What is a datum point in search and rescue planning?
A datum point is the position your search sectors get drawn around, the spot the planning math puts the subject now after wind, current, and drift have been applied to the last known fix. The last known position or last seen point is a historical fact. Datum is a forward projection from it, recalculated as time passes and conditions change.
That distinction trips up people outside the SAR world, and it trips up new planners too. Here's how the three terms stack up against each other.
LKP, PLS, and datum aren't the same point
Last known position (LKP) is the last spot someone can confirm the subject was actually seen, by a witness, a GPS ping, a cell tower hit, a check-in log. It's a fact, not an estimate.
Point last seen (PLS) is close to LKP but used more in land SAR, usually tied to a physical sighting rather than an electronic one. A hiker's last confirmed trailhead sign-in, a vehicle spotted at a pullout.
Datum is neither of those. It's what you get when you take the LKP or PLS and move it forward in time, accounting for everything that could have shifted the subject's position since that last fix: current set and drift for a vessel or person in water, wind and leeway for a life raft or debris field, terrain and likely travel direction for a lost hiker. Datum is a calculation, updated as the elapsed search time number climbs. An LKP from six hours ago and the same LKP from eighteen hours ago produce two very different datum points, because the drift correction keeps compounding.
Plot your sectors against the LKP instead of datum and you're flying a pattern centered on the subject's last known spot while the water or terrain has already carried them somewhere else. On a water case with any current running, that gap can put your whole search box in the wrong stretch of coastline.
Why datum drifts, and what that does to your search sectors
For a water search, drift gets worked up from total water current (the combination of ocean/tidal current and wind-driven surface current) plus leeway, which is the wind's own push on whatever's in the water, since a swimmer, a capsized hull, and a life jacket all catch wind differently and drift at different rates and headings than the current alone would carry them. Run that math forward from the LKP and you get a datum point, plus a radius of uncertainty around it that grows with every hour of elapsed time.
That growing uncertainty circle is the part that matters for sector planning. It's why a case that starts as a tight, high-confidence box at hour two can turn into a sprawling search area by hour fourteen if the first few sorties don't turn anything up and datum gets recalculated. Coordinators who don't recompute datum as new information comes in, a second witness sighting, updated wind data, a debris recovery, end up flying sectors that drifted out from under the actual search area hours ago.
When a sortie turns up debris with a solid GPS fix, maybe a section of hull or a life jacket, that find becomes a new anchor point. The planner backtracks the drift calculation from it and tightens up datum for the next sector draw. The faster that hit gets flagged and fed back into the plan, the less time the next sortie spends covering water or ground that's already been ruled out.
Which is usually the bottleneck in practice. The flight gets flown, the imagery comes back, and someone has to find the one frame that matters out of thousands before that redraw can happen. Search & Rescue Imagery runs detection and similarity search over a sortie's imagery and hands back a short, ranked candidate list instead of a folder nobody has time to scroll through, so a confirmed hit gets back into the datum math while it still matters.
If your bottleneck is turning flown imagery into a usable debrief fast enough to redraw the next sector, that's the problem this tool is built around.