An Extension of Multiquadric Method Based on Trend Analysis for Surface Construction

The representation of spatial discrete data is crucial for the analysis meteorology, agriculture, geological exploration and other fields. Among various methods scattered interpolation, multiquadric (MQ) method most favorable in field surface construction. However, classical MQ has accuracy concern on boundary time consuming large set. This study proposes an algorithm by integrating with trend analysis. In which, low-order polynomial equation firstly used to model overall trend. Then, applied fit residual after removal from data. Our implementation can eliminate distortion missing areas method, modeling efficiency be improved significantly since local divide into a group subsurfaces. proposed are validated synthetic model. performance developed further examined elevation collected Tibet seabed strait Norway. results show that equivalent resolution, much more efficient than well-developed Kriging method.