be that you can use these contributed enhanced products, catalogs or images, for your research. It’s worth checking out!
How:
Go to the “Position Search” page. Ensure that the “Contributed Enhanced Products” box is checked.
USE LARGE SEARCH RADII WITH CAUTION. Some of those catalogs are very large, and may even time out if you are retrieving, say, half a million rows.
Contributed enhanced product search results come up with a tab that is a "Summary View" of the search results. By default, the results are grouped by type -- images, catalogs, or spectra. You can also choose to sort the results by originating project.
The images come up quickly, and the summary page indicates how many images from each project are returned. Note that sometimes this could mean one image for each band, or it could mean image, errors, and coverage for each band, or it could mean multiple resamplings of the same bands, etc. To investigate exactly what images are available, click on the first link in each line. Additional tabs are spawned for each click, containing a summary of the contents of each image. Please consult the original documentation that came with the delivery for definitions, explanations, and caveats that might go with each image. Note that some teams delivered, e.g., GALEX and Halpha images, so you may obtain non-Spitzer data.
Catalogs come up more slowly (because they are harder to search!), and depending on your search, the SHA may tell you that it has completed searching on "8 of 10 catalogs" or a similar phrase. Here too, the catalog contents vary across projects, and even between deliveries within the same project. In the case of the catalogs, the "count" tells you how many individual catalog entries there are in that catalog for your search parameters. Click on the first link in each line to spawn an additional tab with the catalog values. Note that by default, only 50 lines are shown, but as these catalogs can be very long, it may take a second or two to read in and render. The catalog columns as displayed are exactly as provided by the original Legacy team (with some additional columns added by IRSA to reformat the RA/Dec into standard options), so the columns are effectively different for each catalog. Please consult the original documentation that came with the delivery for column definitions, caveats, quality flags, etc.
For information on individual programs, and information on individual deliveries, please see the SSC website.
12.4.6 Finding data from two different instruments or modes at once.
Why? If you have searched on, e.g., the FEPS Legacy program ID, and are looking at a long list of observations using IRAC, MIPS, and IRS, perhaps you wish to just view IRS Stare and IRS Map observations.
How: If you have already searched and are looking at the grid of results, you can impose filters. But, the filters imposed after a search are most commonly logically ANDed together; that is, if you add a filter to only show IRS Staring observations and then add another filter to only show IRS Mapping Observations, you will be left with no data, because the filters are looking for data that are both IRS Stare AND IRS Map, and no data meet (can ever meet) that requirement.
The fastest way to actually weed down the results in the fashion desired in this example is to repeat your search – go back to the Search Pane by clicking on “Search Again”. Click on “More Options” to pull down the additional options. Click on all the observing modes you wish to include, and un-select the observing modes you do not wish to include. Search again, and the search results will then only include, e.g., data from both IRS Staring and IRS Mapping observations.
But… see the next example!
12.4.7 Finding data from two different channels at once.
Why? If you have searched on IRAC data and are looking at a stack of results, perhaps you only want to examine and download the 5.8 and 8 micron results.
How: If you have already searched and are looking at the grid of results, you can impose filters. Although the filters imposed after a search are most commonly logically ANDed together, you can get around this. You can use the “IN” operator which means 'included within this list', e.g., the parameter on which I am querying (such as "Bandpass") is included within the list I am specifying (such as, from the list of bandpasses provided, where more than one can be selected by holding down the command key and clicking on the bands desired). For this example, do exactly that – go to the Level 1 or Level 2 tab, click on “Apply filters”, then pick “Bandpass”, then “IN”, then select 5.8 and 8 microns from the list. Don’t forget to “Apply” the filter, then close the filter window. The rows that should be left in the window are just the ones that correspond to the 5.8 and 8 micron observations.
NOTE THAT the filters as imposed this way only apply to your current tab, even if "restrict data in other tabs" is selected!
12.4.8 Finding just MIPS prime or non-prime data.
Why? MIPS had three cameras on board, and, when MIPS was operating, all three cameras were always on. In certain observing modes (e.g., 70 micron photometry), the commanded action of the spacecraft + scan mirror was designed to image a particular section of sky at that band (the ‘prime’ band), and the other cameras (e.g., 24 micron) imaged whatever section of sky happened to be visible to it during that sequence (the ‘non-prime’ band). When the 24 µm array was the primary array, the Germanium arrays (70 and 160 microns) did not have the right calibration sequence for the data to be scientifically viable, but the 24 µm array did take viable data when the 70 or 160 micron arrays were primary. Observers looking for serendipitously obtained data may wish to specifically obtain non-prime data.
