The main purpose of data mining in private and public sector institutions is to process and analyse data with the aim of generating reliable information for decision-making. Decision-making performance is determined by the availability of the relevant data and the user’s ability to adapt that data for analytical purposes. The popularity of spatial statistical tools is on the rise owing to the complexity of the analysed factors, their variation over time and their correlations with the spatial structure. Popular models should be applied in demographic analyses for the needs of the spatial planning process. The availability of high-resolution data and accurate analytical tools enhances the value of spatial analyses, and the described models can be universally applied to support the decision-making process. The aim of this study was to present the applicability of selected spatial statistical models for analysing demographic data in the planning process and to identify the main advantages of these models.

The possibilities of using the geospatial analysis methods for visualizing land monitoring data and modelling the spatial distribution of the main agrochemical soil indicators are discussed in the article. The research was conducted within the limits of land use of RUP “Uchkhoz BGSHA” (Republic of Belarus, Mogilev region, Goretsky district). The total area of the surveyed territory was 3187.0 hectares. The geospatial analysis of the spatial distribution of humus, mobile phosphorus, mobile potassium and pHKCl was carried out using the Geostatistical Analyst module of the ArcGIS software. Semivariograms were used as the main tool for studying the structure of the spatial distribution of agrochemical indicators. The exponential function was identified as the best variogram model, the type of the circle was standard, the type and the number of sectors was 4 with a displacement of 450, and the lag was 200 metres. The interpolation accuracy was determined from the mean error (ME), mean square error (RMSE) and standard error (RMSS). The universal kriging method was used to perform the forecast and visualize the spatial distribution of agrochemical indicators. The multivariate analysis was performed using the functionality of the Raster Calculator tool, Principal Component analysis and Maximum Likelihood Classification. The search and determination of areas of sites with the most optimal agrochemical indicators were carried out by the multifactor analysis in the GIS environment. Calculation of the area of each circuit within the limits of working parcels was carried out using the utility "Zone Statistics".