SECT. 2] LARGE-SCALE INTERACTIONS 121 



(after Budyko) is presented in Fig. 11. Before deducing any consequences to 

 geophysics, it is well to make careful tests of any residually deduced quantity 

 such as Qio, particularly to determine whether its important features could be 

 due to uncertainties in evaluation. Table VI shows the oceanic heat -balance 

 components integrated longitudinally and averaged over 10°-latitude belts. 

 On the right we compare the earth's surface radiation balance computed from 

 the tables of London (1957) which, unfortunately, could not be separated into 

 land and ocean regions. The agreement with Budyko 's surface radiation 

 balance, B, is excellent equatorward of 30° and fair jJoleward. 



Table VI 



Mean Annual Distribution with Latitude of the Heat-Balance Components 



of the Ocean Surface 



Units : kg cal/cm^ year 



Since the mean latitudinal dependence of Qe (Fig. 4) is agreed on by all 

 authors to a much smaller margin than radiation, or the size of the residual it- 

 self, we may substitute London's R for Budyko's in (1) to test Qvo- While the 

 negative values poleward of 30° are much reduced thereby, its main features 

 from 0-30°N are reproduced, as is the change in sign to negative north of 30°N. 

 The next test is a comparison of Fig. 11 with a similar map constructed by 

 Sverdrup (1957) shown in Table VII. In making the balance in equation (1), 

 Sverdrup used the Qs and Qe of Jacobs together with radiation balance maps of 

 Kimball (1928). The agreement between Budyko and Sverdrup is good in the 

 Atlantic, poor in the Pacific and fair when averaged longitudinally. Briefly, 

 Budyko's chart (relative to Sverdrup's) shows much diminished effects of the 



