dwd_data_access is a micro service that offers a REST-API to query weather forecasts based on data that has been downloaded from opendata.dwd.de by micro service dwd_data_crawler (https://github.com/UdSAES/dwd_data_crawler).

See the LICENSE file for license rights and limitations (MIT).

Part of the data on opendata.dwd.de is provided in compressed form. For performance reasons, dwd_data_crawler changes the compression from bzip2 to lz4 during the download and storage procedure.

In order to provide a forecast, dwd_data_access needs to decompress the lz4 files. As of today, we, the authors of dwd_data_access, are not aware of a high performance implementation of lz4 in plain JavaScript and/or Node.js. Therefore, we have decided to implement the decompression of lz4 files by calling lz4 command via child_process.execFile. Thus, the lz4 command forms a dependency of dwd_data_access.

The micro service is configured via the following environment variables:

• LISTEN_PORT: The TCP port the micro service is listening for incoming requests.
• DATA_ROOT_PATH: The path to the storage directory where dwd_data_crawler is storing the files downloaded from opendata.dwd.de.
• NEWEST_FORECAST_ROOT_PATH: The path to the directory where the micro service shall store the newest forecasts for the predefined pois (points of interest).
• POIS_JSON_FILE_PATH: The path to the JSON file defining pois (points of interest) for which newest forecasts shall be created. An example of such a file can be found in ./configuration/pois.json.
• JWT_PUBLIC_KEY_FILE_PATH: The path to the public key file to be used to verify JSON web tokens in the authentication header. An example of such a file can be found in ./sample_data/sample_public_key.pem.
• LOG_LEVEL: The log level to be used for this service instance (defaults to info).

Sample call:

$ LISTEN_PORT=12345 DATA_ROOT_PATH=/home/service/DWD_data_storage \
  NEWEST_FORECAST_ROOT_PATH=/home/service/forecast_storage \
  POIS_JSON_FILE_PATH=./configuration/pois.json \
  JWT_PUBLIC_KEY_FILE_PATH=./sample_data/sample_public_key.pem \
  node index.js

To use dwd_data_access with docker, the repository includes a Dockerfile to build a docker image based on alpine. To use the docker image the following configuration needs to be done on startup:

• mount source directory of dwd_data_crawler (see env DATA_ROOT_PATH) to /mnt/dwd_raw_data
• mount target directory for newest forecasts of points of interests (see env NEWEST_FORECAST_ROOT_PATH) to /mnt/forecast_cache
• mount directory of configuration files (i.e. directory where pois.json and vois.json are stored) to /mnt/configuration
• mount file with public keys for verifying JSON web tokens (see envJWT_PUBLIC_KEY_FILE_PATH) to /mnt/keys/public_key.pem

Sample call to build the image:

$ docker build -t dwd_data_access .

Sample call to run the image (LISTEN_PORT is statically set to 3000 within the image):

$ docker run -p 12345:3000 \
  -v $DATA_ROOT_PATH:/mnt/dwd_raw_data \
  -v $NEWEST_FORECAST_ROOT_PATH:/mnt/forecast_cache \
  -v $PWD/configuration:/mnt/configuration \
  -v $PWD/sample_data/sample_public_key.pem:/mnt/keys/public_key.pem \
  --name dwd_data_access_instance \
  dwd_data_access

The basic idea of the micro service dwd_data_accessis to provide weather forecast information based on data previously downloaded by the micro service dwd_data_crawler from opendata.dwd.de. Thus, dwd_data_access itself does not query from opendata.dwd.de directly. This approach has two significant advanteges:

1. Forecast information can be provided even if a connection to opendata.dwd.de is not available (e.g. due to a connection malfunction or if a connection is not possible due to security reasons).
2. As dwd_data_crawler is storing data from opendata.dwd.de in a file structure which prevents overriding of data (as it is done by opendata.dwd.de) it is possible to provide historical forecasts which might be interesting for research activities.

The interaction between opendata.dwd.de, dwd_data_crawler and dwd_data_access is shown in the following diagram.

Forecasts are provided for predefined points of interest. The predefined points of interest are stored in a configuration file in the path ./configuration/pois.json.

For each of the predefined points of interest a newest 27 h and 45 h forecast are cyclically derived by the loop as shown in the following diagram.

The forecasts generated by the loop are stored in the files 27h_forecast.json and 45h_forecast.json under the path $NEWEST_FORECAST_ROOT_PATH/$poi_id/ where

• $NEWEST_FORECAST_ROOT_PATH given as parameter to the program
• $poi_id is the id of the point of interest.