Category Values & Cross-Source Mappings

This file contains every category value for each characteristic, organized by data source. Where two sources describe the same characteristic using different terminology, a cross-source mapping table shows how values relate to each other.

Why mappings matter: The HeatWaves app consolidates data from sources that were never designed to interoperate. A mapping marked ≈ partial means the concepts overlap but aren't identical — the app must handle these carefully, either by grouping them or flagging the ambiguity for the user.

Mapping notation

• ✓ exact — same concept, identical or trivially-identical naming (e.g. spelling-only differences)
• ≈ partial — overlapping concept, but with meaningful differences in granularity, framing, or scope. Trivial spelling/spacing differences across sources (e.g. Fellesareal vs Felles areal) are also treated as ≈ partial because they are technically distinct strings even when semantically identical
• ✗ source only — no equivalent in the other source

Table of Contents

1. Cross-Characteristic Overlaps (Heat Source ↔ Primary Causal Factor)
2. Area of Origin (Arnested)
3. Item First Ignited (Objekt)
4. Heat Source (Tennkilde)
5. Primary Causal Factor (Årsak)

---

Cross-Characteristic Overlaps

Before diving into the per-characteristic mappings, it's important to understand a structural quirk of the source data: the boundary between Heat Source and Primary Causal Factor is not consistently drawn across sources. Several values appear as Heat Source in one source and as Primary Causal Factor in another. This is the single most important consolidation challenge in the data.

EU Firestat addresses this conceptually by reserving Heat Source for the energy source and Primary Causal Factor for the underlying reason — but the Norwegian source datasets predate this distinction, so the data is messy.

Values that blur the line

Implications for consolidation

When the HeatWaves app aggregates data across sources, it must decide whether these values are queried under Heat Source or Primary Causal Factor. Three viable strategies exist:

1. Keep them split per source — preserve raw fidelity, accept that comparing across sources for these specific values requires care.
2. Map to a unified characteristic — pick one (typically Heat Source per EU Firestat's framing, since these are sources of energy) and consolidate cross-source data there. Document the choice clearly.
3. Surface the ambiguity in the UI — let users see that the value lives in different characteristics depending on the source. This aligns with HeatWaves' core mission of transparency about source dissonance.

There is no objectively correct choice — the right answer depends on what the user is trying to learn from the data. Surface the trade-off rather than hiding it.

---

1. Area of Origin (Arnested)

Describes the room or area within the building where the fire started.

Sources: DSB, Police

DSB — Antatt arnested

Police — Arnested

Cross-Source Mapping: Area of Origin

Key differences: DSB has 25 categories, Police has 21. DSB includes 4 categories not tracked by Police: Hulrom i konstruksjon, Ventilasjonsrom/teknisk rom, Skorstein, and Badstue.

---

2. Item First Ignited (Objekt)

Describes the first object or material that caught fire.

Sources: Police, DSB

Police — Brannen startet i

DSB — Hva startet brannen/branntilløpet i?

Cross-Source Mapping: Item First Ignited

Key differences: Police uses 22 fine-grained categories; DSB consolidates into 11 broader groups. Most notably, Police splits electrical equipment into 3 types and furnishings/textiles into 8 types, while DSB groups each of these into a single category.

---

3. Heat Source (Tennkilde)

Describes the source of energy that initiates combustion in the item first ignited.

Sources: DSB, BRASK

See Cross-Characteristic Overlaps — Åpen ild, Selvantennelse, and Naturlig fenomen in this characteristic also appear as Primary Causal Factor values in other sources.

DSB — Antatt tennkilde

BRASK — Kilde

Cross-Source Mapping: Heat Source

Key differences: DSB uses 6 broad categories; BRASK uses 7 categories that are more specific for electrical sources (3 subtypes) but lack DSB's Selvantennelse and Naturlig fenomen. The granularity is inverted compared to Item First Ignited — here BRASK is more detailed for electrical, while DSB is more detailed for non-electrical phenomena.

---

4. Primary Causal Factor (Årsak)

Describes the factor most important in explaining why the item first ignited was exposed to the heat source in a way that led to uncontrolled combustion. (See taxonomy-and-eu-firestat-alignment.md for the EU Firestat-aligned definition.)

Sources: Police, Fire Brigade, BRASK — three sources, three different terminologies.

See Cross-Characteristic Overlaps — Åpen ild, Selvtenning/Selvantennelse, and Naturlige fenomener/Lynnedslag in this characteristic also appear as Heat Source values in DSB.

Police — Brannårsak overordnet

Fire Brigade — Brann og redningsvesenets vurdering av hvordan brannen kan ha startet

BRASK — Årsak

Cross-Source Mapping: Primary Causal Factor

Key differences: This is the most fragmented characteristic. Police has 9 categories, Fire Brigade has 6, and BRASK has 7 — but they carve up the concept space differently. The electrical/technical failure area is particularly messy: Police uses one broad Elektrisk årsak, Fire Brigade splits into equipment vs. installation faults, and BRASK offers both Teknisk svikt and Elektrisk fenomenskade with unclear boundaries between them. Arson terminology also varies in certainty level across sources.