Machine Learning to classify the indiscriminate use of land for agricultural purposes and its relationship with climate change, Atacama Desert headwaters

Maquera-Callo E.; Pino-Vargas E.; Choque G.; Huayna G.; Fernández-Cutire O.; Ramos-Fernández L. · Idesia (2023) · DOI: 10.4067/s0718-34292023000400101

Resumen

Climate change has been generating uncertainty in water availability and, in turn, problems in food security worldwide. The objective of this work is to determine the spatial and temporal evolution of land use, to adequately manage the irrigation that comes from groundwater from an overexploited aquifer. Freely available satellite images for the period from 1985 to 2021 and Machine Learning techniques were used for the classification and calculation of biomass indices in the Tacna region in southern Peru. The agricultural area grew by 373.2% in the period 1985 to 2021, being the period of greatest increase between the years 2015 to 2021. The VCI values increase in the period 1985 to 2021 (from 0.1 to 0.98), which represents that there are no drought conditions. The TCI index presents values that do not show a defined pattern and drought is observed in the years 2000, 2005 and 2010, and without drought in the years 1985, 1990, 1995, 2015 and 2021. The VHI index indicates that the years 1990, 2000, 2005 and 2010 presented drought (mild to moderate), while the years 1985, 1995, 2015 and 2021 exhibit conditions without drought. These anomalies are associated with the variability and climate change identified in the study region.

Instituciones: Universidad Nacional Agraria La Molina; Universidad Nacional Jorge Basadre Grohmann