Kriging is a statistical technique used to estimate the value of a random variable at an unobserved point in space. The spelling of "kriging" is based on the name of the Norwegian mathematician, Danie Johan Krig, who developed the method in the 1950s. The IPA phonetic transcription for "kriging" is /ˈkridʒɪŋ/, with the stress on the first syllable. This corresponds to the pronunciation of the word as "KREE-jing", with a hard "g" at the end.
Kriging is a geostatistical technique used to estimate values of unknown variables at unobserved locations in a spatial domain based on observed data at nearby locations. In simpler terms, it is a method for spatial interpolation that predicts values for unsampled areas based on available data from surrounding areas.
The process of kriging involves analyzing the spatial correlation structure of the available data and using this information to create a model for the spatial distribution of the variable of interest. This model includes a best-fitting mathematical function called a variogram, which quantifies the spatial dependency or similarity between pairs of data points. By evaluating the variogram, kriging estimates the spatial autocorrelation and determines the optimal weights to be assigned to each observed data point for predicting values at unsampled locations.
The key advantage of kriging over other interpolation methods lies in its ability to estimate values with a high degree of precision, incorporating both the overall trend of the data as well as the local variability. It provides not only an estimate of the unknown variable but also an assessment of the uncertainty associated with the estimation.
Kriging is commonly employed in various fields such as environmental sciences, geology, agriculture, and remote sensing. By utilizing the spatial patterns observed in the existing data, this technique allows researchers and analysts to make informed decisions and generate accurate maps or models for planning, resource management, risk assessment, and other spatial analysis applications.
The word "kriging" is derived from the last name of Danish mathematician Danie G. Krige. In the early 1950s, Krige developed a statistical interpolation technique, which later came to be known as "kriging". The method involves estimating the unknown values of a spatially distributed variable based on surrounding measurements. The name "kriging" was adopted to honor Krige for his significant contributions to geostatistics.