Abstract. Mountain permafrost and rock glaciers in the dry Andes are of growing interest due to the increase in mining industry and infrastructure development in this remote area. Empirical models of mountain permafrost distribution based on rock glacier activity status and temperature data have been established as a tool for regional-scale assessments of its distribution; this kind of model approach has never been applied for a large portion of the Andes. In the present study, this methodology is applied to map permafrost favourability throughout the semi-arid Andes of central Chile (29–32° S), excluding areas of exposed bedrock. After spatially modelling of the mean annual air temperature distribution from scarce temperature records (116 station years) using a linear mixed-effects model, a generalized additive model was built to model the activity status of 3524 rock glaciers. A permafrost favourability index (PFI) was obtained by adjusting model predictions for conceptual differences between permafrost and rock glacier distribution. The results indicate that the model has an acceptable performance (median AUROC: 0.76). Conditions highly favourable to permafrost presence (PFI  ≥  0.75) are predicted for 1051 km2 of mountain terrain, or 2.7 % of the total area of the watersheds studied. Favourable conditions are expected to occur in 2636 km2, or 6.8 % of the area. Substantial portions of the Elqui and Huasco watersheds are considered to be favourable for permafrost presence (11.8 % each), while in the Limarí and Choapa watersheds permafrost is expected to be mostly limited to specific sub-watersheds. In the future, local ground-truth observations will be required to confirm permafrost presence in favourable areas and to monitor permafrost evolution under the influence of climate change.

El interés científi co y político en los glaciares rocosos andinos ha aumentado fuertemente en los últimos años debido al creciente reconocimiento de su importancia hidrológica y las recientes intervenciones mineras en ellos. Se presenta un resumen de los conocimientos científi cos y el marco político-legal de las intervenciones pasadas y actuales en glaciares rocosos en Chile y, en forma preliminar, en Argentina y Perú. En estos países, 26 faenas mineras han afectado glaciares rocosos o podrían posiblemente intervenirlos en el futuro. Hasta la fecha, el área afectada de glaciares rocosos en Chile es de aproximadamente 3,3 km 2 . Se recomienda ampliar el actual debate sobre intervenciones en glaciares rocosos, basado en la actualidad principalmente en argumentos hidrológicos, para tomar en cuenta peligros ambientales en el largo plazo -como la posible inestabilidad de glaciares rocosos con depósitos mineros-como también la amplia gama de servicios ambientales brindados por los glaciares rocosos.

Abstract. Mountain permafrost and rock glaciers in the dry Andes are of growing interest due to the increase in human activities in this remote area. Empirical models of mountain permafrost distribution based on the spatial analysis of intact and relict rock glaciers and mean annual air temperature (MAAT) have been established as a tool for regional-scale assessments of permafrost favorability across entire mountain ranges; however, this kind of model approach has never been applied for a large portion of the Andes. In the present study, this methodology is applied to map permafrost favorability throughout the semi-arid Andes of central Chile (29° S–32° S), excluding areas of exposed bedrock. After spatially modeling MAAT distribution from scarce temperature records (116 station-years) using a linear mixed-effects model (LMEM), a generalized additive model (GAM) was built to model the activity status of 3524 rock glaciers. A Permafrost Favorability Index (PFI) was obtained by adjusting model predictions for conceptual differences between permafrost and rock glacier distribution. The results indicate that model has an acceptable performance (median AUROC: 0.76). Conditions highly favorable to permafrost presence (PFI ≥ 0.75) are predicted for 1051 km2 of mountain terrain, or 2.7 % of the total area of the watersheds studied. Favorable conditions are expected to occur in 2636 km2, or 6.8 % of the area. Especially in the Elqui and Huasco watersheds in the northern half of the study area, where a substantial surface portion (11.8 % each) was considered to be favorable for permafrost presence, while predicted favorable areas in the southern Limarí and Choapa watersheds are mostly limited to specific sub-watersheds. In the future, local ground-truth observations will be required to confirm permafrost presence in favorable areas, and to monitor permafrost evolution under the influence of climate change.

Microcontrollers such as Arduino have been increasingly used by researchers to create and customise their own tools. In geography, microcontrollers are frequently used to design data loggers for monitoring purposes. We reviewed the use of Arduino in physical geography to unravel the opportunities and challenges of using off‐the‐shelf tools in research. We conducted a literature review, putting the retrieved information in perspective with our experimental work in mountainous and riverine landscapes in Chile and Peru. We show that the low cost and versatility of Do It Yourself (DIY) data loggers open research opportunities, extending the range of application of their expensive commercial counterparts. The possibility of connecting Arduino to a wide range of sensors, actuators, and wireless communication devices has helped to monitor rivers, glaciers, lakes, ice‐waves, caves, and landslides, improving the temporal and spatial resolution of data collected in critical environments. Low‐cost sensors have been extensively compared against expensive alternatives with good results, although they require thorough testing before field deployment due to the common existence of defective equipment. Building research equipment has several challenges. DIY data loggers might not be unconditionally accepted by environmental agencies, partially restricting their use to educational and research purposes. Failures in data loggers can be difficult to track, since they might be related to coding, electronic assemblage, or inadequate housing to withstand outdoor use. Yet, Arduino‐based data loggers have helped scientists around the world in different stages of their career, especially in scarcely funded research endeavours. Arduino has boosted creativity and resourcefulness, paving the way for innovative monitoring strategies in physical geography.

In the tropical Andes of Peru very little is known about the occurrence and extent of mountain permafrost. Only recently systematic studies have been carried out on the high elevation sites of the mountain ranges (cordilleras). In the framework of the first pioneering studies, and with the objective to improve the understanding of characteristics of mountain permafrost and rock glaciers, we analyze how mountain permafrost in the Peruvian Andes is correlated with the altitude of the 0°C isotherm (ZIA). Climate change has generated an increase in air temperature and in the ZIA in the past decades. These temperature changes could lead to impact the state of the mountain permafrost. In this research, we focus on two mountain regions: The Cordillera Central (CC) and the Cordillera Volcánica (CV), the first located in the central zone and the second in the south zone of Peru. The study used air temperature data from 20 weather stations (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016) to calculate the mean annual air temperature (MAAT), interpolated using a multiple linear regression model (MLRM) and digital elevation model (MERIT DEM). Occurrence and extent of 46 intact rock glaciers (IRG) and the global model of permafrost (Permafrost Zonation Index) were used to validate the results. The MAAT of CC has a minimum value around -4.1°C (R2 = 0.8) and a ZIA average of 5152 m a.s.l. None of the IRGs are located above the ZIA. The MAAT of CV has a minimum value around of -5.5°C (R2 = 0.8), a ZIA average of 4861 m a.s.l., and 60% of the IRGs are located above of the ZIA. The results show a greater variation of the position of the ZIA in CC in comparison to CV, which could indicate a possible degradation of mountain permafrost in these mountain ranges.