CrIS Observation Error Covariance
Wed, Nov 23, 20161 Introduction
2 Chris Barnet ()
- Representativenss probably dominated by time with water
- He did linear interpolation of 3-hour to see what is required, and it implied very small times
- NWP interpolates to 1 hour for observations, but sondes are already at the exact analysis times, so no interpolation of model?
- Hamming implies they must use more channels if they can't utilize the low noise of CrIS