Native API Comparison — Smoke Test Datasets

Side-by-side reference of how the native datasource APIs look for a representative set of HeatWaves requests. Datasets 1–11 cover the matrix for BRIS/BRASK smoke tests (dataset 12 is an extra BRIS case). Each example pairs the HeatWaves intent with the exact upstream request body or form. Together with category-native-to-cat-code.md and filter-axes-native-to-filter-id.md, these eleven examples are designed to be sufficient for any HeatWaves ETL developer to compose a working native request for any (datasource, dimension, filter) combination supported by the platform.

Every numeric id, parameter name, and value below is sourced from one of:

category-native-to-cat-code.md — cat_code ↔ native_api_code mapping
filter-axes-native-to-filter-id.md — filter parameter names and values per source
lib/data-model/adapters/_brisRequestBody.ts — BRIS request builder (buildBrisRequestBody)
lib/data-model/adapters/_brisUpstream.constants.ts — BOLIG_IDS, NON_RESIDENTIAL_MATRIKKEL_LEAF_IDS, BRIS_QUESTION_AGGREGATIONS, TIME_OPTIONS
lib/data-model/adapters/brask-passthrough.ts — BRASK form mechanics (shared constants in lib/data-model/adapters/_braskWebForms.constants.ts: BRASK_FILTER_FIELD_MAP, BRASK_AXIS / braskExplorerAxisId, BRASK_UTFORSK_RBL_*)
ssb-pxwebapi-tables.md — SSB table 12058 variable reference

Unless otherwise noted, every example is a time series with years on the x-axis, antall (or kr) on the y-axis, and categories within the chosen dimension as series.

Filter vocabulary in HeatWaves intent lines — per filter-axes-native-to-filter-id.md: — = axis omitted / unfiltered upstream slice (NULL *_code FK on facts; no filter-row standard_code). RESIDENTIAL, DEVELOPED, etc. = resolved filters.standard_code. none / n/a = axis not available on that upstream endpoint. Do not use ALL or ALL_BUILDINGS here — those codes belong on categories (cross-tab dimension totals), not the filter axes.

Shared API Mechanics

Three upstream endpoints are exercised across the eleven datasets:

Source	Endpoint	Body type
BRASK	`POST https://brask.finansnorge.no/`	`application/x-www-form-urlencoded` (ASP.NET form)
BRIS — Alle branner og ulykker	`POST https://brannstatistikk.no/api/v1/missionreports/aggregate`	`application/json` (full filter body)
BRIS — Hvordan bygningsbranner starter	`POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation`	`application/json` (minimal body)
SSB	`POST https://data.ssb.no/api/v0/no/table/12058`	`application/json` (PxWebApi v0 query, returns json-stat2)

All BRASK requests must first GET https://brask.finansnorge.no/ and include the __VIEWSTATE, __VIEWSTATEGENERATOR, and __EVENTVALIDATION hidden fields from the response. Those three fields are omitted from every example below for readability.

Dataset 1 — Police, Arnested, Utviklet brann, Bolig

HeatWaves intent

datasource: BRIS — Police (Sub-A, "Alle branner og ulykker")
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = RESIDENTIAL

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "hitsToReturn": 0,
  "skipped": 0,
  "missionTypes": { "ids": [], "isMissingValue": false },
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "monthNumbers": { "ids": [], "isMissingValue": false },
  "dayOfWeek": { "ids": [], "isMissingValue": false },
  "responseTimeLimitGreaterThan": true,
  "callOutTimeLimitGreaterThan": true,
  "effortDurationGreaterThan": true,
  "counties": { "ids": [], "isMissingValue": false },
  "municipalities": { "ids": [], "isMissingValue": false },
  "basicStatisticalUnits": { "ids": [], "isMissingValue": false },
  "fireDepartments": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "streetAddress": "",
  "geoShapeFilter": { "polygons": [], "circles": [] },
  "fireStations": { "ids": [], "isMissingValue": false },
  "dispatchedFireDepartments": { "ids": [], "isMissingValue": false },
  "buildingTypes": {
    "ids": [
      "111","112","113","121","122","123","124","131","133","135","136",
      "141","142","143","144","145","146","151","152","159","161","162",
      "163","171","172","181","182","183","193","199"
    ],
    "isMissingValue": false
  },
  "naceCodes": { "ids": [], "isMissingValue": false },
  "includeAssistanceMissions": false,
  "includeExercises": false,
  "includeMissionsHandledByHundredAndTen": false,
  "includePoliceCauseWithNoMission": false,
  "includeRVRMissionsFromBris": false,
  "includeApprovedMissions": true,
  "onlyApprovedMissions": true,
  "answers": [],
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — aggregationDefinition.id = "403" is Police-only on /aggregate. HeatWaves ingest sends one Matrikkel leaf id per slice (buildingTypeIdsOverride); the 30-id Bolig body below is the upstream shape for a manual rollup request only — see filter-axes-native-to-filter-id.md. Leaf reference list: BOLIG_IDS in _brisUpstream.constants.ts. Result rows are keyed by 403_* answer ids — see category-native-to-cat-code.md §4 (cat_codes 401-421).

Dataset 2 — DSB, Arnested, Utviklet brann, no building type

HeatWaves intent

datasource: BRIS — DSB ("Hvordan bygningsbranner starter")
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none (not supported on this endpoint)

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "227",
      "name": "Antatt arnested (startsted)",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — aggregationDefinition.id = "227" is restrictedOnly. The restricted body has no buildingTypes field at all (see filter-axes-native-to-filter-id.md §2 — Building Type is unsupported on /restrictedaggregation). Result rows are keyed by 227_* answer ids — category-native-to-cat-code.md §4 (cat_codes 422-448).

Dataset 3 — Police, Objekt, Utviklet brann, Bolig

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 3 (Objekt) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = RESIDENTIAL

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

Same envelope as Dataset 1 with two changes:

{
  "...": "[same as Dataset 1, except:]",
  "buildingTypes": { "ids": ["...30 BOLIG_IDS..."], "isMissingValue": false },
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "405",
      "name": "Brannen startet i - politi",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — aggregationDefinition.id = "405" is fullOnly. Result rows are keyed by 405_* answer ids — category-native-to-cat-code.md §3 (cat_codes 301-322).

Dataset 4 — Police, Objekt, Utviklet brann, all buildings

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 3 (Objekt) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = — (unfiltered — all buildings)

Native API

Identical to Dataset 3 except buildingTypes.ids is empty:

{
  "...": "[same as Dataset 3, except:]",
  "buildingTypes": { "ids": [], "isMissingValue": false }
}

Notes — Compare side-by-side with Dataset 3 to see the building-type filter on/off delta cleanly. Empty buildingTypes.ids array means "all building types" per filter-axes-native-to-filter-id.md §2.

