I'm wondering why such a seemingly simple but very useful tool is not working as it should in ArcGIS 10? The 3D analyst Calculate area and statistic gives an erroneous result. I have tested it with a clipped 1x1km DEM (25 m pixel size), it returns 2D area: 950625 sq m, surface area: 992885.32 sq m, which is less than it should be. This problem was present back since ArcGIS 8, and it is an unpleasant surprise that it was not solved so far (correct me if I'm wrong!).
In the method described by Jakubsisak, in ArcGIS 9.3.1 only the 3D area is calculated, and is still incorrectly less than 1 sq km.
Could anybody please recommend a simple and accurate method of 3D surface calculation?