Risk and prioritization · Lesson 4

Updating a risk register when evidence and footprints conflict

5 min read

Lesson objectives

  • Align footprint and observation time before comparing exposure, then record changed dependencies, uncertainty and the next verification decision.
Reading tools

A new map does not make the old register disposable, and adding old and new counts does not create a fact. First define the unit: household, person, facility or service request. Then identify the footprint, observation period, counting method and source of each record. A modelled inundation footprint and a field report of water seen today are different evidence types; neither automatically replaces the other. If a polygon boundary changes, its household count cannot be compared with the earlier count as if both covered the same population. Retain version, time and owner, and use a spatial intersection or crosswalk before describing a trend. Label a value verified within a defined footprint, conditional pending reconciliation, or unavailable. Silence is not zero exposure.

For a cascading risk, draw at least one service dependency. A blocked road might prevent maintenance from reaching a pump, weaken drainage later and prolong access disruption at a clinic. It does not prove that the pump has failed. Give each link a trigger, evidence needed and person responsible for checking it. If a new observation challenges an old assumption, distinguish when the team learned of it from when the condition changed, if known. Do not reorder priorities by changing the colour of a cell. Ask which decision could change now, which quick check would reduce consequential uncertainty, and who can authorize action. A useful learning indicator is the share of decisions with a defined footprint, time and source, alongside a list of unresolved checks, rather than the number of rows edited.

Worked solution: create two versioned rows rather than describing 185−140=45 as new event exposure. Record old A as 140 households under version one's boundary and new A as 185 under version two, with each map time. Mark numeric comparison invalid until boundaries and common households are reconciled. Do not add the two counts or mix them with the clinic's visitor count. For B, record 'water observed on road 08:10; maintenance-vehicle passage unverified' and 'pump running per technician 08:25; maintenance access untested.' The possible chain is road water, failed maintenance access and a later pumping-capacity loss; only the first condition is observed. Ask for a quick safe route and vehicle-access check, including an alternate, because that finding could change reliance on the pump. Request a crosswalk from mapping staff showing common households, entries and exits before interpreting the difference of 45. Neither footprint has a proven final risk ranking: event frequency, consequence and actual access remain uncertain. Deliver a time-stamped decision row distinguishing observation, inference, open check and owner.

Lesson summary

Record footprint versions and maintenance-vehicle access limits.