Posted Mon, 28 Jul 2014 23:25:36 GMT by Debora Janos

I am using FEPEST for calibrating my 3-D model. My model is of a regional scale and my elements are fairly big. I have a ton of observation points that may fall anywhere within the elements. My observation parameters are the hydraulic heads which are calculated on the nodes. I was wondering if the observation points are compared to the heads at the closest node or to the interpolated calculated heads at the location of the observation points. If so, what interpolation technique is used?

Thanks for helping!

Posted Thu, 07 Aug 2014 01:10:44 GMT by Jarrah Muller Civil Engineer
Hi Debora,

The heads are interpolated at the observation points.

You can test this in a few minutes by making a test model with a single large triangular element, putting an observation point in the middle, and assigning fixed heads to the three vertices.

I don't have a reference for the interpolation method used. However, it appears that the three heads at the triangle vertices are used to define a plane, and that the head observations are then drawn off that plane (i.e. 2D linear interploation).

(see attachment)

You must be signed in to post in this forum.