Because this is where the section is subjected to the most rot. Wood is stable when it is continually dry or continually wet. What it can't take is getting wet and drying out over and over. Even pressure treated lumber will eventually fail this way.
EIT, that sounds very logical, but a thought that occured to me is that trees also tend to fail above the ground. Being that poles are made from trees I would guess that the reason is the same. Trees probably have stronger wood in the lower trunk and fewer knots. Diameter change would also factor in.
After Katrina (LA & MS), I saw the same 6' - 7' height situation on the timber posts supporting homes 10 to 12 feet above grade that got hit directly from the initial storm surge. Usually the home was gone, but the posts remained partially. It was even common when the posts rotated in the soil and the house itself got hit as the surge grew.
Concrete piling just rotated. The best coastal solutions seemed to be 8"x24 reinforced block piers (perpendicular to the coast and parallel to the surge direction) on spread footings about 6' deep. - Masonry blow-out panels between if the lower area was enclosed.
Piles and telephone poles are like a cantelever beam. The inflection point (where the bend starts) is somewhat below ground, dependent on the stiffness of the ground and pole and the depth to which the pole is buried. The poles have a taper to them and the point of maximum stress usually ocurrs somewhere above grade. As others have stated, the mechanical properties of the pole are very dependent on growth patterns, condition and moisture content.
The max moment in an embedded pole is right at the ground level. Right at grade you have load acting in one direction only (presumably wind) contributing to the moment. Once you get below grade the soil is counteracting the applied moment, thus reducing the moment in the section. The max moment doesn't occur at the point of inflection, that's a location of zero moment.
If you check the taper, you'll probably find that those that fail above grade have a greater taper.
The failure doesn't occur at the point of maximum moment, but at the point of maximum bending stress, which in a tapered section is going to be somewhat above grade.
I like to break my spaghetti into thirds before cooking.
You'd think you could just grasp a bundle of it tightly at one end, and snap it off clean between two closely spaced fists.
It doesn't work that way; you have to move one hand away from the 'ground', and then it fails a couple of bundle diameters 'up', and never a nice clean break. Just like when a wooden pole snaps.
The spaghetti flying all over the kitchen usually convinces the wife that she doesn't really need my help...
Maybe the ultimate stress at and near the ground level is increased by the way the dirt confines the fibers in the pole. Sort of like the way ties work in a concrete column.
I agree with EIT's first comment about rot - the most likely culprit.
The typical wood piles have their taper such that the top is a larger diameter than the bottom (tip)... although I've see a few telephone poles that I'd exclude from this.
Combine this with the fact that the max moment (unless there is a collar) IS below the ground surface-at the point of fixity, you would expect it to fail below the ground surface because of a higher moment and a smaller cross-section.
Therefore, I'd conclude that the material is, for some reason, weaker at the location of failure. I would guess that there is typically a small amount of wicking of groundwater to a certain height which would give a continual wetting and drying of the wood with changes in groundwater level, therefore causing the more rapid decay.
Noy too many cars hit them below the groundline...
Seriously though, for a tapered pole, with the section varying dramatically along the length of the pole, it only stands to reason that it will fail at the point where the section becomes insufficient for the applied moment.
I also agree with StructuralEIT about the rot issue here. Seen a lot of that with marine pile that always fail in the area of the waterline region of the pile.
Also, with linemen and treated poles, the spikes they use may, with time, penetrate the region of pressure treating, allowiong rot to start. Stands to reason that this area would be toward the base of the pole where more activity occurs and the moment is greatest.
