openaq-fetch
openaq-fetch copied to clipboard
Update Netherlands adapter to use API
Netherlands data is available through an API https://api.luchtmeetnet.nl/open_api. With it we can:
- get more accurate data (the current method is sometimes returning -999 when it shouldn't be)
- programmatically access station data (including coordinates, local authority data provider, and even some metadata)
- access stations reporting BC data others we may be missing
thx. Already looked into it last year. Then the data interface was crappy then. Will look into it again and hope it is better.
Small update: confirmed with someone from luchtmeetnet that
- the measurements are averaged hourly
- the "timestamp_measured" parameter returned by the API is the start of the measurement period and we want the end, so an hour should be added (their website does this btw)
This looks promising, I want to compare this to the data that we already have and check for duplicates.
Going to table this for now. I made a branch called 'lambda-netherlands'. I created a duplicate detector that returned this:
parameterMatch: 'no2 µg/m³',
distanceToNearest: '0.4 meters',
original: {
location: 'Groningen-Europaweg',
coordinates: [ 53.2178, 6.5789 ],
value: 29.5,
date: '2023-02-13T19:00:00.000Z'
},
nearest: {
location: 'Groningen-Europaweg',
coordinates: [ 53.217796, 6.578901 ],
value: 30.77,
date: '2023-02-13T19:00:00.000Z'
}
},
{
parameterMatch: 'pm25 µg/m³',
distanceToNearest: '0.5 meters',
original: {
location: 'Enschede-Winkelhorst.',
coordinates: [ 52.2345, 6.91949 ],
value: 17.0986,
date: '2023-02-13T19:00:00.000Z'
},
nearest: {
location: 'Enschede-Winkelshorst',
coordinates: [ 52.234504, 6.919494 ],
value: 16,
date: '2023-02-13T19:00:00.000Z'
}
}
]
there are 83 stations with possible duplicates
there are 5 stations with no duplicates
While the values are slightly different (maybe due to a time offset problem) the stations are mostly reporting the same data from the same locations so it does not really enhance our coverage of the Netherlands.