Shooting at the foot (and missing)
Sometimes you can say something that proves to be true but illustrate it poorly enough that readers don’t believe you.
Case in point: effect of basement lithologies on the grade of rivers (in this case, for how we interpret paleoriver systems). Manny Gabet (among others) has suggested that this causes the azimuthal variation in grade of Eocene paleochannels, and he illustrated this with the example shown below:
Now one thing here is the distance axis on the three plots: is it measured along the channel, or airline? One might think along the channel. But in any event, look at the distance from B to C on the map and then on the plots. Airline it eyeballs to about 6 km on the map, but only 4 km on the plots. It is even worse if you measure along the river. So this quick eyeball reality check would make many readers pause and question the conclusion here.
So GG here has carried this slightly further, which results in the map below:
Here the colored channels are from a USGS database placed on a shaded relief DEM. Clearly a few details of the channels in Gabet’s map are a little off, but now the distances are directly calculated more accurately. And that B-C leg seems a real problem. But then the question was, where does the Consumnes River really drop to ~170m elevation? Maybe point C is misplaced. The answer turns out to be, not at point C on Gabet’s map, but at point C’, shown above. Now C’ is actually closer to B than Gabet had plotted, so rather than the B-C leg being shallower, as one’s initial guess from the map error would have suggested, it is in fact steeper.
For grins, GG replotted the points above with the along-channel distances (left, below) or airline distances (right, below) and added a stretch of the Middle Fork of the Consumnes above B (point “F” is just below the “F” in “Mid. Fk”):
Now this hasn’t been adjusted for drainage area size, but the Middle Fork (point F) and North Fork (point A) are pretty similar in drainage area, so that difference is a fair indication of the effect of cross versus parallel to strike channel geometry in the weaker rocks. But you can also see the effect of just crossing resistant rock below B (there is very little increase in drainage area below B). Thus even though the trellis drainage will yield azimuthal variations in grade, the crossing of resistant rock will introduce scatter within an azimuth that could be comparable in magnitude.
Anyways, the point here isn’t to fully address the question of interpreting the paleochannel gradients (though that motivated looking at this closely). There’s a lot more going on than just this. The point is that the author in this case risked having his point questioned by having been a bit sloppy with his map. Moral of the story is that care in the details can matter. GG was ready to write this off after comparing the map to the plots.