Although the amplitude of the fluctuations in ocean level is approximately the same within the polar circles and in the tropical zones, the prerequisites for the formation of submarine valleys in these regions are totally different. In the tropical zones, as the ocean level falls, the rivers follow the advancing coastline, incising their valleys more deeply in the process. With the characteristic abundance of atmospheric precipitation in the tropical zones it is readily seen that the erosion processes will be fairly vigorous, forming submarine canyons along the continental coasts.
The conditions created by the fall of the ocean level will be totally different in high latitudes. In case of a slight shift of poles the fluctuation in ocean level at these latitudes will be small, and the rivers will incise their valleys only to a small depth, so that the submarine canyons will soon be buried under the shifting sands after the return of the sea. If the poles shift considerably and the fall of the ocean level is large, this phenomenon will be accompanied with such a considerable drop in the mean temperatures that glaciation would inevitably occur at these latitudes. The ice sheet would cover the dry land and possibly also the sea shallows, preventing the erosive activity of the rivers.
To illustrate the above hypothesis, let us examine the possibility of the formation of the submarine valleys near the mouths of Siberian rivers flowing into the Arctic Ocean, choosing the largest rivers for this purpose, namely the Ob, the Yenisei, and the Lena. Their respective mouths lie at 67°N., 70°N, and 73°N. At these latitudes the approach of the pole through one degree causes a drop of 220 m in ocean level (see Table 1); thus, the rivers will be able to incise their valleys to a depth of 220m, i. e., they will remain within the boundaries of the continental shelf.
Such valleys may readily become buried afterward. Therefore, in order to allow for the possibility of the formation of sufficiently deep and stable submarine canyons we must assume larger shifts of the pole, of about 2.5-3°. An approach of the pole through 3° will depress the ocean level by 3X220 = 660 m, and in this case the submarine valleys will be much better preserved, provided they have been incised by the river to such a depth.
However, a drop in ocean level will also cause a drop in mean temper ature.
Let us calculate the variation in mean annual temperature by the method provided by Khizanashvili /10, pp.61 —64/. A 660m depression in the ocean level will decrease the mean annual temperature by 660/200 = 3.3 °C, while the increase in the latitude will reduce the mean annual temperature by 3X0.6 = 1.8 °C . The total temperature drop will amount to 3.3 + 1.8 = 5.1 °C . It may be assumed that a drop of 5 °C in the mean annual temperature is sufficient to cause glaciation. Evidently, favorable conditions for the formation of submarine canyons may not occur in this region even in cases of considerable shifts of the poles.
The above discussion was not intended to exclude the possibility of the formation of submarine valleys in high latitudes. The formation of canyons there may be facilitated by other factors, such as warm sea currents limiting the spread of glaciation which might result from the drop in ocean level. Furthermore, it was noted by Saks /9, p. 34/ that in high latitudes one may encounter submerged valleys of glacial or other origin. Consequently, our hypothesis should be interpreted to mean that the probability of the formation of submarine valleys is considerably lower in the polar regions than in the middle and low latitudes. This possibly explains why no submarine canyons have been discovered thus far in the Antarctic /3, 9, 11/, while their known number in the Arctic Ocean is small /9, 11/; in the Pacific basin the northernmost canyon was found at a latitude of 54°, according to Shepard /12, p. 228/.
Thus, in our discussion of the fluctuations in the level of the oceanic spheroid with respect to the possibility of formation of submarine valleys we have seen that a drop in ocean level should create conditions con ducive to such formation in the middle and low latitudes of the Earth, but these conditions may not arise in high latitudes.
Distribution of atmospheric precipitation. Our second prerequisite for the formation of submarine valleys was a sufficient amount of atmospheric precipitation. As in the preceding case, we are interested in the distribution of atmospheric precipitation at different terrestrial latitudes.