Astronomy · How new tools extend what we can know

Survey networks and coordinated messenger observations

Distinguish total coverage from a complete coordinated test, then calculate how alert receipt and setup affect an original two-facility observing window.

Continue in the approved chapter order. Necessary vocabulary, quantities and assumptions are explained in the lesson.

Back to the chapter

After reading, you can

Separate union coverage, simultaneous availability and usable coordinated measurements.
Calculate whether a complete stipulated exposure fits after alert receipt and setup.
Identify the timing, coordinate, response and revision information needed to interpret an event alert and a nondetection.

A collection of instruments is not always one instrument

An observing network can extend sky coverage, observing time, wavelength range or access to different messengers. Those gains address different scientific questions. Combining facilities does not make every location continuously visible or give every detector the same response.

A joint analysis needs to specify what information is required together. Some questions need overlapping exposures to compare a changing spectrum. Others can use measurements at different times or intrinsic messenger delays. Simultaneity is a scientific condition for a particular test, not a universal rule for multi-messenger astronomy.

Response and calibration remain attached to each instrument. A telescope that did not detect a counterpart may have lacked coverage, sensitivity or the relevant band. An event’s localization uncertainty can require several pointings, and weather, acquisition and data quality can reduce otherwise available time.

The original exercise below deliberately strips away those complications to isolate availability and readiness. Its intervals are invented times in one uniform minute coordinate measured from a stipulated event reference. They are not real observatory schedules, measured network latency or a model of Earth rotation.

Total time and shared windows answer different questions

Let facility A be available over [1, 4] and [7, 10] minutes, while facility B is available over [3, 8] minutes. A has six minutes of availability and B has five. Their union is [1, 10], lasting nine minutes; overlapping time must not be counted twice.

Their intersection is [3, 4] together with [7, 8], lasting two minutes in total. This is when both facilities are available simultaneously. The union’s nine minutes does not imply nine minutes of a shared measurement.

For this specific invented scientific test, require a full continuous one-minute exposure from each facility at the same time, entirely during an active phase [0, 6] minutes. Under these exact conditions only the first joint window can contain the required test.

The second shared window [7, 8] has enough simultaneous duration but starts after the stipulated phase ends. Adding the two separated shared windows also does not create one continuous exposure. Duration, continuity and scientific timing are separate requirements.

Original availability intervals in one declared minute coordinate. Endpoint inclusion is stipulated; the full exposure may start at 3 and finish at 4.
Availability quantity	Intervals (min)	Total duration (min)
A	[1, 4] and [7, 10]	6
B	[3, 8]	5
Either available, A union B	[1, 10]	9
Both available, A intersection B	[3, 4] and [7, 8]	2
Stipulated active phase	[0, 6]	6

An alert must leave time for the complete measurement

Stipulate that both facilities require 0.25 minute of setup after alert receipt, with no other delay or scheduling cost in this abstract model. If the alert is received at r, they are ready at r + 0.25. An exposure start s must be no earlier than that readiness time.

To fit the full one-minute exposure within [3, 4], the start must be exactly s = 3. Readiness must therefore be at or before 3, requiring r ≤ 2.75 minutes. This is the first-window receipt deadline under the supplied setup and exposure rules.

Receipt at 2.50 gives readiness at 2.75, so the facilities can wait until 3 and expose through 4. Receipt at 2.75 gives readiness exactly at 3 and also fits. Equality is allowed by the declared closed-interval convention.

Receipt at 3.00 instead gives readiness at 3.25, leaving only 0.75 minute of the first window. A complete shared one-minute exposure first fits at [7, 8], which is outside the active phase. Receipt at 6.50 also allows [7, 8] after readiness at 6.75, with the same scientific limitation.

