A Framework Using Open-Source Software for Land Use Prediction and Climate Data Time Series Analysis in a Protected Area of Portugal: Alvão Natural Park

Abstract

Changes in land use and land cover (LULC) in protected areas can lead to an ecological imbalance in these territories. Temporal monitoring and predictive modeling are valuable tools for making decisions about conserving these areas and planning actions to reduce the pressure caused by activities such as agriculture. This study accordingly developed an LULC analysis framework based on open-source software (QGIS and R language) and predictive methodology using artificial neural networks in the Alvão Natural Park (PNA), a protected area in northern Portugal. The results show that in 2041, Agriculture and Open Space/Non-vegetation classes will evidence the greatest decrease, while Forest and Bushes will have expanded the most. Spatially, the areas to the west and northeast of the protected area will experience the most significant changes. The relationship of land use classes with data from the climate model HadGEM3-GC31-LL (CMIP6) utilizing scenarios RCP 4.5 and 8.5 demonstrates how through the period 2041–2060 there is a tendency for increased precipitation, which when combined with the dynamics of a retraction in classes such as agriculture, favors the advancement of natural classes such as bushes and forest; however, the subsequent climate data period (2061–2080) projects a decrease in precipitation volumes and an increase in the minimum and maximum temperatures, defining a new pattern with an extension of the period of drought and precipitation being concentrated in a short period of the year, which may result in a greater recurrence of extreme events, such as prolonged droughts that result in water shortages and fires.