Search

Observations in channel 1 (Ch. 1, 50.3 GHz) and channel 2 (Ch. 2, 53.74 GHz) of the Microwave Sounding Unit (MSU) over the convective areas of tropical oceans are analysed to reveal the nature of extinction (contamination) in these data. From this analysis we find Ch. 2 data are not free from the influence of clouds and rain. Extinction due to clouds and rain manifests primarily as emission in Ch. 1, and as absorption in Ch. 2. Scattering due to hydrometeors in these channels apparently is of secondary importance. Furthermore we show, in the convective areas of tropical oceans, contamination due to hydrometeors in MSU Ch. 2 data is significant and it is extensive in area. Based on this study we conclude Spencer, Christy, and Grody (this issue) underestimate this contamination.

Using Microwave Sounding Unit (MSU) channel 2 (Ch. 2, 53.74 GHz) data, Spencer and Christy (1992a) determined that the earth exhibits no temperature trend in the period 1979-90, while other authors find a temperature increase of roughly 0.1 K. Based on a theoretical analysis Prabhakara et al. (1995) showed that the information about the global atmospheric temperature deduced from MSU Ch. 2 observations has a small contamination, [Delta][T.sub.2], as a result of the attenuation due to hydrometeors in the atmosphere. A method is developed in this study, that utilizes coincident measurements made by MSU in Ch. 1 (50.3 GHz), to estimate this [Delta][T.sub.2] over the global oceans. The magnitude of [Delta][T.sub.2] is found to be about 1 K over significant pans of the tropical oceanic rain belts and about 0.25 K over minor portions of the mid-latitude oceanic storm tracks. Due to events such as El Nioo, there is variability from year to year in the rain areas and rain intensity leading to significant change in the patterns of [Delta][T.sub.2]. The patterns of [Delta][T.sub.2] derived for March 82 and March 83 reveal such a change. When averaged over the global oceans, from 50 [degrees] N to 50 [degrees] S, [Delta][T.sub.2] has a value of 0.25 and 0.29 K for March 1982 and 1983, respectively. Due to these reasons the interannual temperature change derived by Spencer and Christy from MSU Ch. 2 will contain a residual hydrometeor effect. Thus in evaluating decadal trend of the global mean temperature of the order of 0.1 K from MSU Ch. 2 data one has to take into account completely the contamination due to hydrometeors.