Original alert/readiness cases. The earliest complete shared exposure refers to availability; the last column separately checks the scientific active phase.
Receipt r (min)	Ready (min)	Earliest complete shared exposure (min)	Entirely within [0, 6]?
2.50	2.75	[3, 4]	Yes
2.75	3.00	[3, 4]	Yes
3.00	3.25	[7, 8]	No
6.50	6.75	[7, 8]	No

Interpret the alert before using its numbers

An event measurement time identifies when the reported observation occurred under its time convention. Packet creation time identifies when the report was assembled. Receipt time identifies when a subscriber obtained it. Confusing those stages can incorrectly suggest an intrinsic source delay or an observing opportunity that had already passed.

The IVOA VOEvent specification separates packet curation from observation coordinates and times. Its roles also distinguish observations, predictions and tests. A test message should not become evidence for an actual celestial event merely because it has a realistic timestamp and sky coordinate.

A scientifically interpretable alert needs units, time scale and reference location, coordinate frame, localization and timing uncertainty, and relevant instrument context. A numerical classification probability is a supplied inference that needs its model and calibration; a field in a standard format does not independently validate it.

Persistent identifiers and links to follow-up, superseding reports or retractions preserve context as information changes. The metadata standard defines meaning, rather than guaranteeing transport speed, successful follow-up or physical association. Two packets can also reuse the same observations or calibration; their existence alone does not establish statistical independence.

A missed interval limits a nondetection

In the receipt-at-3 case, a nondetection during [7, 8] cannot by itself exclude emission confined to the earlier [0, 6] phase. The required coordinated test was missed. It may still constrain another specified late-emission hypothesis if the instrument response and detection capability support that comparison.

Even receipt before 2.75 guarantees only the abstract scheduling condition in this model. Real visibility, response, localization coverage, weather, setup variability and data quality could prevent an adequate measurement. A useful coordination result states which conditions were modelled and which remain untested.

Provenance links the alert and later products to observations, calibration and processing. It helps distinguish a revised interpretation of the same data from new measurements. A coherent event record retains nondetections and actual observing windows alongside detections rather than treating every absence as an equally strong constraint.

A network increases scientific capability when it creates relevant, calibrated observations that can be compared with serious alternatives. Timing coincidence alone does not establish a common origin. The next lesson examines the supporting calibration, archives and numerical representation needed to preserve what such observations actually contain.

Worked examples

Find the first-window deadline

When must the alert arrive to fit a one-minute exposure in [3, 4] after 0.25 minute of setup?

The exposure must start at 3 to finish by 4.
Require readiness r + 0.25 ≤ 3.
Solve for the latest allowed receipt time.

r ≤ 2.75 minutes under the declared exact scheduling assumptions.

Test receipt at 3 minutes

Can receipt at 3.00 complete the stipulated shared test during [0, 6]?

Readiness is 3.25, leaving only 0.75 minute before the first window ends.
The next full shared exposure is [7, 8].
That interval lies outside the active phase.

No. A full shared observation remains possible later, but it misses this particular phase-dependent test.

Try the reasoning

Why does 6 minutes for A plus 5 for B give a union of 9 rather than 11?
Show the answer and reasoning
Their two minutes of overlap would otherwise be counted twice.
Union duration is 6 + 5 − 2 = 9 minutes.
Is [7, 8] a complete shared one-minute window in the supplied availability?
Show the answer and reasoning
Yes.
It satisfies shared duration but not the separate active-phase condition.
Is packet creation time automatically the event measurement time?
Show the answer and reasoning
No.
They identify different stages and need their respective conventions.
Can a [7, 8] nondetection alone rule out a signal confined to [0, 6]?
Show the answer and reasoning
No.
The relevant earlier interval was not covered by that observation.

What to carry forward

Union coverage, simultaneous availability and a usable coordinated test are different quantities.
Readiness and complete exposure duration must fit the scientific observing window.
Alerts and nondetections need timing conventions, response, coverage, uncertainty and provenance.

References

Teaching resources and research supporting this lesson’s explanations and examples.