How: Search on your desired position/observation/program/etc. Under “More options” on the search page, ask it to only return MIPS observations, and ensure that you are asking it to return Level 1 data as well as anything else you want returned. When the search results appear, click on the Level 1 tab to bring it to the foreground. Scroll over, and find the column labeled “Primary field”. This value is 1 if that array was prime, and 0 if it was not. Go up to the upper right of the search results window, and click on “Add filters.” Choose “Primary field”, then the operator you want (probably “=”) and the value you want to select (1= yes, it was prime, 0= no, it was not prime but instead ancillary). Apply the filter and close the window. The remaining observations as displayed are the observations that meet the criterion you have imposed.
NOTE THAT the filters as imposed this way only apply to your current tab, even if "restrict data in other tabs" is selected!
12.4.9 Filtering by PI.
Why? If you have a large list of observations taken by many people, you can impose a filter to find observations just by one PI, but the filters are very sensitive to exact matching of PI names.
How: Search on your desired position/observation/program/etc. When the search results appear, scroll over until you find the PI column. Make a note of exactly how the PI name appears. Go up to the upper right of the search results window, and click on “Add filters.” Choose “PI”, then the “=” operator, and enter the name exactly as it appears in the list. Apply the filter and close the window. The remaining observations as displayed are the observations that meet the criterion you have imposed.
NOTE THAT the filters as imposed this way only apply to your current tab, even if "restrict data in other tabs" is selected!
12.4.10 Finding AORs that were constrained together.
Why? In general, Spitzer discouraged use of constraints to link AORs together; fewer constraints on the observations made them easier to schedule. However, for some observations, it was scientifically necessary to require that, say, observation “b” be obtained after 4 hours but not more than 6 hours after observation “a”. You may want to obtain observations that were constrained to a given observation.
How: Go to the AORKEY search. Enter the AORKEY you want to search. Ensure that the “Get all AORs in the same scheduling constraints” checkbox is checked, and submit the search. All AORs that are linked to that AOR will be returned.
The SSC website summarizes all of the flavors of constraints that were available.
NOTE THAT some time series observations were obtained in the following manner. For a series of observations (a, b, c, d, e), a was tied to b with a particular constraint, b tied to c, c tied to d, and d tied to e. If you search on the AORKEY corresponding to observation b, and ask the SHA to give you all of the AORs constrained to that observation, it does exactly (and only) what you asked it to do -- b is tied to a and c, but not explicitly tied to d or e. So it returns to you observations a, b, and c, and not d or e.
12.4.11 Finding IRAC subarray Level 2 data.
Why? When you do a search on most IRAC data, individual data frames (Level 1) as well as mosaics (Level 2) are returned. But no mosaics are produced for IRAC subarray data! If you search on subarray data (or if subarray data are some of the results you have uncovered), and if you do not select Level 1 data as one of your desired results, these subarray data will appear to vanish because they have no data to display in the Level 2 tab.
How: Since no subarray mosaics are created, there will never be any to display. You must download the Level 1 BCD frames, and make your own mosaic from the BCD data using MOPEX or another software package of your choice.
Potential gotcha: If you search for a list of observations of, e.g., certain exoplanet targets, you will return many observations in the AOR tab, some of which will be subarray and some of which will be full array. When you go directly to the Level 2 tab for this search, there will be entries from the full array observations, but not the subarray observations. If you have searched only for subarray observations, it is obvious that something unusual is going on when you go to the Level 2 tab and get ‘no data to display.’ But, when you have a mixture of kinds of observations in the AOR tab, then there are still rows to display in Level 2, and it is not obvious that something unusual has happened.
Why the heck does the pipeline not make mosaics for subarray data? The post-BCD pipeline was not designed to work with the three-dimensional data cubes that are the subarray BCDs. When the software was developed it was deemed unlikely that subarray observations would be used for anything other than photometry of single bright objects and mosaics would not be a desired product. Please see the IRAC Instrument Handbook for more information on dealing with subarray data.
