I have to fill the missing data for precipitation stations. I have 28 stations data and I am going to select 6 stations. The data availability differs for different stations. Can you suggest me some methods for filling missing data for the precipitation stations.
Regards,
Juthika
I am working on a river basin. I have calculated the distance based on "Haversine' equation based on their latitude and longitude
Juthika, I have done this for the Columbia River basin, USA. Please have a look at this paper. Best,
Article Estimation of Columbia River Virgin Flow: 1879 to 1928
If you keep it simple by ignoring the time-component, you might use k-nearest neighbor imputation or an EM-algorithm, both e.g. available in R package VIM.
If the time component is important, R package Amelia might be your friend.
Another question is why you have missing values. Are they missing completely at random or missing at random. Or are they missing because the rainfall was under the detection limit of the measurement unit?
Thanks Matthias for your answer. Actually I do not have proper information about why the data is missing. I have collected the data from Meteorological Department.
- Badges
- Science topic