First, the attributes to be copied are defined.

In case of a 3D model, you can select one of three algorithms to compute the mean value:

The horizontal mean from an element layer to a node layer is computed from the element values horizontally adjacent to the node in the selected element layer and assigned to the corresponding node in the selected node layer (e.g. UNTE layer 1 is averaged to ZKOR layer 5).

The horizontal mean obtained from a node layer applied to an element layer is computed from the (3 or 4) corner node values in the selected node layer, and is assigned to the corresponding element in the selected element layer (e.g. ZKOR layer 7 is averaged to UNTE layer 1).

The 3D mean value for only one node layer is computed in the same way as the element values of the two element layers adjacent to the selected node layer. In case the node layer k was selected, the algorithm uses the values from the element layers k-1 and k (e.g. KSPE layer 7 as mean value from SPEI).

The 3D mean value for only one element is computed from the corner node values of the 3D element (3 or 4 nodes at the top and the bottom). In case the element layer k was selected, the algorithm uses the values from the 3 or 4 corner nodes in the node layers k and k+1 (e.g. SPEI layer 1 as mean value from KSPE).

The 3D mean value for all node layers is computed in the same way as the 3D mean value for one node layer. But the resulting value is assigned to all node layers instead (e.g. KSPE for all node layers as mean value from SPEI)

The 3D mean value for all element layers is computed in the same fashion as the 3D mean value for one element layer. However, the resulting value is assigned to all element layers instead (e.g. SPEI for all element layers as mean value from KSPE).

You can choose between a normal and a logarithmic mean value determination:

Normal (arithmetic) mean value μ:

, w_i are the values in the adjacent elements or nodes which should be converted and n is the number of values.

Logarithmic mean value μ:

, w_i are the values in the adjacent elements or nodes which should be converted and n is the number of values. This method is useful for positive element values with different magnitudes i.e. permeabilities.

Alternative to the arithmetic mean the algorithm can perform weighting by element areas where values from larger elements are weighted more heavily than those from smaller elements. A weighting of the element area is via the computation rule:

Normal:

e.g.

Logarithmic:

Where F_i is the area oft he adjacent elements.

For the conversion from node to element values just a normal (uniform) weighting is possible:

,

w_i are the values in the adjacent nodes which should be converted and n is the number of values.

Consider mass conservation:

If the mass conservation is switched on (e.g. conversion of FLAE into KNOT), the element values (e.g. inflow rates) are proportionally distributed to the corner nodes associated with the element. Then weighting and the type of mean (logarithmic / normal) are unimportant. Each node receives from an adjacent quadrilateral a quarter, or rather from an adjacent triangle a third, of the element rate, etc.