In order to assess the value of remote sounding data for numerical weather prediction, parallel sets of analyses were produced with (SAT) and without (NOSAT) the sounding data from the experimental Nimbus-6 and operational NOAA-4 satellites for the Data Systems Test periods, 18 August–4 September 1975 (DST-5), and 1 February–4 March 1976 (DST-6). All other factors, i.e., the assimilation method and remainder of the data base, were identical for both the SAT and NOSAT modes of each set. For selected days of DST-5 and DST-6, forecasts were generated through 72 h over the Northern Hemisphere. Differences between corresponding SAT and NOSAT analyses and the forecasts produced therefrom were assessed via a set of objective and subjective procedures, including evaluation of standard skill scores and judgment by experienced meteorologists.
The effect of remote temperature soundings in the NMC DST experiments was generally small and of inconsistent sign, i.e., beneficial in some cases, harmful in others. The average of these positive and negative contributions over the cases considered proved slightly positive for the DST-6 period and slightly negative for the DST-5 period. Neither result was judged of much meteorological consequence. Overall, we conclude that the remote soundings had little impact on forecasts in the Northern Hemisphere.
However, systematic differences were noted between the SAT and NOSAT analyses—the amplitude of weather systems was consistently less in the SAT mode. The reduced amplitude reflected an intrinsic characteristic of the remote soundings; viz., the tendency for the satellite temperature retrievals to underestimate the spatial variance in the thermal structure of the atmosphere.