I have downloaded the climate time series for 1200 weather stations in europe but some values can be missing for several months sometimes.
I am looking to have hourly resolution. I figured that if the the missing values are only about few hours, then I could fill them in with the values of the previous day - the differences should be minimal.
But how to engineer time series data when I need to fill in missing data for a whole month for example?
I figured that I could cluster together all the stations that belong to the same climate type (mediterrenean, temperate etc) based on their longitude and latitude, average the values and scale them to fill in the mising values of the stations with missing values. The thing is I wasn't able to find the "borders" of the climate types.
Where can I find such data and do you know if there is a better method to engineer missing values?
thanks!
أكثر...
I am looking to have hourly resolution. I figured that if the the missing values are only about few hours, then I could fill them in with the values of the previous day - the differences should be minimal.
But how to engineer time series data when I need to fill in missing data for a whole month for example?
I figured that I could cluster together all the stations that belong to the same climate type (mediterrenean, temperate etc) based on their longitude and latitude, average the values and scale them to fill in the mising values of the stations with missing values. The thing is I wasn't able to find the "borders" of the climate types.
Where can I find such data and do you know if there is a better method to engineer missing values?
thanks!
أكثر...