Explore the open-data infrastructure behind IMBER, then access UK forecasts and near-real-time rainfall radar through documented, read-only endpoints.
Currently available as a developing public service. No API key is required for the endpoints below. Availability, response formats and access limits may change while we improve the platform.
Independent software connects openly published meteorological data to a UK-focused forecast and radar service. This overview explains the infrastructure and what is checked live.
Status checks test publicly accessible data endpoints when this page is opened; they are not a guarantee of uptime, a continuous infrastructure monitor or confirmation that every forecast model is available. No API key is required for the documented read-only endpoints.
Interested in the code behind the website? Explore the IMBER GitHub repository ↗. The published source currently covers the WordPress plugin, not the complete API backend.
Use GET requests over HTTPS. The WordPress endpoints below power our website and are suitable for browser-based requests to this domain. The direct forecast service is also available at api.imberweather.com.
Search the published settlement catalogue. Results include names and geographical coordinates. An empty search lists available forecast locations.
https://imberweather.com/wp-json/imber/v1/locations?search=CardiffReturns future forecast times and available weather variables for a cached location, together with its actual coordinates, source, model and run timestamp.
https://imberweather.com/wp-json/imber/v1/forecast?place=CardiffProvide decimal latitude and longitude. IMBER selects the nearest cached town where one is within 35 km. The response identifies the forecast location and distance: it is not an exact-location interpolation.
https://imberweather.com/wp-json/imber/v1/forecast?lat=51.45&lon=-3.31Unlike the UKV forecast, radar images show observed, radar-derived rainfall rate across the UK. IMBER downloads the Met Office’s 1 km composite from AWS Open Data every 15 minutes when new observations are available, converts the ODIM HDF5 data into geographically referenced, transparent PNG overlays, and retains a short rolling sequence for animation.
The radar metadata response lists each available frame’s UTC observation time and image URL, along with geographic bounds and source attribution. A browser or mapping library can draw an image above its own basemap. The overlays are observational estimates, not forecasts or rainfall probability.
Returns the available radar frames (normally up to nine, covering roughly two hours), geographic bounds, UTC timestamps, attribution and image URLs. The collection updates as new Met Office observations are processed.
https://imberweather.com/wp-json/imber/v1/radarRead the frames[].url property from the radar metadata response to fetch a specific transparent PNG. Do not hard-code a timestamp: old frames are automatically removed.
https://api.imberweather.com/v1/radar/frames/{YYYYMMDDHHMM}.pngThe input radar grid is approximately 1 km. New observations are usually published at 15-minute intervals, with a publication and processing delay. The colourised PNG is resampled for web mapping and is not itself a 1 km analytical grid.
Rainfall colours represent intensity in mm/h. Transparent areas can mean no detected rainfall, values below the display threshold or unavailable coverage. The observation timestamp, rather than the page load time, indicates data age.
The PNG is presented with latitude/longitude bounds in the metadata. Use the provided bounds with a compatible image-overlay map layer. Basemap imagery has separate licensing and attribution requirements.
Radar can miss or misestimate rain due to beam geometry, terrain, range and processing artefacts. Do not use the map alone for flood warnings, aviation, emergency operations or critical decisions.
Radar source: © Crown copyright, Met Office, UK composite radar-derived rain rate, via AWS Open Data (CC BY-SA 4.0). Reprojected and colourised by IMBER. The source dataset’s licensing obligations continue to apply when reusing imagery or data.
IMBER processes Met Office UKV numerical weather prediction data into a compact catalogue of forecast values for selected UK towns. Source NetCDF grids are sampled at the nearest model cell for each catalogue location, converted into readable units and cached as a JSON dataset. Our server refreshes the dataset when new model runs become available.
Our current source is the deterministic UKV model, at approximately 2 km grid spacing. Forecast values are model predictions, not live observations; geographic detail and forecast accuracy are not the same thing.
Forecasts are provided at three-hourly intervals, for up to five days. Timestamps use UTC in ISO 8601 format. Data availability can vary by variable and run, and publication may lag the model run.
Available fields include temperature and dew point (°C), humidity and cloud cover (%), wind and gust (mph), direction (degrees), pressure (hPa), precipitation rate (mm/h), visibility (km) and solar radiation (W/m²). Missing fields should not be treated as zero.
Our cache currently covers approximately 1,500 named locations. A coordinate request can be matched to a nearby cached town, not necessarily the requested point. Mountainous, coastal and rural areas may differ meaningfully from the town forecast.
{
"forecast_location": "Cardiff",
"model": "ukv",
"run_time": "ISO-8601 UTC timestamp",
"forecasts": [
{
"time": "ISO-8601 UTC timestamp",
"variables": {
"temperature": { "value": 0, "unit": "°C" }
}
}
]
}Values in this example are placeholders to demonstrate the response format. Live requests return real model values. The schema may be extended as we add models and parameters.
IMBER combines an independently developed presentation and processing pipeline with third-party data. Those underlying data rights do not automatically become unrestricted because they are available through this API.
Our current forecast source is the Met Office UK deterministic 2 km dataset distributed through AWS Open Data. The AWS dataset registry identifies this release as British Crown copyright, Met Office, licensed under Creative Commons Attribution–ShareAlike (CC BY-SA). Reuse may require attribution, a licence link, indication of changes and ShareAlike treatment of adapted material.
Some town names and coordinates are drawn from GeoNames, which publishes data under a Creative Commons Attribution licence and permits commercial reuse with credit. The catalogue also retains locations added independently by IMBER.
Check the current source-specific licence yourself, provide appropriate attribution and assess whether ShareAlike obligations apply to your adapted output. Met Office datasets supplied through other channels may have materially different terms; their licences should not be assumed to apply to this AWS release. IMBER does not grant additional rights over third-party source data.
Suggested source credit: “Forecast data: © Crown copyright, Met Office, via AWS Open Data (CC BY-SA). Place information: GeoNames (CC BY). Processed and presented by IMBER.” Include links to the source and applicable licence, and identify transformations where relevant.
The Met Office describes this AWS release as a free, unsupported dataset and advises against its use for critical business purposes. IMBER is an independent service, not an official Met Office forecast or warning service. Neither availability nor forecast accuracy is guaranteed.
Do not rely on these forecasts alone for safety-critical decisions, emergency response, regulated aviation, or other activities requiring an operationally supported service. Rainfall rate is not a probability of precipitation; a deterministic model does not provide a calibrated confidence percentage. The API is currently read-only and provided on a best-efforts development basis.
For the source documentation and model publication details, see the Met Office external data channels and the AWS dataset registry.