Search sector planning: how to divide a search area so nothing gets missed

Every search starts with the same argument. One person wants teams on the ground at the last known point, right now. Another wants twenty more minutes with the map. The second person is usually right, and the first person usually wins.

Sectors are how that argument gets settled without anyone losing. They turn an area that is too large to think about into a list of tasks that a coordinator can assign, track and close. Done well, they are the difference between an operation and a crowd of volunteers walking through a forest. Done badly, they produce a map that looks organised and a search that is not.

This is how experienced coordinators build them, and the parts that training courses tend to skip.

01First the boundary, then the sectors

Before you divide anything, you have to decide how much there is to divide. That is the search area boundary, and it is a different kind of decision from everything that follows: it is statistical, and it belongs to the person who understands the subject, not the terrain.

Start from the initial planning point — usually the place last seen or the last known position. Then ask what category of person you are looking for. A dementia patient who walked out of a care home, a hiker overdue on a marked trail, a child, a despondent adult and a hunter each move in a completely different way, and the research on lost person behaviour gives you distance rings for each of them. Use the data your national organisation trains on. If you do not have national data, ISRID-based tables are the standard reference, and they are far better than the number someone feels is right.

Now trim the boundary with terrain. A cliff band, a river in flood, a motorway with a fence — these are barriers, and a subject who did not cross one does not need to be searched for on the other side. Roads, ridgelines and streams are also handrails: features a disoriented person tends to follow. Handrails pull probability towards them. Barriers push it away. That reshaping matters more than the radius you started with.

02What makes a sector a good sector

A sector is not a shape. It is a task, and the shape is just how you describe it. The single most useful rule is this: one team should be able to search one sector to the assigned standard within one operational period, including the walk in and the walk out.

That rule silently constrains everything else. A four-hour period, a team of five, and steep beech forest with heavy undergrowth is a very different sector than the same team on open pasture. In practice the difference is not a factor of two, it is a factor of five or more. Coordinators who size sectors in hectares on a screen consistently get this wrong; coordinators who size them in hours get it right.

Boundaries a tired person can see

Use streams, ridges, forest edges, tracks and fence lines. A sector edge drawn as a straight line across a slope exists on the map and nowhere else, and the two teams on either side of it will both assume the other one covered the gap.

Terrain, not geometry

A sector that spans a ridge is really two sectors. The team drops into one drainage, works it, and never realistically climbs out to do the other with the same energy or attention.

One access route, thought through

Note how the team gets in and out, and how long that takes. Twenty minutes of approach at each end is nearly an hour of a four-hour task.

Numbered, not named

Short, unambiguous identifiers over the radio. "Sector 7" survives a bad connection. "The area below the old sheepfold" does not.

03Ranking: which sector gets your best team

Once the area is divided you have to decide the order, and this is where most operations quietly lose time. The instinct is to start with the sector closest to the command post, or the one the first arriving team can reach fastest. Sometimes that is also the most likely sector. Usually it is not.

The tool for this is probability of area — an estimate, for each sector, of the chance the subject is inside it. You will never know these numbers. That is fine; you are ranking, not measuring. The classic way to produce them with a group is the Mattson consensus method: each experienced person on the planning team independently distributes 100 points across the sectors, and you average the results. It takes about ten minutes and it is remarkably effective at surfacing the sector that one quiet person on the team has been worried about since the beginning.

Then combine that ranking with effort. A sector with 15% probability that a hasty team can cover in forty minutes may well go before a 30% sector that will take two teams a full period. You are trying to maximise probability found per hour, not to work down a list in order.

The first two hours are worth more than the next ten. Spend them on the sectors that would embarrass you if the subject were found there tomorrow.

04What "searched" actually means

A team comes back and reports the sector as clear. What have you learned? Depending on how they searched it, somewhere between "the subject is almost certainly not there" and "the subject could have been fifteen metres away, asleep under a spruce, and we would have walked past."

This is probability of detection — the chance a team finds the subject given that the subject is in the sector. It depends on spacing between searchers, vegetation, light, weather, the subject's ability to respond, and how tired everybody is. A hasty search along trails and drainages is fast and has a low POD. A tight grid with searchers a few metres apart is slow and has a high one. Neither is wrong. They are different tools.

The practical consequence: never record a sector as a binary. Record the search type, the number of searchers, the spacing, the conditions, and the team leader's honest estimate of coverage. Then a sector searched at low thoroughness stays on the board as a candidate for a second pass, at higher effort, if the high-probability sectors come up empty.

Click to enlarge

05Recording it so it survives the night

Operations that run past dark have a shift change, and shift changes are where information dies. The incoming coordinator inherits a map covered in marker, a stack of task forms in three handwritings, and the previous coordinator's memory, which is leaving in ten minutes to sleep.

Everything about a sector should be recorded the moment it is decided: who is assigned, when they went in, what search type, what came back. If you are working on paper, use one numbered task form per sector and never let a team leave without one. If you are working in software, the discipline is the same — the software only makes it faster to be honest.

The record is also what you hand to whoever asks questions afterwards. Sometimes that is a debrief. Sometimes it is a family. Occasionally it is a court. "We searched that area" is not an answer. "Sector 4, two teams, grid at eight metres, 14:20 to 17:05, coverage estimated high, GPS track attached" is.

06Five mistakes worth naming

• Sectors sized for the map, not for the team. If a sector takes seven hours, the team will cut corners in hour five and report it clear anyway.
• Searching the easy ground first because it is easy. It feels productive. It is the most common way to lose a day.
• Treating a hasty search as a completed search. Hasty coverage is a first pass. Mark it as one.
• Boundaries drawn as straight lines. Every straight sector edge on a slope is a gap waiting to happen.
• No record of what was not searched. The sectors you skipped are as important as the ones you cleared, and by hour eighteen nobody remembers which was which.

None of this requires software. Sectors were planned well for decades with acetate, a chinagraph pencil and a good coordinator. What software changes is the cost of keeping the record accurate while the operation is moving — and that cost is exactly what gets abandoned first when things get busy.