LDT4QGIS: An open-source tool to enhance landscape analysis

Abstract

Despite the variety of landscape software available, new tools keep being developed in order to improve the existing analytical capabilities. Scripts, packages, or standalone software are built to enhance performance, add features or fulfil specific needs. This paper introduces LDT4QGIS, an open-source tool built exclusively to operationalize the Landscape Dynamics Typology method. LDT4QGIS is a Python-based tool developed to assess the landscape composition and configuration and to classify different land cover / land-use changes using QGIS. By combining the metrics ‘area’ and ‘number of patches’, it is possible to define different patterns of landscape changes (types of dynamics). The tool, which uses vectorial binary landscapes, allows squares or irregular polygons to be used as analytical units. The end goal of the procedure is to assign a type of dynamic to each analytical unit. To do this, metrics and their variations are calculated and then combined through a series of queries and selections that operationalize the type of dynamic assignment. LDT4QGIS has additional function- alities that are computed at the class level; they are related to the calculation and spatially explicit representation of gained and lost areas, as well as perforations in the land cover under study. Depending on what functionalities are being used, two or three analytical moments are required as inputs. The outputs produced by LDT4QGIS can be used in landscape assessment procedures and therefore can be helpful to decision-makers, allowing them to better understand and anticipate the consequences of policies related to the use of the land. LDT4QGIS will be updated regularly and improved whenever we develop a new functionality, but the real strength of LDT4QGIS is that it is free and open-source, and it can be adapted by each user according to their own needs. This paper describes LDT4QGIS and includes three illustrative practical examples to demonstrate the variety of scripts and outputs. The first case study concerns the dynamics of grasslands in Slovenia, and it uses two analytical moments and squares as analytical units. The second example focuses on olive groves in Portugal; it uses administrative boundaries as analytical units and three analytical moments. The last example concerns forests in Germany and showcases additional features like the detection of gained areas, lost areas, and the associated dynamics.