The quality of the publicly available Lidar topo data in our area is very good and in some areas excellent.... Except for the centerlines/beds of creeks, branches, etc. Has anyone else noticed that it looks like someone left off a breakline? I've overlaid our field run topo on the Lidar contours and am amazed how everything seems to fit well within national mapping standards. Until I look at streams, etc. The Lidar topo shows a bunch of "pools" along the streams. The only time I've seen that is when someone(yes, me) left off a breakline and the program interpolated from one bank to the other. I'm not at all complaining, but would like to see the Lidar data get even better.
Me. "What's the difference?"
T.C. Carroll "It's the difference between right and wrong!"
LiDAR output is somewhat uniform grid of discrete ground shots. Break lines can be combined with a LiDAR data set to create a better DTM but that is typically out of scope not for small scall (large area) project.
My guess would be there was water flowing or it was extremely wet in those areas when the flights took place. If they can't be confident in the returns, they'll often just remove those areas from the dataset. If it's a large enough area, like a lake or bay, it might be designated with NoData values - but if not the data is "bridged" across the centerline gap, resulting in the pool effect.
It might also be an effect of sampling the data. Lots of large-scale datasets only contain a few points per square meter by the time processing is complete, and that might get downsampled even more when generating the raster DEMs, depending on the resolution.
Just the cartographer in me speaking, but the larger the map scale the smaller the area mapped on a given sheet size.