Dataset 5 — DSB, Tennkilde, Begrenset branntilløp, all buildings

HeatWaves intent

datasource: BRIS — DSB ("Hvordan bygningsbranner starter")
dimension: dim_code = 2 (Tennkilde) — series
x-axis: year
y-axis: antall
filters: fire_type = CONFINED, building_type = none (not supported on this endpoint)

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10086, 10212], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "264",
      "name": "Antatt tennkilde",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — revisedMissionTypes.ids = [10086, 10212] covers Branntilløp komfyr + Branntilløp i bygg annet. Brann i skorstein (10175) is omitted here — it does not exist in the restricted "Hvordan bygningsbranner starter" dataset (see filter-axes-native-to-filter-id.md §1). Result rows are keyed by 264_* — category-native-to-cat-code.md §2 (cat_codes 208-213).

Dataset 6 — BRASK, Tennkilde, Begrenset, Bolig, Oslo, alder 1–5, value=kr

HeatWaves intent

datasource: BRASK
dimension: dim_code = 2 (Tennkilde / Kilde) — series
x-axis: year
y-axis: erstatningsbeløp (kr)
filters: fire_type = CONFINED, building_type = RESIDENTIAL, region = Oslo, building_age = 1-5

This is the most filter-heavy BRASK example — exercises every BRASK filter mechanic in one request.

Native API

POST https://brask.finansnorge.no/
Content-Type: application/x-www-form-urlencoded

__VIEWSTATE=<from initial GET>
__VIEWSTATEGENERATOR=<from initial GET>
__EVENTVALIDATION=<from initial GET>
ctl00$Innhold$ddlRad=5
ctl00$Innhold$ddlKolonne=7
ctl00$Innhold$rblVerdi=2
ctl00$Innhold$rblBeregning=1
ctl00$Innhold$lbBransje=
ctl00$Innhold$lbAldersgruppe=2
ctl00$Innhold$lbFylke=3
ctl00$Innhold$lbDag=
ctl00$Innhold$lbKilde=
ctl00$Innhold$lbKvartal=
ctl00$Innhold$lbMåned=
ctl00$Innhold$lbNæring=0
ctl00$Innhold$lbType=2
ctl00$Innhold$lbUkedag=
ctl00$Innhold$lbÅr=
ctl00$Innhold$lbÅrsak=
ctl00$Innhold$btnKjør=Lag tabell

Notes — Field-by-field justification:

ddlRad=5 — row dimension = kilde (Utforsk ddlRad index = BRASK_AXIS.kilde (5) in _braskWebForms.constants.ts; not fire_data.categories.dim_code).
ddlKolonne=7 — column dimension = ar (BRASK_AXIS.ar = 7).
rblVerdi=2 — value type = erstatningsbelop (kr); 1 would be antallSkader.
lbAldersgruppe=2 — native value 2 = "1-5" per filter-axes-native-to-filter-id.md §3 (normalizes to HeatWaves bucket 0-5).
lbFylke=3 — Oslo per filter-axes-native-to-filter-id.md §4.
lbNæring=0 — Beboelse = RESIDENTIAL per filter-axes-native-to-filter-id.md §2.
lbType=2 — Kald = CONFINED per filter-axes-native-to-filter-id.md §1.
All other lb* fields sent empty (= Alle / no filter).

Dataset 7 — Brigade, Årsak, Komfyr-only confined, all buildings

HeatWaves intent

datasource: BRIS — Fire Brigade ("Hvordan bygningsbranner starter")
dimension: dim_code = 1 (Årsak) — series
x-axis: year
y-axis: antall
filters: fire_type = CONFINED — komfyr only, building_type = none

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10086], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "269",
      "name": "Brannvesenets vurdering",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — Single-mission-type slice; the narrowest fire-type filter possible on BRIS. Useful as an ETL ingestion canary because the row count is small and easy to verify by hand. aggregationDefinition.id = "269" is restrictedOnly. Result rows are keyed by 269_* — category-native-to-cat-code.md §1 (cat_codes 117-122).

Dataset 8 — BRASK, Årsak, all fire types, all buildings, Oslo, value=kr

HeatWaves intent

datasource: BRASK
dimension: dim_code = 1 (Årsak) — series
x-axis: year
y-axis: erstatningsbeløp (kr)
filters: fire_type = — (unfiltered), building_type = — (unfiltered), region = Oslo

Native API

POST https://brask.finansnorge.no/
Content-Type: application/x-www-form-urlencoded

__VIEWSTATE=<from initial GET>
__VIEWSTATEGENERATOR=<from initial GET>
__EVENTVALIDATION=<from initial GET>
ctl00$Innhold$ddlRad=6
ctl00$Innhold$ddlKolonne=7
ctl00$Innhold$rblVerdi=2
ctl00$Innhold$rblBeregning=1
ctl00$Innhold$lbBransje=
ctl00$Innhold$lbAldersgruppe=
ctl00$Innhold$lbFylke=3
ctl00$Innhold$lbDag=
ctl00$Innhold$lbKilde=
ctl00$Innhold$lbKvartal=
ctl00$Innhold$lbMåned=
ctl00$Innhold$lbNæring=
ctl00$Innhold$lbType=
ctl00$Innhold$lbUkedag=
ctl00$Innhold$lbÅr=
ctl00$Innhold$lbÅrsak=
ctl00$Innhold$btnKjør=Lag tabell

Notes

ddlRad=6 — row dimension = arsak (BRASK_AXIS.arsak = 6 in _braskWebForms.constants.ts).
ddlKolonne=7 — column dimension = ar (year).
rblVerdi=2 — kr.
lbFylke=3 — Oslo only.
lbType= and lbNæring= empty — all fire types / all building types.
The only example showing kr value-type without fire-type or building-type narrowing — useful as a "broad financial loss by cause" smoke test.

Dataset 9 — BRIS Police, Arnested, Utviklet brann, all buildings, Oslo

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = — (unfiltered — all buildings), region = Oslo

Direct cross-source counterpart to Datasets 6 and 8 — shows how Oslo is filtered on the BRIS side.

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "...": "[same as Dataset 1 envelope, except:]",
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [3], "isMissingValue": false },
  "buildingTypes": { "ids": [], "isMissingValue": false },
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — cross-system caveat

BRASK lbFylke = 3 and BRIS counties.ids = [3] both happen to equal Oslo, but the systems align only by coincidence for Oslo. Every other county diverges:
- BRASK uses pre-2020 cresta-sone (e.g. 2 = Akershus, 7 = Vestfold, 19 = Troms).
- BRIS uses SSB klass 104 (e.g. 30 = Viken for 2020-2023 incidents, 32 = Akershus for post-2024 incidents — see fylke-code appendix below).
For finer granularity, municipalities.ids = [301] would also work for Oslo in BRIS.

Dataset 10 — SSB, Boligbranner (RESIDENTIAL dataset), Oslo

HeatWaves intent

datasource: SSB (third datasource — first appearance in this doc)
dimension: total counts (no within-dimension series — SSB does not break down causal/object dimensions)
x-axis: year
y-axis: antall
filters: building_type = RESIDENTIAL (encoded by selecting the Boligbranner statistical variable, not a parameter), region = Oslo

Cross-source counterpart to Datasets 6, 8, 9 — same Oslo geography across all four sources.

