I think you're bringing up a good point that I should have been clearer about. You're right that you can't completely separate accuracy from resolution since features would be unrecognizable regardless of resolution if the individual points weren't sufficiently accurate relative to each other. I should have made it clear that I was talking about absolute accuracy, which is generally what would be considered for the contours Randy wanted. I really don't have enough information from what Randy said to tell what he really needs. All I can really tell is that he thinks he needs a certain density of points in order to use a contour interval of a certain amount, and that isn't the way it works. For many practical applications, you really would need both sufficient resolution and accuracy. For instance, if you need to judge if a levee is sufficiently high, you need to have the resolution to see the levee and the accuracy (at least relative to the water level) to make the judgment.
I completely agree with you that increasing point density does not necessarily result in better feature depiction. However, I don't think there is a one size fits all answer for point density either. Every case is different and it depends on what you're looking for. Also pulse density is not equal to point density, so if you're trying to see features under trees, you'll need a higher pulse density just to get enough points on the ground to see those features you wanted. That might make the non-tree areas seem over sampled.
Kirk Waters
July 9, 2012
Patrick,
I think you're bringing up a good point that I should have been clearer about. You're right that you can't completely separate accuracy from resolution since features would be unrecognizable regardless of resolution if the individual points weren't sufficiently accurate relative to each other. I should have made it clear that I was talking about absolute accuracy, which is generally what would be considered for the contours Randy wanted. I really don't have enough information from what Randy said to tell what he really needs. All I can really tell is that he thinks he needs a certain density of points in order to use a contour interval of a certain amount, and that isn't the way it works. For many practical applications, you really would need both sufficient resolution and accuracy. For instance, if you need to judge if a levee is sufficiently high, you need to have the resolution to see the levee and the accuracy (at least relative to the water level) to make the judgment.
I completely agree with you that increasing point density does not necessarily result in better feature depiction. However, I don't think there is a one size fits all answer for point density either. Every case is different and it depends on what you're looking for. Also pulse density is not equal to point density, so if you're trying to see features under trees, you'll need a higher pulse density just to get enough points on the ground to see those features you wanted. That might make the non-tree areas seem over sampled.
Kirk