KMA Weather API
Weather forecasts from the Korea Meteorological Administration KMA
Data Sources
This API uses global KMA GDPS 10 km weather forecasts and combines them with high-resolution LDPS 1.5 km model for the North and South Korea. Information about KMA weather models is available here. For GDPS Global, values are interpolated from 3-hourly to 1-hourly time-steps.
| Weather Model | Region | Spatial Resolution | Temporal Resolution | Forecast Length | Update frequency |
|---|---|---|---|---|---|
| KMA GDPS | 🌍 | 0.13° (~12 km) | 3-Hourly | 12 days | Every 6 hours |
| KMA LDPS |  | 1.5 km | Hourly | 2 days | Every 6 hours |
Native Model Variables
KMA directly predicts the fields listed below. Open-Meteo retains these fields or uses them to calculate more convenient API variables. KMA provides both global and direct solar radiation and a native 2 m cloud cover (fog) field.
| Native KMA field | Level | Use in the Open-Meteo API |
|---|---|---|
| Temperature | 2 m, surface | temperature_2m, surface_temperature |
| Relative humidity | 2 m | Relative humidity and dew point |
| U and V wind components | 10 m, 50 m (GDPS) | Wind speed and direction |
| Wind gusts | 10 m | wind_gusts_10m |
| Mean sea-level pressure | Mean sea level | pressure_msl and derived surface pressure |
| Precipitation, convective precipitation (GDPS) and snowfall | Surface | precipitation, rain, showers, snowfall |
| Global and direct solar radiation | Surface | Global, direct, diffuse radiation, DNI and GTI |
| Cloud cover low, mid, high and fog fraction | Surface | Cloud cover layers and cloud_cover_2m |
| Visibility and CAPE (GDPS) | Surface | visibility, cape and weather-code inputs |
CAPE, 50 m wind and convective precipitation are only available in the global GDPS model. The total cloud cover is combined from the layers because the native total reads too high.
Derived Variables
KMA publishes relative humidity, cloud cover layers, a 2 m cloud cover (fog) field and both global and direct solar radiation natively. Open-Meteo derives the remaining API variables from these native fields.
| Derived Variable | How it is derived? |
|---|---|
| Total cloud cover | Combined from the native low, mid and high cloud layers. KMA's own total cloud field is not used because it reads too high. |
| Weather code | Computed from cloud cover, rain, snowfall, wind gusts, visibility and, for the global GDPS model, CAPE and showers. |
| Wind speed and direction | Calculated from the native U and V wind components at 10 m and 50 m. |
| Snowfall and rain | Snowfall converts the native snowfall water equivalent with 0.7 cm per mm. Rain is total precipitation minus snowfall water equivalent (and showers for GDPS). |
| Diffuse radiation and DNI | Diffuse radiation is separated from global radiation with the Razo, Müller Witwer model. DNI follows from native direct radiation, together with GTI and instant values. |
| Dew point, vapour pressure deficit and wet bulb temperature | Calculated from native 2 m temperature and relative humidity. |
| Surface pressure | Calculated from mean sea-level pressure, 2 m temperature and terrain elevation. |
| Apparent temperature, ET₀ and sunshine duration | Combine temperature, humidity, wind speed and solar radiation. ET₀ follows the FAO-56 Penman-Monteith equation; sunshine duration counts time with native DNI above 120 W/m². |
Sunrise, sunset, daylight duration and the day-or-night flag are astronomical calculations. Daily values are aggregated from hourly data.
API Documentation
For a detailed list of all available weather variables please refer to the general Weather Forecast API. Only notable remarks are listed below
- Direct Solar Radiation: KMA provides direct solar radiation. Many other weather models only provide global solar radiation and direct solar radiation must be calculated user separation models.
- Cloud Cover (2m): KMA provides cloud cover at 2 metre above ground which can be interpreted as fog. This is remarkable, because only very weather models are capable of modeling low level cloud cover and fog with a good degree of accuracy.
- Showers:Showers are only computed in the global GDPS model. Since LDPS is a convection-resolving model, it incorporates showers within the regular rain parameter, resulting in a constant value of 0 for the showers variable in LDPS.