Native API

POST https://data.ssb.no/api/v0/no/table/12058
Content-Type: application/json

{
  "query": [
    {
      "code": "KOKkommuneregion0000",
      "selection": { "filter": "item", "values": ["0301"] }
    },
    {
      "code": "ContentsCode",
      "selection": { "filter": "item", "values": ["KOSboligbranner0000"] }
    },
    {
      "code": "Tid",
      "selection": { "filter": "top", "values": ["10"] }
    }
  ],
  "response": { "format": "json-stat2" }
}

Notes

Table 12058 is the SSB KOSTRA fire and safety table. Variable codes are documented in ssb-pxwebapi-tables.md.
For the all-buildings total slice (building_type_code NULL; Bygningsbranner / KOSbygningsbrann0000) instead of RESIDENTIAL, swap the ContentsCode value accordingly. For kr instead of antall: KOSerstatnbygnin0000.
KOKkommuneregion0000: municipality codes (0301 = Oslo) or EAK for national total. PxWeb → region_code: filter-axes-native-to-filter-id.md §4; upstream codes: ssb-pxwebapi-tables.md.
Tid is years natively in SSB — no equivalent of BRIS's nextLevelAggregationDefinition is needed. The "top": "10" value means "the most recent 10 years"; replace with { "filter": "item", "values": ["2020","2021","2022","2023","2024"] } for an explicit list.
Response format json-stat2: row-major value array — decode contract in pipeline-slices-idempotency-and-triggers.md §4.3; PxWebApi mechanics in ssb-pxwebapi-tables.md.

Dataset 11 — BRIS DSB, Tennkilde, no time sub-aggregation

HeatWaves intent

datasource: BRIS — DSB
dimension: dim_code = 2 (Tennkilde) — series
x-axis: none (categories only — no time series)
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none

This is the only example in the doc that does NOT match the "year on x-axis" rule — included expressly to show the simpler payload shape for ETL slices that only need categorical totals.

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "264",
      "name": "Antatt tennkilde",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": []
    }
  }
}

Notes — Identical to Dataset 5 with two changes: revisedMissionTypes.ids = [10026] (DEVELOPED), and nextLevelAggregationDefinition is omitted entirely from the aggregationDefinition object. The response shape is then a flat result[0].buckets[<category>].count map (no time sub-buckets); the BRIS adapter parses that into fact_yearly rows for /explore charts via get_fire_data.

Dataset 12 — BRIS Police, Arnested, Utviklet brann, DAY granularity, single year window

HeatWaves intent

datasource: BRIS — Police (Sub-A, "Alle branner og ulykker")
dimension: dim_code = 4 (Arnested) — series
x-axis: date (one bucket per calendar day)
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none, period = 2024

This is the canonical day-grain BRIS request used by the daily ingest adapter (lib/data-model/adapters/bris.daily.ts). Tier-2 sends the same body with interval: "MONTH" per drifted year; Tier-3 sends this body with the period narrowed to a single (year, month).

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "hitsToReturn": 0,
  "skipped": 0,
  "missionTypes": { "ids": [], "isMissingValue": false },
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "monthNumbers": { "ids": [], "isMissingValue": false },
  "dayOfWeek": { "ids": [], "isMissingValue": false },
  "responseTimeLimitGreaterThan": true,
  "callOutTimeLimitGreaterThan": true,
  "effortDurationGreaterThan": true,
  "counties": { "ids": [], "isMissingValue": false },
  "municipalities": { "ids": [], "isMissingValue": false },
  "basicStatisticalUnits": { "ids": [], "isMissingValue": false },
  "fireDepartments": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "streetAddress": "",
  "geoShapeFilter": { "polygons": [], "circles": [] },
  "fireStations": { "ids": [], "isMissingValue": false },
  "dispatchedFireDepartments": { "ids": [], "isMissingValue": false },
  "buildingTypes": { "ids": [], "isMissingValue": false },
  "naceCodes": { "ids": [], "isMissingValue": false },
  "includeAssistanceMissions": false,
  "includeExercises": false,
  "includeMissionsHandledByHundredAndTen": false,
  "includePoliceCauseWithNoMission": false,
  "includeRVRMissionsFromBris": false,
  "includeApprovedMissions": true,
  "onlyApprovedMissions": true,
  "answers": [],
  "periodStart": "2024-01-01",
  "periodEnd":   "2024-12-31",
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_DAY",
        "name": "Dag",
        "type": "DATE_HISTOGRAM",
        "interval": "DAY",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Response shape — verified live

The full raw response is committed at .live_responses/dataset_12_police_arnested_DAY_2024.json. Key invariants the daily adapter relies on:

result[0].buckets[<answer_id>].subAggregationBuckets keys are Unix-ms timestamps that decode via new Date(Number(ts)).toISOString().slice(0, 10) to ISO dates (UTC midnight, no timezone offset).
result[0].total === Σ inner.count across every category × day bucket (3244 / 3244 in the sample).
BRIS zero-pads every calendar day in the period — the 2024 sample has 366 unique day buckets per category (leap year), most with count: 0. transformBrisDailyResponse skips zero cells the same way transformBraskDailyTable does.
The -1 bucket ("Ikke besvart") resolves through resolveCatCodeFromBrisBucket (native_name = 'Ikke besvart' in the catalog).

Notes — Same body as Dataset 1 with buildingTypes.ids = [], nextLevelAggregationDefinition.interval = "DAY", and explicit periodStart / periodEnd. Switching interval to "MONTH" is the Tier-2 monthly-probe variant of the same body.

Norwegian Fylke Code Reference

Authoritative codes for BRIS counties.ids come from SSB klass 104 (Standard for fylkesinndeling). Post-2024 incidents (from 2024-01-01): BRIS uses the same numeric ids as the HeatWaves canonical county spine — one table only, in regions-and-cresta.md (section “Canonical region set (2024, county level)”). Do not copy that table here; update the canonical doc when the spine changes.

Pre-2024 (incidents 2020-01-01 to 2023-12-31) — BRIS klass 104 historical fylke ids for request bodies only (differs from the 2024 canonical list):

Code	Fylke
03	Oslo
11	Rogaland
15	Møre og Romsdal
18	Nordland
30	Viken
34	Innlandet
38	Vestfold og Telemark
42	Agder
46	Vestland
50	Trøndelag
54	Troms og Finnmark
99	Uoppgitt

ETL caveat — BRIS returns historical incidents under the codes that were valid at the time of the incident. A query that filters counties.ids = [30] (Viken) only matches incidents from 2020-2023; post-2024 incidents in the same geography are split across [31, 32, 33]. HeatWaves should normalize across both code sets when building cross-year region series. Source: https://data.ssb.no/api/klass/v1/classifications/104/codes.

