# Data Sources

WeatherLayers GL can be used either with custom self-hosted data or with WeatherLayers Cloud.

For integrating any custom data (NetCDF, GRIB), the data needs to be transformed by your backend server to a supported data type, data format and map projection.

### Data sources

* [WeatherLayers Cloud](/weatherlayers-cloud.md) public data sources - NOAA (GFS, GFS Wave), Copernicus (CMEMS, CAMS)
* other public data sources - ECMWF, ICON, Copernicus (ERA5), ...
* commercial data sources
* custom data sources - your own data from scientific research

### Supported data types

* Uint8
  * quantized data into 256 possible values
  * lower precision, higher compression ratio for lower file size
  * recommended for visualization purposes
  * original data bounds need to be provided to unscale the data into the original data, see [Data Properties imageUnscale](/weatherlayers-gl/layers/data-properties.md#imageunscale)
* Float32
  * original data
  * better precision, lower compression ratio and larger file size
  * recommended for scientific purposes, or for use cases where exact values with no quantization errors are needed

### Supported data formats

* PNG, WebP (Uint8)
  * scalar - R channel
    * nodata - `0` in A channel
  * vector - RG channels
    * nodata - `0` in A channel
* GeoTIFF (Uint8)
  * scalar - band 1
    * nodata - `0` in band 2
  * vector - bands 1 and 2
    * nodata - `0` in band 4
* GeoTIFF (Float32)
  * scalar - band 1
    * nodata - `NaN` in band 1
  * vector - bands 1 and 2
    * nodata - `NaN` in bands 1 and 2

### Supported projections

* equirectangular projection (EPSG:4326)

### Data transformation

Data transformation into a supported format can be done on your servers with GDAL.

See [gdal\_translate](https://gdal.org/programs/gdal_translate.html) for transformations between data types and data formats.

See [gdalwarp](https://gdal.org/programs/gdalwarp.html) for transformations between map projections.

See [gdal\_calc](https://gdal.org/en/stable/programs/gdal_calc.html) for calculations and [gdalbuildvrt](https://gdal.org/en/stable/programs/gdalbuildvrt.html) for merging files.

### Example – Temperature from GRIB to PNG

Scale from \[213.15, 325.15] to \[0, 255], disable GDAL unit normalization from K to C:

{% code overflow="wrap" %}

```sh
gdal_translate -ot Byte -scale 213.15 325.15 0 255 --config GRIB_NORMALIZE_UNITS=NO temperature.grib temperature.png
```

{% endcode %}

WeatherLayers GL configuration:

* `imageType: WeatherLayers.ImageType.SCALAR`
* `imageUnscale: [213.15, 325.15]`

### Example – Wind from GRIB to PNG

Merge U and V files:

* R channel: U variable
* G channel: V variable
* B channel: anything, it's ignored, duplicating V variable is the easiest with GDAL, otherwise it would need a new separately created dataset with zeroes

{% code overflow="wrap" %}

```sh
gdalbuildvrt -separate wind.vrt wind_u.grib wind_v.grib wind_v.grib
```

{% endcode %}

Scale from \[-128, 127] to \[0, 255]:

{% code overflow="wrap" %}

```sh
gdal_translate -ot Byte -scale -128 127 0 255 wind.vrt wind.png
```

{% endcode %}

WeatherLayers GL configuration:

* `imageType: WeatherLayers.ImageType.VECTOR`
* `imageUnscale: [-128, 127]`


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.weatherlayers.com/weatherlayers-gl/data-sources.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.