Download Met Office data

The service below allows you to download hourly or daily data published by the Met Office, you can download by individual sites or all sites.
If your system does not allow active content, you can enter the date manually with the following format: dd/mm/yyyy. Please note, this is a beta service at the moment.

Comments

trailing comma after each row

Dear Data.Gov guys

First of all, great service, thanks!

Observation Data

Issue 1

Each row of data (except the header row) has a comma at the end. This is causing my ETL to think there is one more column than there is. Please remove

Eg:

Current:

"3002","BALTASOUND (3002)","60.7490","-0.8540","Orkney & Shetland","00:00","2014-06-01","SSW","6","","22000","8.20","1021","F","(Black) Low-level cloud",

should be:

"3002","BALTASOUND (3002)","60.7490","-0.8540","Orkney & Shetland","00:00","2014-06-01","SSW","6","","22000","8.20","1021","F","(Black) Low-level cloud"

Issue 2

The filename is some sort of hash (eg 9e5b22cc3327481991e5ffb3245a34a8.csv). Can you change this to "Observations_[Location code or All]_yyyymmdd_hhmm.csv"

Daily Forecast and Three-Hourly Forecast

Only text datums have speech-mark text qualifiers. Please can you use speech-mark text qualifiers on all the fields or none of the fields. Preferably all.

Other comments

If poss, please can we choose alternative delimiters. Eg tab, pipe (|), dollar ($).

Thanks

Fil

 Flag as offensive 

Observational data

Just a few comments about the observational data:

  1. To keep the data download as small as possible it seems to me that lat and longs to 4 decimal places is a bit over the top, most of the time (98%) the 4th decimal place is a zero anyway, and dropping it to 3 dp would save 130+ superfulous characters.
  2. Why not just include the stations site code and dispense with the site name, latitude, longitude and region which make up at least 50% of the string of data for each observation? You could add a link to a separate text files thet includes all location data, and by doing this you could drastically reduce the overall download size. This is how relational databases came about but I'm sure you know all about that...
  3. If you reduced the size of one hours observational data (see 2), you could then offer users a way of downloading data for an entire day in one go instead of coming back 24 times.
  4. Reduce the precision for air temperature from 2 to 1, the second 2nd decimal place is redundant as it is always zero.
  5. Instead of rounding MSLP values as you do, why not list them to one decimal place just as they are reported?
  6. Why not include a dew point or a relative humidity to make the content a bit richer.
  7. All these sites have automatic rainfall gauges and a lot of them have sunshine recorders, so why not finish the job and add hourly rainfall and sunshine values for sites that have the data and make it so much more useful for anyone interested in investigating the climate of any site in the UK.

Bruce.

 Flag as offensive