# Gridding interpolation

**Keywords:** mapping, gridding, interpolation, approximation, surface interpolation, kriging, minimum curvature,a powerful contouring, gridding, and surface mapping package for scientists and engineers smooth surface fitting interpolation of a function of many variables representation and approximation of surfaces

**Description:** Interpolation and approximation of surface. Result is comparable with the surface created by Kriging or Minimum curvature method.

- It is very fast. surfit is multithreading application, it can use all of your computer processors. Also you can use boosters to speed up computation speed even more.
- Quality of resulting surface is comparable with a surface created by Kriging or Minimum curvature method.
- It can deal with
**HUGE**data sets. - With surfit you can build your surface from very various data with various uncertainty. In short, you can use scattered points, fault lines, trend surfaces, contour lines, inequalities, integral conditions and other. Also it is possible to order data by priority or combine with weights. Read more about surfit gridding algorithm .
- Supports various data formats .

The screenshots and examples describe ` surfit` features in detail, but don't hurry, look at the pictures below, they will give you idea of that surfit can do (some images were made in Surfer 11):

surfit allows you to easy choose between interpolation or approximation of data points. The strength of approximation is a changable parameter:

Look at the images below - there are results of porosity field reconstruction. The first surface was built from porosity measurments, given in scattered data points (black dots). The second surface was built from the same data plus new condition was added: average porosity value inside data points convex hull (purple line on images) should be equal to 25%. This value is litte lower than average porosity calculated from data points (25.7%). Adding average value condition is one of surfit features. Using it you can add additional information about estimating field. Also you can set "weighted average value" condition.

Consider usual task - we need to reconstruct reservoir top surface from the set of digitized isolines (level curves). Common way to solve this task is to convert isolines into scattered data points and then build the grid using some gridding algorithm. On the first picture below you can see these scateered data points and fault lines. Result of gridding with surfit is on the second picture. The third picture - result of gridding by Surfer 11 (MINC algorithm). As you can see level curves aren't smooth as we want:

Now we will try to solve this problem in other way. We will not convert isolines into scattered data points - we will use isolines as isolines in gridding method, i.e. we add condition that isolines, traced from the resulting grid should be equal to given isolines:

On the first picture sorce isolines are blue, fault lines are red. On the second picture is the result of gridding with surfit. So, as you can see resulting grid isolines are smooth. And I want to notice that this result was obtained without any additional smoothing!

In this example we have setof 1005 scattered data points with very noisy data. So, we going to approximate these points, instead of building interpolation surface

3D seismic data presented with 33673 data points, this data is more accuarate then 2D seismic and covers smaller area. Again we are going to approximate this data, but resulting surface should be not so smooth:

We have wellpicks data in points and we absolutely belive to this data. That's why our top surface should interpolate wellpicks. We will use 2D and 3D seismic surfaces as trends with 3D seismic higher priority