The way I was taught to find depth with L-Rods (with a sample on the tips) was to find what we called the "aura" (circle of energy) around the target. However, for an underground stream of water, which is a long line instead of a single point, I assume the "aura" would be two parallel lines on either side (I assume, because I have never water dowsed; I am only interested in large caches). By the way, if you want to use electrical dowsing to find water, then you can use an harmonic of the output frequency listed on your microwave oven (for example: mine is 2450 MHz, so you can use 245 Hz, which I have verified; this implies molecular resonance). If you hum the frequency, then you can use mechanical resonance. The way we used to get the aura was to walk towards the target and detect 3 points (nearest edge, target, farthest edge). Normally, the radius of the aura was the depth, which implies the energy was traveling to the surface at a 45 degree angle (vertical triangulation). But sometimes there might be a double aura, and neither aura would be the correct depth. Also, the aura is typically slanted to one side, like an ellipse, which is probably some kind of magnetic or gravitational distortion (in those cases, take the average). Checking the aura as you dig is a smart thing to do; by doing so, we noticed that sometimes the aura would decrease in radius, and other times it would increase in radius. Whenever it decreased, it would disappear at the hole just as you reached the target. If you reach the predicted depth and no object is found, but the target energy jumps to a new location nearby, then you just dug a "false target" (a reflection of a real "aged target" nearby). Of course, determining depth means nothing, if you're on the wrong spot (except it tells you when to stop digging).
