Spatial Semantic Kriging

The spatial semantic kriging (SemK) based spatial interpolation method is applied for the interpolation of meteorological parameters, aiming to enhance the accuracy in results. The SemK considers the semantic properties of the terrain, which is influential to the meteorological parameters and incorporates into the prediction process. One such property is the terrestrial land-use/land-cover (LULC) distribution. An ontology hierarchy is built with the available LULC classes to find the influence of each of the classes to the land surface temperature (LST). This interpolation process belongs to the family of kriging and extends the the ordinary kriging (OK) based spatial interpolation with LULC knowledge. The empirical experiments show that this auxiliary knowledge is highly significant to achieve more accuracy in prediction. The theoretical performance analysis is also carried out in this chapter to prove the efficiency of SemK over other existing methods.