No PM10 stations available for GB since 2021 for CAMS2_82

[andagit]

I have raised this issue before (see issue: #940), but perhaps not clear enough. It is peculiar that CAMS2_83 can show stats for GB stations for 2023, but CAMS2_82 can not - I realise that the projects are reading different data, but the origin is EEA for both. I also claim that there are data for GB in our database EEA_AQeRep.NRT/: 2022: 17 stations 2023: 17 stations 2024: 16 stations For 2024 it is only until 19.jan currently for GB stations, but for the other years I can not see any missing data in those files for GB_5*

[avaldebe]

Some of the 2024 stations are getting new data.

$ for F in  GB_*; do tail -1 $F | cut -d, -f9,11,13-14; done | sort -u | 'grep' 2024-02
CO,hour,mg/m3,2024-02-05 03:00:00 +01:00
CO,hour,mg/m3,2024-02-06 16:00:00 +01:00
CO,hour,mg/m3,2024-02-06 19:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 01:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 02:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 12:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 13:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 14:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 16:00:00 +01:00
NO2,hour,µg/m3,2024-02-05 18:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 01:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 02:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 09:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 11:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 12:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 14:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 15:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 16:00:00 +01:00
NO2,hour,µg/m3,2024-02-06 19:00:00 +01:00
O3,hour,µg/m3,2024-02-05 14:00:00 +01:00
O3,hour,µg/m3,2024-02-05 15:00:00 +01:00
O3,hour,µg/m3,2024-02-05 18:00:00 +01:00
O3,hour,µg/m3,2024-02-06 01:00:00 +01:00
O3,hour,µg/m3,2024-02-06 11:00:00 +01:00
O3,hour,µg/m3,2024-02-06 12:00:00 +01:00
O3,hour,µg/m3,2024-02-06 14:00:00 +01:00
O3,hour,µg/m3,2024-02-06 15:00:00 +01:00
O3,hour,µg/m3,2024-02-06 16:00:00 +01:00
O3,hour,µg/m3,2024-02-06 17:00:00 +01:00
O3,hour,µg/m3,2024-02-06 19:00:00 +01:00
PM10,hour,µg/m3,2024-02-05 17:00:00 +01:00
PM10,hour,µg/m3,2024-02-06 14:00:00 +01:00
PM10,hour,µg/m3,2024-02-06 16:00:00 +01:00
PM10,hour,µg/m3,2024-02-06 19:00:00 +01:00
PM2.5,hour,µg/m3,2024-02-05 12:00:00 +01:00
PM2.5,hour,µg/m3,2024-02-05 13:00:00 +01:00
PM2.5,hour,µg/m3,2024-02-06 16:00:00 +01:00
PM2.5,hour,µg/m3,2024-02-06 19:00:00 +01:00
SO2,hour,µg/m3,2024-02-06 11:00:00 +01:00
SO2,hour,µg/m3,2024-02-06 14:00:00 +01:00
SO2,hour,µg/m3,2024-02-06 15:00:00 +01:00
SO2,hour,µg/m3,2024-02-06 16:00:00 +01:00
SO2,hour,µg/m3,2024-02-06 19:00:00 +01:00

Should I re-download the 2022 and 2023 GB stations?

[avaldebe]

Should I re-download the 2022 and 2023 GB stations?

I downloaded all 2022 and 2023 GB stations and there is no new data.

[avaldebe]

looks like the station metadata has changed a lot since the last time it was updated

$ cd /lustre/storeB/project/aerocom/aerocom1/AEROCOM_OBSDATA/EEA_AQeRep.NRT/
$ wget http://discomap.eea.europa.eu/map/fme/metadata/PanEuropean_metadata.csv
$ wc -l renamed/metadata.csv PanEuropean_metadata.csv
   67257 renamed/metadata.csv
   69272 PanEuropean_metadata.csv

[avaldebe]

should I update the metadata file?

[avaldebe]

should I update the metadata file?

I updated the metadata files on storeA/B, hope it helps

$ wc -l /lustre/store?/project/aerocom/aerocom1/AEROCOM_OBSDATA/EEA_AQeRep.NRT/download/metadata.csv
   69274 /lustre/storeA/project/aerocom/aerocom1/AEROCOM_OBSDATA/EEA_AQeRep.NRT/download/metadata.csv
   69274 /lustre/storeB/project/aerocom/aerocom1/AEROCOM_OBSDATA/EEA_AQeRep.NRT/download/metadata.csv

[jgriesfeller]

The location of a stations is only noted in the metadata file. So it's vital to have that up to date. Without it, the station's data cannot be read. https://github.com/metno/pyaerocom/blob/f8596b2cca36bf53226d1f224c95a01564888b2b/pyaerocom/io/read_eea_aqerep_base.py#L678-L682

[jgriesfeller]

After @andagit complained again today about the missing stations, I checked the existing 17 stations for 2023 against the metadata file. It turned out that all stations had a station classification of either urban-background or urban-traffic.

The problems is that in our config file we have the following station classification we accept:

"area_classification": [
          "rural",
          "rural-nearcity",
          "rural-regional",
          "rural-remote"
        ]

So basically none of the british stations match our classification criteria. Another way of looking at it is that the UK does not submit rural stations anymore.

[andagit]

Just for the record - from EEA we got no. of stations from GB: 2015: 85, 2016: 77, 2017: 76, 2018: 70, 2019: 89, 2020: 34 (brexit,Corona), 2021: 90 (last v2), 2022: 17, 2023: 17

[andagit]

So could report back to ECMWF (Richard Engelen)/CAMS2_82 about the lack of GB stations.