A Data Warehouse for Spatial Soil Data Analysis and Mining: Application to the Maghreb Region

Soil health affects soil functions, food production, and climate change. Understanding our soil is important for our environment and food security. Therefore, we propose a data warehouse architecture for storing, processing, and visualizing soil data. In this study, we focused on the Maghreb region for the first time. This region is highly vulnerable to climate change and food insecurity, even though, as far as we know, it has not yet been considered. The proposed data warehouse architecture involves data mining and data science techniques to extract value from soil data and support decision-making at various levels. In our case, the warehoused data were analyzed using exploratory spatial data analysis tools to explore the spatial distribution, heterogeneity, and autocorrelation of soil properties. In our experiments, we presented the results of organic carbon (OC) analysis because of its importance to regulate climate change. The highest values of OC were situated in Morocco and Tunisia and we noticed a great distribution similarity in Algeria and Libya. One of the most important results of the autocorrelation analysis is the presence of positive spatial autocorrelation that leads to a perfect spatial clustering situation in the whole Maghreb region. This motivated us to a future work that aims to detect and interpret these clusters.