The loss of glacier resilience due to climate change throughout the Cordillera Blanca, Peru between 1984 and 2023

The loss of glacier resilience due to climate change throughout the Cordillera Blanca, Peru between 1984 and 2023 Item Type info:eu-repo/semantics/article Authors Giraldo Malca, Ulises Francisco; Yauri Solano, Lilian Netsy; Choroco Carranza, Sofia Valeria; Camacho Alvarez, Daniela Geraldine; Quispe Quispe, Fernanda Cryztal; Chávez García, Johann Alexis; Mark, Bryan G. DOI 10.1016/j.qsa.2025.100286 Publisher Elsevier Ltd Journal Quaternary Science Advances Rights info:eu-repo/semantics/openAccess; Attribution 4.0 International Download date 03/11/2025 16:12:54 Item License http://creativecommons.org/licenses/by/4.0/ Link to Item http://hdl.handle.net/10757/686713 The loss of glacier resilience due to climate change throughout the Cordillera Blanca, Peru between 1984 and 2023 Ulises Francisco Giraldo Malca a,* , Lilian Netsy Yauri Solano a , Sofia Valeria Choroco Carranza a , Daniela Geraldine Camacho Alvarez a , Fernanda Cryztal Quispe Quispe a , Johann Alexis Ch´ avez García a , Bryan G. Mark b a Universidad Peruana de Ciencias Aplicadas, Prolongaci´ on Primavera 2390, Santiago de Surco, 15023, Peru b Department of Geography and Byrd Polar Research Center, The Ohio State University, Columbus, OH, USA ARTICLE INFO Keywords: NDWI Climate patterns Santa river basin Glaciated massif Huascaran national park ABSTRACT The loss of mountain glaciers has accelerated in recent decades, linked to global warming, which in Peru alone has caused the loss of more than half of its glaciated area in fifty years. The Cordillera Blanca is the highest and most extensively glacierized tropical mountain range in the world, and glacier-fed streams provide water for hundreds of thousands of people living downstream. Previous inventories and glacier-specific mass balance studies have documented persistent and sustained mass loss. Yet the range-wide resilience of glaciers – the capacity to accumulate annual snowfall to offset area loss – remains an unquantified variable that is important to understand the evolution and climate response of glaciers over time and better project future mass changes for the coming decades. Therefore, we analyze the relationship between the annually clean glacier area and snow cover fluctuations and climate variability throughout the entire glacierized Cordillera Blanca between 1984 and 2023. To this end, we used multispectral Landsat imagery to identify clean glaciers and distinguish accumulation areas by calculating the Normalized Water Differential Index. The results show a 44 % reduction in glacier area, reflected in a decrease from the pre-2013 annual average of 54,469 ha to 42,700 ha in subsequent years. Our results suggest glaciers have passed a significant mass balance threshold, such that since 2012, glaciers have lost their ability to regain mass. We also document a strong inverse correlation of glacier area with the increase in global mean temperature, with the greatest loss occurring during the lasts strong El Ni˜ no-Southern Oscillation (ENSO) phases. We conclude that glaciers have become less resilient over the past decade, that the deglaciation of the Cordillera Blanca is primarily driven by increasing average temperatures and that the glaciers with the greatest retreat are those with perimeters proportionally more exposed to other types of surfaces (i.e., bedrock or lakes),. 