How to calculate probability of detection for an air search sector
POD gets tossed around in briefings like everyone already agrees on the number, but the math underneath it is simple enough to do on a kneeboard. What takes the time is getting the inputs right, and that's usually where a sector gets re-flown.
The coverage factor math
Probability of detection for a given sector starts with coverage factor, not POD itself. Coverage factor (C) is sweep width (W) divided by track spacing (S):
C = W / S
Sweep width comes from the search object type, the sensor (eyeball, EO/IR, radar), and conditions on scene: visibility, sea state or ground cover, and light. Track spacing is whatever your search pattern flew, parallel track, creeping line, whatever the sector called for. Once you have C, you don't calculate POD from scratch. You pull it off the lookup curve in the National SAR Supplement (or whatever planning tool your unit uses), which maps coverage factor to POD for that detection method. A coverage factor of 1.0 on a visual search typically lands POD somewhere in the 60-70% range on the curve, not 100%, because detection isn't linear with coverage. That's the part that trips up new coordinators: doubling your coverage factor does not double your POD.
Weather degrades sweep width before it degrades anything else. Recompute sweep width for the conditions the crew flew, especially if the ceiling dropped or haze rolled in mid-sortie. A sortie flown in deteriorating visibility with the briefed sweep width will overstate POD if you don't adjust.
From POD to the POS call
POD alone doesn't tell you whether to stand the sortie down or send it back out. That's where probability of containment (POC) comes in, your confidence that the search object was inside the sector in the first place, based on the datum and drift calculation. Multiply the two:
POS = POD × POC
A high POD on a sector built from a shaky datum still gives you a weak POS. Some planning references bundle this whole chain under a PODAR-style designator, POD, area, and the resulting assessment rolled into one line on the planning worksheet, but the underlying arithmetic is the same POD × POC relationship whatever your unit calls the row.
Cumulative POS across all sorties flown on that sector is the number that drives the next decision, more than any single sortie's POD. If POS is still below your unit's threshold after the flight, the sector gets re-searched, usually at a tighter track spacing to push coverage factor up. If POS clears threshold and nothing turned up, that sector gets marked searched and effort shifts.
Where this gets slow in practice is the debrief itself. You've got a POD number on paper from the planning math, and then you've got a few thousand frames of imagery from the sortie that someone has to look at before anyone signs off that the sector was really covered the way the math says it was. Scrolling that frame by frame after a long sortie is where coordinators lose the hours they don't have. That's the gap Search & Rescue Imagery is built for: it runs detection and similarity search over the flown imagery and hands back a short, ranked list of candidates instead of the whole folder, so the visual confirmation catches up to the POD math instead of lagging behind it.
If your sector debrief is still a frame-by-frame scroll while the POD number sits finished on the planning sheet, it's worth seeing what a ranked candidate list does to that gap.