I have encountered four consecutive years and several consecutive months of missing observed temperature data in a data obtained from National Meteorological Agency. How can fill the missing data using the satellite temperature data? Please help me in this regard.
Efrem
Case-I : There are many techniques to fill the missing value without any alternative source of data sources namely MICE, nearest neighbour approach, NIPALS algorithm.
Case-II : If you have satellite temperature data as an alternative source, that gives you an additional advantage. Follow Case-I and you can do a bias correction using QQ mapping as said by Mr. Bai. We have tested MICE technique to fill missing data in a 40-year monthly dataset of temperature. We got good accuracy with an alternate data source.
P.S. : Before doing this all exercise you may take a portion of you observational data where there is not a single missing data. Remove 10-20% of data intentionally and follow Case-I. Compare that with your original data as well as satellite data and do a bias correction.
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere