Premium
Variational cloud‐clearing with TOVS data
Author(s) -
Joiner J.,
Rokke L.
Publication year - 2000
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.49712656316
Subject(s) - cloud computing , data assimilation , clearing , environmental science , satellite , remote sensing , meteorology , radiative transfer , computer science , engineering , physics , geology , finance , quantum mechanics , economics , aerospace engineering , operating system
A number of studies have shown that the use of passive microwave and infrared satellite observations in data assimilation systems can increase forecast skill. Considerable effort has been expended over the past two decades, particularly with the TIROS Operational Vertical Sounder (TOVS), to achieve this result. The positive impact on forecast skill is a result of more rigorous treatment of quality control, improvements in systematic error correction schemes, and advances in data assimilation systems. Yet, there Still remains potential for improving the use of satellite data, particularly cloud‐contaminated observations, in data assimilation. Here, we use a one‐dimensional variational framework (1DVAR) as a first step towards improving the treatment of cloudy data by cloud‐clearing. Cloud‐clearing is a procedure that removes cloud radiative effects through comparison of partly cloudy adjacent pixels. The JDVAR approach simultaneously extracts cloud‐clearing parameters and information about the atmospheric and surface state from microwave and infrared observations. The variational framework ensures that the state estimate is consistent with all available measurements. The 1DVAR cloud‐clearing approach can also be extended to three or four dimensions (3DVAR, 4DVAR). Our TOVS cloud‐clearing implementation allows for complex cloud structures, including multiple cloud layers with wavelength‐dependent radiative properties. We present preliminary results of our 1DVAR cloud‐clearing implementation with TOVS. The results suggest that there is useful information in the cloud‐cleared data.