12.4.12 Finding other IRAC Level 2 data that appears to be missing.
Why? When you do a search on most IRAC data, individual data frames (Level 1) as well as mosaics (Level 2) are returned. But, for some other kinds of IRAC observations (some IERs), there are also no Level 2 products produced.
How: Since no mosaics are or will be produced, there will never be any Level 2 products for download. You must download the Level 1 data and make your own mosaic from the BCD data using MOPEX or another software package of your choice.
The IERs where there are missing Level 2 products are ones where there are more than 8000 frames within a single observation. These are most commonly exoplanet monitoring observations, in which case a single mosaic is not particularly the desired product in any case.
12.4.13 Downloading raw data.
Why? Some expert users will want to download the raw data for an observation.
How: When requesting packaging of the data, you can request that the raw data also be packaged for download. Most users will never need to access the raw data. You cannot search on the raw data directly.
NOTE THAT not every frame (DCE, or data collection event) was used in the creation of the BCD and PBCD products. All of the original DCEs are still in the SHA, and can still be obtained, even if you cannot search on the raw data directly.
When packaging by AOR and asking for the raw data, all products matching the designated AORKEY are packaged, regardless of whether “apply spatial constraints to individual data products” is selected and the DCEs fall out of that cone.
When packaging by BCD or PBCD and asking for the raw data, all products matching the designated AORKEY and channel are packaged, regardless of whether “apply spatial constraints to individual data products” is selected and the DCEs fall out of that cone.
12.4.14 Apparent logical inconsistencies in position searches.
Why? If you search by position with the default radius, you may see that, for example, both IRAC warm channels were observed in the AOR tab, and are there in the Level 2/PBCD tab but find that one channel appears to be missing in the Level 1/BCD tab.
How: This is a weird one, but it is actually doing exactly what it is supposed to be doing, for several reasons. By default, the SHA applies the spatial constraints to the individual data products – this radio button option is the first one that appears under “More options” so you have to consciously find this button if you want to change it. The default search radius of ~8 arcmin is smaller than two fields of view for IRAC. The Level 2/PBCD mosaics are produced on an AOR basis, and include the area observed by both fields of view even if it is filled with NaNs. It is quite possible, therefore, to have AORs covering a default search region, and Level 2 mosaics apparently covering a default search region, but because of the “apply spatial constraints to individual data products” option, the Level 1/BCD images are not returned.
Here is the real-life case that prompted this example. Search by position with the defaults in place for NGC 2264, asking it to return the AOR, Level 1, and Level 2 tabs. These observations are single pointings (not mosaics), for the purposes of monitoring young stars in this cluster. Thus, the 3.6 micron channel always points at the same portion of sky and the 4.5 micron channel points at a slightly different portion of sky, and the observation does not dither back to make sure that there is 3.6 micron coverage of the original 4.5 micron field of view.
Pick any of the PC data AORs, such as 29340160 or 29347328 or 29347072. Click "restrict data in other tabs". Click on Level 2. See two mosaics there. Click on the 3.6 um one, with "restrict data in other
tabs" still checked. Click on Level 1, and there are BCDs there. Go back to Level 2. Unclick the 3.6 um one and click the 4.5 um one. Go back to Level 1. "There are no data to display."
This is because the 4.5 BCDs are just outside of the search radius. These data were obtained as part of the observation, and the resultant pipeline mosaic includes the search position, but that search position does not actually include any 4.5 micron data.
Of course, searching by AORKEY does NOT reproduce this, because searching by AORKEY returns the whole AOR. Searching using a larger radius will recover the 4.5 micron data. Most non-monitoring observations tiled the larger area such that most of the sky covered by 3.6 microns is also covered by 4.5 microns, and thus BCDs would be returned for both channels, appropriately limited by position. BUT in this particular case, because this was a single field, monitored in one IRAC channel, you can get this seemingly inconsistent result, where there is a 4.5 micron mosaic but ‘no BCDs’.
Click on that to bring up a set of image manipulation tools:
Click on that to bring up a set of image manipulation tools. Click on the “spider webby” icon to add coordinates. You can change the color map and stretch; click on the icon that looks like a color cube to change the colors and the one that looks like a squashed color cube to change the stretch. 