Summary Matrix

#	Datasource	Dimension	Fire type	Building type	Region	Endpoint	Body type
1	Police (Sub-A)	Arnested	DEVELOPED	RESIDENTIAL	—	`/aggregate`	full JSON
2	DSB	Arnested	DEVELOPED	n/a	—	`/restrictedaggregation`	minimal JSON
3	Police (Sub-A)	Objekt	DEVELOPED	RESIDENTIAL	—	`/aggregate`	full JSON
4	Police (Sub-A)	Objekt	DEVELOPED	—	—	`/aggregate`	full JSON, empty `buildingTypes.ids`
5	DSB	Tennkilde	CONFINED (no skorstein)	n/a	—	`/restrictedaggregation`	minimal JSON
6	BRASK	Tennkilde	CONFINED	RESIDENTIAL	Oslo	form POST	URL-encoded form, `value=kr`, age 1-5
7	Brigade	Årsak	CONFINED komfyr-only	n/a	—	`/restrictedaggregation`	minimal JSON, single mission
8	BRASK	Årsak	—	—	Oslo	form POST	URL-encoded form, `value=kr`
9	Police (Sub-A)	Arnested	DEVELOPED	—	Oslo	`/aggregate`	full JSON, `counties.ids=[3]`
10	SSB	(totals)	n/a	RESIDENTIAL via dataset	Oslo (`0301`)	PxWebApi v0	JSON-stat2 query
11	DSB	Tennkilde	DEVELOPED	n/a	—	`/restrictedaggregation`	minimal JSON, no time sub-agg
12	Police (Sub-A)	Arnested	DEVELOPED	—	—	`/aggregate`	full JSON, `interval: DAY`, `periodStart`/`periodEnd`

Live API Smoke Test Results (for validation)

Every native request above was executed by scripts/smoke_api_comparison.mjs. Full raw responses are saved under .live_responses/ (one file per dataset — .json for BRIS/SSB, .html for BRASK). That directory is tracked in git so documentation links to pinned samples stay valid in fresh clones; re-run the smoke script when upstream responses or slice templates change, then commit refreshed files. The summary below lets you cross-check key totals against the source websites without opening the raw files.

Smoke-test environment

Run command: node scripts/smoke_api_comparison.mjs
All 12 calls returned HTTP 200.
BRIS year coverage: 2016–2026 (11 buckets, current data range of brannstatistikk.no).
BRASK year coverage: 1985–2025 (41 columns, full archive range).
SSB year coverage requested: top 10 (2016–2025).
The label -1 in BRIS results is the "missing answer" bucket returned because aggregationDefinition.includeMissing = true. Subtract it from grandTotal to get the answered-only count.

Cross-source sanity checks

These three identities hold across the responses and confirm the slice logic in filter-axes-native-to-filter-id.md is internally consistent:

Dataset 2 (DSB Arnested DEVELOPED) grandTotal = 32,320 = Dataset 4 (Police Objekt DEVELOPED, all buildings) grandTotal = Dataset 11 (DSB Tennkilde DEVELOPED, no sub-agg) grandTotal. All three slice the same population (DEVELOPED Brann i bygning nationwide, 2016–2026).
Dataset 1 (Police Arnested DEVELOPED, RESIDENTIAL) grandTotal = 15,879 = Dataset 3 (Police Objekt DEVELOPED, RESIDENTIAL) grandTotal. Same building-type filter, same mission slice.
Dataset 9 (Police Arnested DEVELOPED, Oslo, all buildings) grandTotal = 3,930 — strictly less than Dataset 1's 15,879. Oslo's share is roughly 25% of national, which matches the public Oslo brannvesen statistics for the period.

BRIS results (datasets 1, 2, 3, 4, 5, 7, 9, 11, 12)

#	Endpoint	Question id	Mission types	Cats	Grand total	Year range	Top 3 (cat → count)
1	`/aggregate`	403 (Arnested - politi)	[10026]	21	15,879	2016–2026	-1 → 10,092 · Kjøkken → 1,055 · Stue → 930
2	`/restrictedaggregation`	227 (Antatt arnested)	[10026]	28	32,320	2016–2026	Kjøkken → 4,708 · Stue → 3,996 · Utvendig → 3,897
3	`/aggregate`	405 (Brannen startet i - politi)	[10026]	23	15,879	2016–2026	-1 → 10,151 · Ukjent → 1,444 · Løst elektrisk utstyr → 842
4	`/aggregate`	405	[10026]	23	32,320	2016–2026	-1 → 21,074 · Ukjent → 2,735 · Løst elektrisk utstyr → 1,515
5	`/restrictedaggregation`	264 (Antatt tennkilde)	[10086, 10212]	7	31,392	2016–2026	-1 → 31,313 · Elektrisitet → 48 · Annen → 14
7	`/restrictedaggregation`	269 (Brannvesenets vurdering)	[10086]	4	17,801	2016–2026	-1 → 17,751 · Feil bruk → 46 · Ukjent → 3
9	`/aggregate`	403	[10026] + counties=[3]	22	3,930	2016–2026	-1 → 2,176 · Kjøkken → 360 · Stue → 229
11	`/restrictedaggregation`	264, no time sub-agg	[10026]	7	32,320	n/a (categories-only)	Ukjent → 10,940 · Elektrisitet → 7,539 · Åpen ild → 6,530
12	`/aggregate`	403, `interval: DAY`, period 2024	[10026]	22	3,244	2024-01-01..2024-12-31 (366 day buckets)	-1 → 1,683 · Ukjent → 211 · Stue → 210

Dataset 5 has very few non--1 answers (79 of 31,392) because Tennkilde (264) is rarely filled in for confined incidents. Dataset 7 likewise has only 50 non--1 answers because Brigade-vurdering (269) is rarely populated for Branntilløp komfyr events.

BRASK results (datasets 6, 8)

BRASK numbers are reported in tusen kr (thousands of NOK) when rblVerdi=2. The "Grand total" column below is the SUM × SUM cell of the returned table (sum across all categories AND all years).

#	Row dim	Filters	Cats	Year range	Grand total (tusen kr)	Sample row totals
6	`kilde` (5)	Kald + Bolig + Oslo + 1–5 år + kr	7	1985–2025	2,042	Ildsted → 18 · Varme arbeider → 0 · Åpen ild → 0 · Elektroniske apparater → 60 · Husholdning → 166
8	`arsak` (6)	Oslo + kr (no fire-type / building-type filter)	7	1985–2025	14,554,292	Antatt påsatt → 706,655 · Menneskelig feil → 1,725,625 · Teknisk svikt → 1,257,457 · Selvantennelse → 351,404 · Lynnedslag → 231,198

Dataset 6's filters are narrow enough (Oslo + 1–5 år + Kald + Bolig) that the entire 41-year dataset only sums to ~2 million NOK in claims — verify on the BRASK UI by setting Aldersgruppe = 1–5, Fylke = Oslo, Type = Kald, Næring = Beboelse, Verdi = Erstatningsbeløp. The mostly-zero year columns are accurate. Dataset 8 returns 7 Årsak rows = the seven canonical BRASK cause categories from category-native-to-cat-code.md §1 (cat_codes 101–107).

