Analyzing land use trajectories. A clusterization proposal

This paper presents a methodological alternative to analyze land use trajectories. It takes as a case study and prototype an analysis of changes in land use at a general level in Argentina (the Pampas region, the Northwest, the Northeast, and Cuyo, excluding the Patagonia). The objective is to obtain a map with the highest possible resolution that allows to detect the following situations: 1) areas of recent agrarian frontier expansion; 2) areas of "consolidated" agriculture, and 3) areas of recent urban expansion. This paper focuses especially on the development of the applied methodology and a first attempt to validate it. Based on data from the European Space Agency (ESA) linked to the "Climate Change Initiative-Land" (ESA-CCI-LC) for the 1992-2020 period, a clustering of land use trajectories from distances is generated by means of editing distances (used for sequence analysis). There were 29 types of pixel-level trajectories obtained, validated by visual analysis, with an acceptable performance for detection of different trajectories.