A crossvalidation-based comparison of kriging and IDW in local GNSS/levelling quasigeoid modelling

Abstract

This study compares two interpolation methods in the problem of a local GNSS/levelling (quasi) geoid modelling. It uses raw data, no global geopotential model is involved. The methods differ as to the complexity of modelling procedure and theoretical background, they are ordinary kriging/least-squares collocation with constant trend and inverse distance weighting (IDW). The comparison itself was done through leave-one-out and random (Monte Carlo) cross-validation. Ordinary kriging and IDW performance was tested with a local (using limited number of data) and global (using all available data) neighbourhoods using various planar covariance function models in case of kriging and various exponents (power parameter) in case of IDW. For the study area both methods assure an overall accuracy level, measured by mean absolute error, root mean square error and median absolute error, of less than 1 cm. Although the method of IDW is much simpler, a suitably selected parameters (also trend removal) may contribute to differences between methods that are virtually negligible (fraction of a millimetre).

Elevation Accuracy of Forest Road Maps Derived from Aerial Imaging, Airborne Laser Scanning and Mobile Laser Scanning Data

“…In all these cases, an alternative remote sensing technology or geospatial technique could be used within a 95% confidence interval without significant changes in the elevation accuracy. Since these findings are in line with other studies, e.g., [23,[57][58][59][60], it can be stated that in specific cases it is possible to replace MLS technology with the less economically demanding ALS technology. This is particularly feasible if there is no need for forest road maps with a high level of detail.…”

Research on gravity field modelling and gravimetry in Poland in 2019–2022

“…However, differences between results within a method and between methods in the whole range of steering parameters may become significant. Generally, for the area investigated, the accuracy in terms of absolute, root mean square and median absolute errors, is below 1 cm (Ligas et al, 2022).…”

Spatial Variability of Soil Physical Properties in Southern Mankayan, Benguet, Philippines

“…The interpolations were performed using geostatistical and spatial analyst tools in ArcGIS Pro 3.1 in order to generate spatial distribution of soil physical properties in southern Mankayan. IDW does not require solving any system of equations for the weights since it rests on a priori assumption of change of weights with distance-decay (Ligas et al, 2022). The cross-validation technique was applied using the geostatistical analyst feature in ArcGIS Pro.…”

“…The regression model for particle density (Figure 8C) was deemed as the most accurate interpolation among the distribution maps since the predicted values have a lower margin of error from the measured value when taken out as part of the cross validation. The limitation behind the low accuracy of interpolation of the distribution maps can stem from number of sampling sites and the even distribution of sampling points in the study site since inverse distance weighting method, in essence, relies on the weighted average of neighboring values, assigning larger weights to closer points (Ligas et al, 2022).…”