SSB results (dataset 10)

SSB Boligbranner for Oslo (region 0301), 2016–2025, raw value array from json-stat2:

Year	Antall boligbranner
2016	265
2017	214
2018	247
2019	228
2020	238
2021	272
2022	204
2023	219
2024	250
2025	289
Total 2016–2025	2,426

SSB numbers are independent of BRIS/BRASK (different methodology — SSB sources from KOSTRA municipal reporting). For 2016–2025 SSB reports 2,426 boligbranner in Oslo. Compare to BRIS Dataset 9 (Police Arnested, Oslo, DEVELOPED, all buildings, 2016–2026) which reports 3,930 incidents — the BRIS number is larger because it includes non-RESIDENTIAL buildings, the -1 missing bucket, and one extra year.

How to validate manually

Source	Manual check
BRIS	Open `https://brannstatistikk.no/search`, build the same query in the UI (mission type → group by → time), and compare the displayed counts.
BRASK	Open `https://brask.finansnorge.no/`, set the same Rad/Kolonne dropdowns and filter checkboxes (Type, Næring, Aldersgruppe, Fylke), and compare the table.
SSB	Open `https://www.ssb.no/statbank/table/12058/`, select region = Oslo, statistikkvariabel = Antall boligbranner, år = last 10, and compare the values.

Re-run with node scripts/smoke_api_comparison.mjs. Raw responses overwrite the files in .live_responses/ so you can diff before/after on subsequent runs.

Loading documentation...

Native API Comparison — Smoke Test Datasets

Every numeric id, parameter name, and value below is sourced from one of:

category-native-to-cat-code.md — cat_code ↔ native_api_code mapping
filter-axes-native-to-filter-id.md — filter parameter names and values per source
lib/data-model/adapters/_brisRequestBody.ts — BRIS request builder (buildBrisRequestBody)
lib/data-model/adapters/_brisUpstream.constants.ts — BOLIG_IDS, NON_RESIDENTIAL_MATRIKKEL_LEAF_IDS, BRIS_QUESTION_AGGREGATIONS, TIME_OPTIONS
lib/data-model/adapters/brask-passthrough.ts — BRASK form mechanics (shared constants in lib/data-model/adapters/_braskWebForms.constants.ts: BRASK_FILTER_FIELD_MAP, BRASK_AXIS / braskExplorerAxisId, BRASK_UTFORSK_RBL_*)
ssb-pxwebapi-tables.md — SSB table 12058 variable reference

Unless otherwise noted, every example is a time series with years on the x-axis, antall (or kr) on the y-axis, and categories within the chosen dimension as series.

Shared API Mechanics

Three upstream endpoints are exercised across the eleven datasets:

Source	Endpoint	Body type
BRASK	`POST https://brask.finansnorge.no/`	`application/x-www-form-urlencoded` (ASP.NET form)
BRIS — Alle branner og ulykker	`POST https://brannstatistikk.no/api/v1/missionreports/aggregate`	`application/json` (full filter body)
BRIS — Hvordan bygningsbranner starter	`POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation`	`application/json` (minimal body)
SSB	`POST https://data.ssb.no/api/v0/no/table/12058`	`application/json` (PxWebApi v0 query, returns json-stat2)

Dataset 1 — Police, Arnested, Utviklet brann, Bolig

HeatWaves intent

datasource: BRIS — Police (Sub-A, "Alle branner og ulykker")
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = RESIDENTIAL

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "hitsToReturn": 0,
  "skipped": 0,
  "missionTypes": { "ids": [], "isMissingValue": false },
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "monthNumbers": { "ids": [], "isMissingValue": false },
  "dayOfWeek": { "ids": [], "isMissingValue": false },
  "responseTimeLimitGreaterThan": true,
  "callOutTimeLimitGreaterThan": true,
  "effortDurationGreaterThan": true,
  "counties": { "ids": [], "isMissingValue": false },
  "municipalities": { "ids": [], "isMissingValue": false },
  "basicStatisticalUnits": { "ids": [], "isMissingValue": false },
  "fireDepartments": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "streetAddress": "",
  "geoShapeFilter": { "polygons": [], "circles": [] },
  "fireStations": { "ids": [], "isMissingValue": false },
  "dispatchedFireDepartments": { "ids": [], "isMissingValue": false },
  "buildingTypes": {
    "ids": [
      "111","112","113","121","122","123","124","131","133","135","136",
      "141","142","143","144","145","146","151","152","159","161","162",
      "163","171","172","181","182","183","193","199"
    ],
    "isMissingValue": false
  },
  "naceCodes": { "ids": [], "isMissingValue": false },
  "includeAssistanceMissions": false,
  "includeExercises": false,
  "includeMissionsHandledByHundredAndTen": false,
  "includePoliceCauseWithNoMission": false,
  "includeRVRMissionsFromBris": false,
  "includeApprovedMissions": true,
  "onlyApprovedMissions": true,
  "answers": [],
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Dataset 2 — DSB, Arnested, Utviklet brann, no building type

HeatWaves intent

datasource: BRIS — DSB ("Hvordan bygningsbranner starter")
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none (not supported on this endpoint)

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "227",
      "name": "Antatt arnested (startsted)",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Dataset 3 — Police, Objekt, Utviklet brann, Bolig

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 3 (Objekt) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = RESIDENTIAL

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

Same envelope as Dataset 1 with two changes:

{
  "...": "[same as Dataset 1, except:]",
  "buildingTypes": { "ids": ["...30 BOLIG_IDS..."], "isMissingValue": false },
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "405",
      "name": "Brannen startet i - politi",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — aggregationDefinition.id = "405" is fullOnly. Result rows are keyed by 405_* answer ids — category-native-to-cat-code.md §3 (cat_codes 301-322).

Dataset 4 — Police, Objekt, Utviklet brann, all buildings

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 3 (Objekt) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = — (unfiltered — all buildings)

Native API

Identical to Dataset 3 except buildingTypes.ids is empty:

{
  "...": "[same as Dataset 3, except:]",
  "buildingTypes": { "ids": [], "isMissingValue": false }
}

Dataset 5 — DSB, Tennkilde, Begrenset branntilløp, all buildings

HeatWaves intent

datasource: BRIS — DSB ("Hvordan bygningsbranner starter")
dimension: dim_code = 2 (Tennkilde) — series
x-axis: year
y-axis: antall
filters: fire_type = CONFINED, building_type = none (not supported on this endpoint)

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10086, 10212], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "264",
      "name": "Antatt tennkilde",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Dataset 6 — BRASK, Tennkilde, Begrenset, Bolig, Oslo, alder 1–5, value=kr

HeatWaves intent

datasource: BRASK
dimension: dim_code = 2 (Tennkilde / Kilde) — series
x-axis: year
y-axis: erstatningsbeløp (kr)
filters: fire_type = CONFINED, building_type = RESIDENTIAL, region = Oslo, building_age = 1-5

This is the most filter-heavy BRASK example — exercises every BRASK filter mechanic in one request.

