View Presentation [PDF]
Paola Peroni and Gareth Smith, Exprodat Consulting Ltd.
Geocientists are commonly faced with the problem of selecting the most appropriate interpolation algorithm when generating grids from point data. Often, the geoscientist will select the defaults provided in their E&P mapping tools without understanding the implications of their choices. GIS provides a rich set of tools for evaluating their data before making these decisions, and increasingly for generating the final interpolated data sets, especially for more regional analysis work, without having to switch to other E&P mapping applications. Choosing a suitable interpolation method for the type of phenomenon we are trying to model and for the specific distribution of our sampled population is not a matter of luck: we need to understand the spatial behavior of the phenomenon we are investigating. And we need to answer to some critical questions: Is the variable we are interpolating normally distributed? Are there directional components in our dataset? How much does proximity matter for the specific phenomenon we are modelling? How spatially dependent are values across the study area?
Exploratory Spatial Data Analysis(ESDA) is a group of techniques used to describe and visualize spatial distributions, to highlight patterns affecting the distribution of sampled values and to identify outliers or any non-typical values. In this paper we focus on the use of ESDA techniques available in ArcGIS Geostatistical Analyst and to guide the user in choosing a suitable interpolation method among the many available. Results of the ESDA are evaluated along with other considerations about the aim of the interpolation process itself as well as the distribution of sample locations and the type of phenomenon being studied. Examples used in this paper focus specifically on the type of datasets geoscientists may commonly encounter. A workflow aimed at guiding the user to choose among interpolation algorithms is also presented. Users can greatly benefit from the results of ESDA, particularly when these results are critically used to support the spatial modelling approach and to add value to the interpolation process itself, either within GIS or other E&P mapping applications.