Accurately estimating and forecasting evapotranspiration is one of the most important tasks to strengthen water resource management, especially in desert areas such as La Yarada, Tacna, Peru, a region located at the head of the Atacama Desert. In this study, we used temperature, humidity, wind speed, air pressure, and solar radiation from a local weather station to forecast potential evapotranspiration (ETo) using machine learning. The Feedforward Neural Network (Multi-Layered Perceptron) algorithm for prediction was used under two approaches: “direct” and “indirect”. In the first one, the ETo is predicted based on historical records, and the second one predicts the climate variables upon which the ETo calculation depends, for which the Penman-Monteith, Hargreaves-Samani, Ritchie, and Turc equations were used. The results were evaluated using statistical criteria to calculate errors, showing remarkable precision, predicting up to 300 days of ETo. Comparing the performance of the approaches and the machine learning used, the results obtained indicate that, despite the similar performance of the two proposed approaches, the indirect approach provides better ETo forecasting capabilities for longer time intervals than the direct approach, whose values of the corresponding metrics are MAE = 0.033, MSE = 0.002, RMSE = 0.043 and RAE = 0.016.

Desertification is one of the main threats to high Andean ecosystems, particularly in arid and semi-arid regions subject to increasing climatic and anthropogenic pressures. This study evaluated the spatial-temporal dynamics of desertification in the province of Candarave (Tacna, Peru) by integrating the Remote Sensing-based Desertification Index (RSDI), constructed from a principal component analysis incorporating four biophysical indicators: vegetation greenness, surface moisture, soil grain size, and fraction of solar radiation reflected (albedo), derived from Landsat 5 and 8 satellite images processed in Google Earth Engine. Temporal trends were analyzed using the Mann–Kendall test, while system stability was evaluated using the coefficient of variation, allowing different degrees of stability and environmental degradation to be characterized during the period 2010–2025. The results show that moderate and severe desertification classes predominate in higher altitude areas, covering approximately 92% of the study area, and are characterized by insignificant to weakly significant negative trends associated with high to relatively high temporal volatility. In contrast, stable areas with no significant changes represent 5.3% of the territory, while restoration processes occupy a small proportion, close to 2.7%. The high variability observed in the high Andean sectors is mainly linked to the interaction between reduced water availability, climate variability, and extreme events, as well as anthropogenic pressures, particularly overgrazing and aquifer exploitation. This multitemporal analysis allows us to anticipate the evolution of desertification and highlights the need to strengthen conservation planning in order to reduce the degradation of strategic high Andean ecosystems in the Tacna region.

This work focuses on studying the relationship that existed between the use of the learning management system (LMS) and the academic performance of the students of the Jorge Basadre Grohmann National University of Tacna-Perú. For this, we use the data provided by the LMS (access virtual classroom) and the university's academic management system (grades). For that, we perform various classification machine learning algorithms to predict academic performance with two classes SATISFACTORY or POOR where Gradient Boosted Trees algorithm had the best accuracy 91.79%. However, with three classes, SATISFACTORY, REGULAR AND POOR, Random Forest algorithm had the best accuracy of 89.26%.

Polylepis forests are the highest-elevation woody ecosystems on Earth and are severely threatened by climate change and land use, yet habitat models that go beyond binary presence/absence remain scarce. The current and future habitat suitability (2021–2100; CMIP6, four to five GCMs × four SSPs) of two co-occurring Andean species, Polylepis rugulosa and Polylepis tarapacana, was modelled with MaxEnt and Random Forest calibrated across their full range (Peru, Bolivia, Chile and Argentina) and clipped to the Tacna region of southern Peru. Internal spatial cross-validation was complemented by two external validations: 228 field-mapped stand polygons and 35 inventory plots (883 trees), the latter used to test whether suitability also tracks measured stand structure. Both algorithms performed well (AUC $=0.78$–$0.94$ across internal and external validation). Once within-plot non-independence and residual spatial autocorrelation were accounted for, no structural attribute was significantly associated with suitability: a positive suitability–multi-stemming trend persisted in Polylepis rugulosa ($p=0.06$–$0.09$) but not in Polylepis tarapacana, so this coupling is reported as a preliminary, hypothesis-generating result rather than as a validated predictive relationship. Within Tacna, suitable habitat above the maxSSS threshold is projected to contract by $17.4\%$ for Polylepis rugulosa and by $100\%$ for Polylepis tarapacana under SSP585 by 2081–2100 (MaxEnt ensemble; the near-total loss is unanimous across the contributing GCMs), revealing divergent climatic vulnerabilities. Reading suitability as a map of habitat quality rather than of habitat presence therefore remains a working hypothesis that larger plot networks and demographic data are needed to test.