Native API

POST https://brask.finansnorge.no/
Content-Type: application/x-www-form-urlencoded

__VIEWSTATE=<from initial GET>
__VIEWSTATEGENERATOR=<from initial GET>
__EVENTVALIDATION=<from initial GET>
ctl00$Innhold$ddlRad=5
ctl00$Innhold$ddlKolonne=7
ctl00$Innhold$rblVerdi=2
ctl00$Innhold$rblBeregning=1
ctl00$Innhold$lbBransje=
ctl00$Innhold$lbAldersgruppe=2
ctl00$Innhold$lbFylke=3
ctl00$Innhold$lbDag=
ctl00$Innhold$lbKilde=
ctl00$Innhold$lbKvartal=
ctl00$Innhold$lbMåned=
ctl00$Innhold$lbNæring=0
ctl00$Innhold$lbType=2
ctl00$Innhold$lbUkedag=
ctl00$Innhold$lbÅr=
ctl00$Innhold$lbÅrsak=
ctl00$Innhold$btnKjør=Lag tabell

Notes — Field-by-field justification:

ddlRad=5 — row dimension = kilde (Utforsk ddlRad index = BRASK_AXIS.kilde (5) in _braskWebForms.constants.ts; not fire_data.categories.dim_code).
ddlKolonne=7 — column dimension = ar (BRASK_AXIS.ar = 7).
rblVerdi=2 — value type = erstatningsbelop (kr); 1 would be antallSkader.
lbAldersgruppe=2 — native value 2 = "1-5" per filter-axes-native-to-filter-id.md §3 (normalizes to HeatWaves bucket 0-5).
lbFylke=3 — Oslo per filter-axes-native-to-filter-id.md §4.
lbNæring=0 — Beboelse = RESIDENTIAL per filter-axes-native-to-filter-id.md §2.
lbType=2 — Kald = CONFINED per filter-axes-native-to-filter-id.md §1.
All other lb* fields sent empty (= Alle / no filter).

Dataset 7 — Brigade, Årsak, Komfyr-only confined, all buildings

HeatWaves intent

datasource: BRIS — Fire Brigade ("Hvordan bygningsbranner starter")
dimension: dim_code = 1 (Årsak) — series
x-axis: year
y-axis: antall
filters: fire_type = CONFINED — komfyr only, building_type = none

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10086], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "269",
      "name": "Brannvesenets vurdering",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Dataset 8 — BRASK, Årsak, all fire types, all buildings, Oslo, value=kr

HeatWaves intent

datasource: BRASK
dimension: dim_code = 1 (Årsak) — series
x-axis: year
y-axis: erstatningsbeløp (kr)
filters: fire_type = — (unfiltered), building_type = — (unfiltered), region = Oslo

Native API

POST https://brask.finansnorge.no/
Content-Type: application/x-www-form-urlencoded

__VIEWSTATE=<from initial GET>
__VIEWSTATEGENERATOR=<from initial GET>
__EVENTVALIDATION=<from initial GET>
ctl00$Innhold$ddlRad=6
ctl00$Innhold$ddlKolonne=7
ctl00$Innhold$rblVerdi=2
ctl00$Innhold$rblBeregning=1
ctl00$Innhold$lbBransje=
ctl00$Innhold$lbAldersgruppe=
ctl00$Innhold$lbFylke=3
ctl00$Innhold$lbDag=
ctl00$Innhold$lbKilde=
ctl00$Innhold$lbKvartal=
ctl00$Innhold$lbMåned=
ctl00$Innhold$lbNæring=
ctl00$Innhold$lbType=
ctl00$Innhold$lbUkedag=
ctl00$Innhold$lbÅr=
ctl00$Innhold$lbÅrsak=
ctl00$Innhold$btnKjør=Lag tabell

Notes

ddlRad=6 — row dimension = arsak (BRASK_AXIS.arsak = 6 in _braskWebForms.constants.ts).
ddlKolonne=7 — column dimension = ar (year).
rblVerdi=2 — kr.
lbFylke=3 — Oslo only.
lbType= and lbNæring= empty — all fire types / all building types.
The only example showing kr value-type without fire-type or building-type narrowing — useful as a "broad financial loss by cause" smoke test.

Dataset 9 — BRIS Police, Arnested, Utviklet brann, all buildings, Oslo

HeatWaves intent

datasource: BRIS — Police (Sub-A)
dimension: dim_code = 4 (Arnested) — series
x-axis: year
y-axis: antall
filters: fire_type = DEVELOPED, building_type = — (unfiltered — all buildings), region = Oslo

Direct cross-source counterpart to Datasets 6 and 8 — shows how Oslo is filtered on the BRIS side.

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "...": "[same as Dataset 1 envelope, except:]",
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [3], "isMissingValue": false },
  "buildingTypes": { "ids": [], "isMissingValue": false },
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_YEAR",
        "name": "År",
        "type": "DATE_HISTOGRAM",
        "interval": "YEAR",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Notes — cross-system caveat

BRASK lbFylke = 3 and BRIS counties.ids = [3] both happen to equal Oslo, but the systems align only by coincidence for Oslo. Every other county diverges:
- BRASK uses pre-2020 cresta-sone (e.g. 2 = Akershus, 7 = Vestfold, 19 = Troms).
- BRIS uses SSB klass 104 (e.g. 30 = Viken for 2020-2023 incidents, 32 = Akershus for post-2024 incidents — see fylke-code appendix below).
For finer granularity, municipalities.ids = [301] would also work for Oslo in BRIS.

Dataset 10 — SSB, Boligbranner (RESIDENTIAL dataset), Oslo

HeatWaves intent

datasource: SSB (third datasource — first appearance in this doc)
dimension: total counts (no within-dimension series — SSB does not break down causal/object dimensions)
x-axis: year
y-axis: antall
filters: building_type = RESIDENTIAL (encoded by selecting the Boligbranner statistical variable, not a parameter), region = Oslo

Cross-source counterpart to Datasets 6, 8, 9 — same Oslo geography across all four sources.

