A comparison of the impact of TOVS and ATOVS satellite sounding data on the accuracy of numerical weather forecasts

Citation
Sj. English et al., A comparison of the impact of TOVS and ATOVS satellite sounding data on the accuracy of numerical weather forecasts, Q J R METEO, 126(569), 2000, pp. 2911-2931
Citations number
26
Categorie Soggetti
Earth Sciences
Journal title
QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY
ISSN journal
00359009 → ACNP
Volume
126
Issue
569
Year of publication
2000
Part
A
Pages
2911 - 2931
Database
ISI
SICI code
0035-9009(200010)126:569<2911:ACOTIO>2.0.ZU;2-0
Abstract
The Advanced TIROS Operational Vertical Sounder (ATOVS) was launched on the NOAA-15 satellite in May 1998. This provided a very significant improvemen t in the information available from meteorological polar-orbiting satellite s compared with the previous TIROS Operational Vertical Sounder system, par ticularly for humidity and vertical resolution of temperature in cloudy are as. In preparation for assimilation of the observations into a three-dimens ional analysis of atmospheric temperature and humidity, the observations ha ve been compared with calculated top-of-atmosphere brightness temperatures computed from numerical weather prediction model profiles of temperature an d humidity. Differences between observed and modelled brightness temperatur e are small. In some parts of the tropics and northern hemisphere the stand ard deviation of these differences for the tropospheric Advanced Microwave Sounding Unit sounding channels is only marginally higher than the radiomet ric noise of the observations. Early in 1999 a series of observation-system experiments were completed in which ATOVS observations were assimilated us ing a one-dimensional variational analysis. No use of the new humidity info rmation could be made because of interference problems experienced by the m icrowave humidity sounder on ATOVS. Nonetheless, these experiments showed t hat the assimilation of the new temperature information provided by the rad iance observations reduces forecast errors by as much as 20% in the souther n hemisphere and 5% in the northern hemisphere. Further improvements have b een found by assimilating more data over land. The major impact arises from the microwave channels. Whilst forward-model errors may be slightly lower for the microwave channels than the infrared channels the primary reason is the provision of sounding information in active weather systems, which are usually cloudy.