Spatial Analysis Using Big Data

4.5. Universal Kriging

In OK, it was assumed that the variable y(s), including the trend component m(s), are second-order stationary, and E [ y ( s ) ] = m ( s ) = m ¯ holds at all points within the area. Meanwhile, in universal Kriging (UK), y(s) can be expressed as follows:

y ( s ) = x ( s ) β + u ( s ) ,

(4.4.18)

where x(s) is a 1 × K explanatory variable vector. Therefore, UK is a Kriging method when the trend components can be expressed by a linear combination of the known variables. According to ...