Native API

POST https://data.ssb.no/api/v0/no/table/12058
Content-Type: application/json

{
  "query": [
    {
      "code": "KOKkommuneregion0000",
      "selection": { "filter": "item", "values": ["0301"] }
    },
    {
      "code": "ContentsCode",
      "selection": { "filter": "item", "values": ["KOSboligbranner0000"] }
    },
    {
      "code": "Tid",
      "selection": { "filter": "top", "values": ["10"] }
    }
  ],
  "response": { "format": "json-stat2" }
}

Notes

Table 12058 is the SSB KOSTRA fire and safety table. Variable codes are documented in ssb-pxwebapi-tables.md.
For the all-buildings total slice (building_type_code NULL; Bygningsbranner / KOSbygningsbrann0000) instead of RESIDENTIAL, swap the ContentsCode value accordingly. For kr instead of antall: KOSerstatnbygnin0000.
KOKkommuneregion0000: municipality codes (0301 = Oslo) or EAK for national total. PxWeb → region_code: filter-axes-native-to-filter-id.md §4; upstream codes: ssb-pxwebapi-tables.md.
Tid is years natively in SSB — no equivalent of BRIS's nextLevelAggregationDefinition is needed. The "top": "10" value means "the most recent 10 years"; replace with { "filter": "item", "values": ["2020","2021","2022","2023","2024"] } for an explicit list.
Response format json-stat2: row-major value array — decode contract in pipeline-slices-idempotency-and-triggers.md §4.3; PxWebApi mechanics in ssb-pxwebapi-tables.md.

Dataset 11 — BRIS DSB, Tennkilde, no time sub-aggregation

HeatWaves intent

datasource: BRIS — DSB
dimension: dim_code = 2 (Tennkilde) — series
x-axis: none (categories only — no time series)
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none

This is the only example in the doc that does NOT match the "year on x-axis" rule — included expressly to show the simpler payload shape for ETL slices that only need categorical totals.

Native API

POST https://brannstatistikk.no/api/v1/missionreports/restrictedaggregation
Content-Type: application/json

{
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "counties": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "includePoliceCauseWithNoMission": false,
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "264",
      "name": "Antatt tennkilde",
      "type": "TERMS",
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": []
    }
  }
}

Dataset 12 — BRIS Police, Arnested, Utviklet brann, DAY granularity, single year window

HeatWaves intent

datasource: BRIS — Police (Sub-A, "Alle branner og ulykker")
dimension: dim_code = 4 (Arnested) — series
x-axis: date (one bucket per calendar day)
y-axis: antall
filters: fire_type = DEVELOPED, building_type = none, period = 2024

Native API

POST https://brannstatistikk.no/api/v1/missionreports/aggregate
Content-Type: application/json

{
  "hitsToReturn": 0,
  "skipped": 0,
  "missionTypes": { "ids": [], "isMissingValue": false },
  "revisedMissionTypes": { "ids": [10026], "isMissingValue": false },
  "monthNumbers": { "ids": [], "isMissingValue": false },
  "dayOfWeek": { "ids": [], "isMissingValue": false },
  "responseTimeLimitGreaterThan": true,
  "callOutTimeLimitGreaterThan": true,
  "effortDurationGreaterThan": true,
  "counties": { "ids": [], "isMissingValue": false },
  "municipalities": { "ids": [], "isMissingValue": false },
  "basicStatisticalUnits": { "ids": [], "isMissingValue": false },
  "fireDepartments": { "ids": [], "isMissingValue": false },
  "hundredAndTenCentrals": { "ids": [], "isMissingValue": false },
  "streetAddress": "",
  "geoShapeFilter": { "polygons": [], "circles": [] },
  "fireStations": { "ids": [], "isMissingValue": false },
  "dispatchedFireDepartments": { "ids": [], "isMissingValue": false },
  "buildingTypes": { "ids": [], "isMissingValue": false },
  "naceCodes": { "ids": [], "isMissingValue": false },
  "includeAssistanceMissions": false,
  "includeExercises": false,
  "includeMissionsHandledByHundredAndTen": false,
  "includePoliceCauseWithNoMission": false,
  "includeRVRMissionsFromBris": false,
  "includeApprovedMissions": true,
  "onlyApprovedMissions": true,
  "answers": [],
  "periodStart": "2024-01-01",
  "periodEnd":   "2024-12-31",
  "aggregationQuery": {
    "includeNationalFigures": false,
    "aggregationDefinition": {
      "id": "403",
      "name": "Arnested - politi",
      "type": "TERMS",
      "interval": null,
      "question": true,
      "groupAnswer": false,
      "restricted": false,
      "includeMissing": true,
      "calculatedSubAggregationDefinitions": [],
      "nextLevelAggregationDefinition": {
        "id": "MISSION_CALL_TIME_DAY",
        "name": "Dag",
        "type": "DATE_HISTOGRAM",
        "interval": "DAY",
        "question": false,
        "groupAnswer": false,
        "restricted": false,
        "includeMissing": true,
        "calculatedSubAggregationDefinitions": []
      }
    }
  }
}

Response shape — verified live

The full raw response is committed at .live_responses/dataset_12_police_arnested_DAY_2024.json. Key invariants the daily adapter relies on:

result[0].buckets[<answer_id>].subAggregationBuckets keys are Unix-ms timestamps that decode via new Date(Number(ts)).toISOString().slice(0, 10) to ISO dates (UTC midnight, no timezone offset).
result[0].total === Σ inner.count across every category × day bucket (3244 / 3244 in the sample).
BRIS zero-pads every calendar day in the period — the 2024 sample has 366 unique day buckets per category (leap year), most with count: 0. transformBrisDailyResponse skips zero cells the same way transformBraskDailyTable does.
The -1 bucket ("Ikke besvart") resolves through resolveCatCodeFromBrisBucket (native_name = 'Ikke besvart' in the catalog).

Norwegian Fylke Code Reference

Pre-2024 (incidents 2020-01-01 to 2023-12-31) — BRIS klass 104 historical fylke ids for request bodies only (differs from the 2024 canonical list):

Code	Fylke
03	Oslo
11	Rogaland
15	Møre og Romsdal
18	Nordland
30	Viken
34	Innlandet
38	Vestfold og Telemark
42	Agder
46	Vestland
50	Trøndelag
54	Troms og Finnmark
99	Uoppgitt

Summary Matrix

#	Datasource	Dimension	Fire type	Building type	Region	Endpoint	Body type
1	Police (Sub-A)	Arnested	DEVELOPED	RESIDENTIAL	—	`/aggregate`	full JSON
2	DSB	Arnested	DEVELOPED	n/a	—	`/restrictedaggregation`	minimal JSON
3	Police (Sub-A)	Objekt	DEVELOPED	RESIDENTIAL	—	`/aggregate`	full JSON
4	Police (Sub-A)	Objekt	DEVELOPED	—	—	`/aggregate`	full JSON, empty `buildingTypes.ids`
5	DSB	Tennkilde	CONFINED (no skorstein)	n/a	—	`/restrictedaggregation`	minimal JSON
6	BRASK	Tennkilde	CONFINED	RESIDENTIAL	Oslo	form POST	URL-encoded form, `value=kr`, age 1-5
7	Brigade	Årsak	CONFINED komfyr-only	n/a	—	`/restrictedaggregation`	minimal JSON, single mission
8	BRASK	Årsak	—	—	Oslo	form POST	URL-encoded form, `value=kr`
9	Police (Sub-A)	Arnested	DEVELOPED	—	Oslo	`/aggregate`	full JSON, `counties.ids=[3]`
10	SSB	(totals)	n/a	RESIDENTIAL via dataset	Oslo (`0301`)	PxWebApi v0	JSON-stat2 query
11	DSB	Tennkilde	DEVELOPED	n/a	—	`/restrictedaggregation`	minimal JSON, no time sub-agg
12	Police (Sub-A)	Arnested	DEVELOPED	—	—	`/aggregate`	full JSON, `interval: DAY`, `periodStart`/`periodEnd`

Live API Smoke Test Results (for validation)

Smoke-test environment

Run command: node scripts/smoke_api_comparison.mjs
All 12 calls returned HTTP 200.
BRIS year coverage: 2016–2026 (11 buckets, current data range of brannstatistikk.no).
BRASK year coverage: 1985–2025 (41 columns, full archive range).
SSB year coverage requested: top 10 (2016–2025).
The label -1 in BRIS results is the "missing answer" bucket returned because aggregationDefinition.includeMissing = true. Subtract it from grandTotal to get the answered-only count.

