earthobservations / wetterdienst

Open weather data for humans.
https://wetterdienst.readthedocs.io/
MIT License
358 stars 55 forks source link

Obtaining metadata about specific data sets #112

Closed amotl closed 4 years ago

amotl commented 4 years ago

Hi there,

together with @wetterfrosch, we are also looking forward to bring metadata about different datasets into this program. The best way to do that would be by directly ingesting resources available on the DWD CDC server or elsewhere from the content published by DWD. Hereby, I wanted to start collecting some resources.

Weather observations

1. Field names and descriptions (German) as XLSX.

2. Field names and descriptions per data set (English and German) as PDF.

Field names per data set are available within the respective folder within the "Parameters" section of corresponding PDF files, for example see data set description for recent daily station observations. Within dwd_description_pdf.py, I've tried to get some Python together to parse these PDF files and it seems to work reasonably, e.g. for 10_minutes/air_temperature/recent:

+-------------+--------------------------------------------------------------------------+
| STATIONS_ID | station identification number                                            |
| MESS_DATUM  | measurement time yyyymmddhhmi                                            |
| QN          | quality level of next columns coding see paragraph "Quality information" |
| PP_10       | pressure at station height hPa                                           |
| TT_10       | air temperature at 2m height                                             |
| TM5_10      | air temperature at 5cm height                                            |
| RF_10       | relative humidity at 2m height                                           |
| TD_10       | dew point temperature at 2m height                                       |
+-------------+--------------------------------------------------------------------------+

MOSMIX forecasts

With kind regards, Andreas.

amotl commented 4 years ago

We've implemented variant 2 for observation data with #199. All things related to forecast data (MOSMIX) will be handled within #70.