Bergen

Apparent time between sunrise and sunset on the annual shortest day, calculated from the reviewed city latitude using NOAA's documented solar geometry.

Calculation. NOAA apparent-sunrise hour-angle geometry at the local winter solstice, using solar zenith 90.833 degrees and declination 23.44 degrees at the reviewed city latitude; duration = 24 x hour angle / pi

Limits. An astronomical estimate, not observed sunshine or cloud cover. Terrain and changing atmospheric refraction can shift the sunrise and sunset a person actually sees. Astronomical apparent daylight, not sunshine. Local terrain and actual atmospheric refraction can change the visible sunrise and sunset.

Sources [3]

Annual mean concentration of fine particulate matter under 2.5 micrometres, estimated as the arithmetic mean of validated EEA 1 km grid cells within 10 km of the reviewed city point.

Calculation. unweighted arithmetic mean of the 312 valid 1 km cells whose centre lies within 10 km of the reviewed city point (60.39, 5.32), transformed to EPSG:3035 as (4062422.4, 4151816.8); no-data cells excluded, no station or wider-radius fallback

Limits. This is a model-assisted circular-area estimate, not a monitoring-station reading, municipality average or regulatory-compliance assessment. The 10 km radius, grid resolution, valid-cell count and confidence B must remain visible. This is a 10 km-radius PM2.5 estimate around the reviewed city point, combining monitoring data, chemical-transport modelling and supplementary inputs at 1 km. It supports European comparison but does not resolve street-level exposure, indoor air or short pollution episodes.

Sources [4]

Peak ground acceleration with a ten per cent probability of exceedance in fifty years, for reference rock, read at the nearest node of the pinned European Seismic Hazard Model's own computational grid to the reviewed city point.

Calculation. value at the nearest node of the 0.1 degree ESHM20 computational grid to the reviewed city point (60.39, 5.32); the node lies at (60.4000, 5.2821), 2.4 km away. No interpolation between nodes and no averaging over neighbouring nodes

Limits. Does not change between model releases, so it is collected once and re-checked annually. Three things are part of the definition rather than context, and must stay visible with the value: the 475-year return period, the reference rock condition (vs30 800 m/s), and the logic-tree mean, which is one of six aggregations ESHM20 publishes for the same site. The value is read at the nearest node of a nominally 0.1-degree grid, up to about eight kilometres from the city point, so it describes the surrounding area and not an address; the node distance is recorded on every observation. Local ground conditions can raise or lower shaking at a specific site. This is a hazard estimate, not a forecast, a building-code value, or a statement about any particular building. Peak ground acceleration with a 10 per cent chance of being exceeded in 50 years, the 475-year return period, for reference rock (rock_vs30_800ms-1) and the logic-tree mean. Read at the nearest node of a 0.1 degree grid, 2.4 km from the reviewed city point, so it describes the surrounding area rather than a specific address. Local ground conditions can raise or lower shaking at a given site, and this is a hazard estimate, not a forecast, a building-code value or a statement about any building.

Sources [5]

Employed people aged 25 to 54 as a share of people of the same age living in private households, for the region containing the city.

Calculation. sum employed people aged 25-34, 35-44 and 45-54 / sum private-household population in the same age bands x 100

Limits. Derived from Eurostat's compatible 25-34, 35-44 and 45-54 employed-person and private-household population series for one region and year. It is not an average of the three published age-band rates. Derived from six compatible EU Labour Force Survey series for the same NUTS 2 region and year; age-band percentages were not averaged.

Sources [6][7][8][9][10][11]

Unemployed people aged 15 to 74 as a share of the economically active population of the same age, for the region containing the city.

Sources [12]

Unemployed people aged 15 to 24 as a share of the economically active population of the same age, for the region containing the city.

Sources [13]

Gross domestic product per inhabitant in purchasing power standards, for the region containing the city.

Limits. the most recent period this value could be taken from is 2021, which is older than the 36-month freshness this dimension declares Regional GDP reflects where output is produced, which can skew commuter regions.

Sources [14]

Life expectancy at birth, for the region containing the city.

Sources [15]

Years expected to be lived without activity limitation, at birth. Published at country level only.

Sources [16]

Practising physicians per 100,000 inhabitants, for the region containing the city.

Limits. Counts registered physicians, not appointment availability.

Sources [17]

Practising nurses with an EU-recognised qualification per 100,000 inhabitants, published at country level.

Limits. National health-system capacity context, not a city staffing level, appointment guarantee or measure of care quality. Eurostat flags national definition differences and estimates next to affected values.

Sources [18]

Available staffed hospital beds for total care per 100,000 inhabitants, published at country level.

Limits. National capacity context, not the number of beds in the city or proof that a bed is available to a newcomer. Day-care and outpatient places are excluded.

Sources [19]

Police-recorded intentional homicides per 100,000 inhabitants, for the region containing the city.

Sources [21]

Police-recorded intentional homicides per 100,000 person-years for the NUTS 3 region containing the city. The estimate pools the compatible years available within the reviewed 2020-2024 window and identifies incomplete evidence.

Calculation. sum intentional homicides / sum 1 January population for the stated available years x 100,000; paired one-sided 95% exact Garwood Poisson bounds

Limits. A regional recorded-homicide measure, not a measure of all crime, neighbourhood safety or personal risk. See docs/methodology/recorded-homicide.md. Available years=2020,2021,2022,2023,2024 (5 of 5); H=16; E=3214148; R=0.49779911; L=0.31224314; U=0.75606922; threshold=1.0.

Sources [23][24][25][26][27][28][29][30][31][22]

People killed in road traffic accidents per 100,000 inhabitants, for the region containing the city.

Calculation. road fatalities / resident population x 100,000

Limits. Rate derived from reported fatalities and population. Derived from Eurostat fatalities and 1 January population for the same region and year.

Sources [20][32]

Net migration including statistical adjustment, per 1,000 inhabitants, for the region containing the city.

Sources [34]

Change in resident population over the ten years ending in the observed period, for the region containing the city.

Calculation. (population 2025 - population 2015) / population 2015 x 100

Limits. Ten-year change from 2015 to 2025 on the same NUTS 3 boundary.

Sources [35][33]

People aged 25 to 64 living in private households whose highest completed education is ISCED 2011 level 5 to 8, as a share of the same population in the region containing the city.

Limits. Regional adult attainment context from the EU Labour Force Survey. It does not measure school quality, current university capacity or the qualifications of a specific city.

Sources [36]