Cross-source sanity checks

These three identities hold across the responses and confirm the slice logic in filter-axes-native-to-filter-id.md is internally consistent:

Dataset 2 (DSB Arnested DEVELOPED) grandTotal = 32,320 = Dataset 4 (Police Objekt DEVELOPED, all buildings) grandTotal = Dataset 11 (DSB Tennkilde DEVELOPED, no sub-agg) grandTotal. All three slice the same population (DEVELOPED Brann i bygning nationwide, 2016–2026).
Dataset 1 (Police Arnested DEVELOPED, RESIDENTIAL) grandTotal = 15,879 = Dataset 3 (Police Objekt DEVELOPED, RESIDENTIAL) grandTotal. Same building-type filter, same mission slice.
Dataset 9 (Police Arnested DEVELOPED, Oslo, all buildings) grandTotal = 3,930 — strictly less than Dataset 1's 15,879. Oslo's share is roughly 25% of national, which matches the public Oslo brannvesen statistics for the period.

BRIS results (datasets 1, 2, 3, 4, 5, 7, 9, 11, 12)

#	Endpoint	Question id	Mission types	Cats	Grand total	Year range	Top 3 (cat → count)
1	`/aggregate`	403 (Arnested - politi)	[10026]	21	15,879	2016–2026	-1 → 10,092 · Kjøkken → 1,055 · Stue → 930
2	`/restrictedaggregation`	227 (Antatt arnested)	[10026]	28	32,320	2016–2026	Kjøkken → 4,708 · Stue → 3,996 · Utvendig → 3,897
3	`/aggregate`	405 (Brannen startet i - politi)	[10026]	23	15,879	2016–2026	-1 → 10,151 · Ukjent → 1,444 · Løst elektrisk utstyr → 842
4	`/aggregate`	405	[10026]	23	32,320	2016–2026	-1 → 21,074 · Ukjent → 2,735 · Løst elektrisk utstyr → 1,515
5	`/restrictedaggregation`	264 (Antatt tennkilde)	[10086, 10212]	7	31,392	2016–2026	-1 → 31,313 · Elektrisitet → 48 · Annen → 14
7	`/restrictedaggregation`	269 (Brannvesenets vurdering)	[10086]	4	17,801	2016–2026	-1 → 17,751 · Feil bruk → 46 · Ukjent → 3
9	`/aggregate`	403	[10026] + counties=[3]	22	3,930	2016–2026	-1 → 2,176 · Kjøkken → 360 · Stue → 229
11	`/restrictedaggregation`	264, no time sub-agg	[10026]	7	32,320	n/a (categories-only)	Ukjent → 10,940 · Elektrisitet → 7,539 · Åpen ild → 6,530
12	`/aggregate`	403, `interval: DAY`, period 2024	[10026]	22	3,244	2024-01-01..2024-12-31 (366 day buckets)	-1 → 1,683 · Ukjent → 211 · Stue → 210

Dataset 5 has very few non--1 answers (79 of 31,392) because Tennkilde (264) is rarely filled in for confined incidents. Dataset 7 likewise has only 50 non--1 answers because Brigade-vurdering (269) is rarely populated for Branntilløp komfyr events.

BRASK results (datasets 6, 8)

BRASK numbers are reported in tusen kr (thousands of NOK) when rblVerdi=2. The "Grand total" column below is the SUM × SUM cell of the returned table (sum across all categories AND all years).

#	Row dim	Filters	Cats	Year range	Grand total (tusen kr)	Sample row totals
6	`kilde` (5)	Kald + Bolig + Oslo + 1–5 år + kr	7	1985–2025	2,042	Ildsted → 18 · Varme arbeider → 0 · Åpen ild → 0 · Elektroniske apparater → 60 · Husholdning → 166
8	`arsak` (6)	Oslo + kr (no fire-type / building-type filter)	7	1985–2025	14,554,292	Antatt påsatt → 706,655 · Menneskelig feil → 1,725,625 · Teknisk svikt → 1,257,457 · Selvantennelse → 351,404 · Lynnedslag → 231,198

Dataset 6's filters are narrow enough (Oslo + 1–5 år + Kald + Bolig) that the entire 41-year dataset only sums to ~2 million NOK in claims — verify on the BRASK UI by setting Aldersgruppe = 1–5, Fylke = Oslo, Type = Kald, Næring = Beboelse, Verdi = Erstatningsbeløp. The mostly-zero year columns are accurate. Dataset 8 returns 7 Årsak rows = the seven canonical BRASK cause categories from category-native-to-cat-code.md §1 (cat_codes 101–107).

SSB results (dataset 10)

SSB Boligbranner for Oslo (region 0301), 2016–2025, raw value array from json-stat2:

Year	Antall boligbranner
2016	265
2017	214
2018	247
2019	228
2020	238
2021	272
2022	204
2023	219
2024	250
2025	289
Total 2016–2025	2,426

SSB numbers are independent of BRIS/BRASK (different methodology — SSB sources from KOSTRA municipal reporting). For 2016–2025 SSB reports 2,426 boligbranner in Oslo. Compare to BRIS Dataset 9 (Police Arnested, Oslo, DEVELOPED, all buildings, 2016–2026) which reports 3,930 incidents — the BRIS number is larger because it includes non-RESIDENTIAL buildings, the -1 missing bucket, and one extra year.

How to validate manually

Source	Manual check
BRIS	Open `https://brannstatistikk.no/search`, build the same query in the UI (mission type → group by → time), and compare the displayed counts.
BRASK	Open `https://brask.finansnorge.no/`, set the same Rad/Kolonne dropdowns and filter checkboxes (Type, Næring, Aldersgruppe, Fylke), and compare the table.
SSB	Open `https://www.ssb.no/statbank/table/12058/`, select region = Oslo, statistikkvariabel = Antall boligbranner, år = last 10, and compare the values.

Re-run with node scripts/smoke_api_comparison.mjs. Raw responses overwrite the files in .live_responses/ so you can diff before/after on subsequent runs.