1. Introduction Glaciers are essential natural components of most ecosystems on Earth, covering 10 % of the Earth’s surface and storing around 70 % of the planet’s freshwater (Synnøve et al., 2018). However, glaciers are not immutable ice masses in space and time, but are subject to global and regional climate variations (Shoobridge, 2005). Glaciers exist in highest areas of mountains as perennial mass balances between annual accumulation of snowfall and loss through melt, sublimation and calving (Martin et al., 2020). Among the myriad Earth systems affected by climate change, mountain glaciers (Lozano-Povis et al., 2021) are the most sensitive to these processes (INAIGEM, 2017; Liu et al., 2022). Recent studies show that global glacier net mass loss has increased rapidly in recent years in response to global warming, rising from an annual loss of 227 gigatons of ice per year between 2000 and 2004 to 292 gigatons per year between 2015 and 2018 (Hugonnet et al., 2021). In the 50 years between 1961 and 2001, the planet lost more than 9625 gigatons of glacier ice (Zemp et al., 2019), leading in some cases to their complete disappearance (Inocente et al., 2020; Synnøve et al., 2018). At the same time, air temperature is increasing at a rate of 1.7 ◦C per decade in areas above 5000 m a.s.l. (Aguilar-Lome et al., 2019). These trends * Corresponding author. E-mail addresses: [email protected] (U.F. Giraldo Malca), [email protected] (L.N. Yauri Solano), [email protected] (S.V. Choroco Carranza), [email protected] (D.G. Camacho Alvarez), [email protected] (F.C. Quispe Quispe), [email protected] (J.A. Ch´ avez García), mark.9@osu. edu (B.G. Mark). Contents lists available at ScienceDirect Quaternary Science Advances journal homepage: www.sciencedirect.com/journal/quaternary-science-advances https://doi.org/10.1016/j.qsa.2025.100286 Received 25 March 2025; Received in revised form 21 June 2025; Accepted 25 June 2025 Quaternary Science Advances 19 (2025) 100286 Available online 26 June 2025 2666-0334/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). are likely to continue as the World Meteorological Organization has recorded the warmest global mean temperatures on record since 2015, with an increase of 1.2 ◦C above pre-industrial levels by 2020 (OMM, 2021) that is estimated to exceed 1.5 ◦C from 2024 onwards (Nature Climate Change, 2023). This is attributed to the fact that anthropogenic activities have increased greenhouse gas emissions in recent decades (Hugonnet et al., 2021; Inocente et al., 2020), whose increased concentration in the atmosphere is driving the accelerated warming observed, especially in low-latitude areas (Thompson et al., 2021), driving massive changes in glacier loss. The retreat of tropical Andean glaciers is one of the fastest in the world (Synnøve et al., 2018). The Andes in South America contain the largest proportion of tropical glacier area on Earth (Vuille et al., 2008), most of which (70 %) is in Peru (Chevallier et al., 2011; INAIGEM, 2017), And this rapid glacial retreat is one of the clearest manifestations of global climate change that has also affected the Andean water cycle and the human systems that depend on it (Mark et al., 2017). Various inventories using aerial photographs and satellite imagery have documented persistent and accelerating decrease in glacier area throughout Peru, with about 43 % of glacier extent having been lost between the first glacier inventory in 1970 and 2010 (Seehaus et al., 2019), and 51 % lost over the last fifty years (ANA, 2019; García-Tadeo et al., 2023). Complete glacier extinction has been recorded in two of Peru’s ranges, the Cordillera Volc´ anica and Cordillera El Barroso (Wood et al., 2021). The Cordillera Blanca, located in the Peruvian department of Ancash, is the most extensively glacierized tropical mountain range in the world (Gordon et al., 2015), containing approximately 25 % of Earth’s tropical glaciers (Glasser et al., 2009). Since the 1930s, it has lost more than 30 % of its glacier area (Schauwecker et al., 2014). In 1970, glaciers of the Cordillera Blanca covered an area of more than 723 km 2 , representing 35 % of Peru’s glacier area and 40 % of its total volume (Mark et al., 2017). By 1990, it had fallen to 620 km 2 , and it is projected to drop below 600 km 2 by the end of the decade (Georges, 2004), and it is estimated that the retreat of its glaciers could reach up to 75 % of its extent by 2080 (Juen et al., 2007). Glacier-specific measurements have quantified glacier area loss: Huascar´ an decreased by 6.24 km 2 (approximately 8.6 %) between 1987 and 2008 (Gilberto Medina et al., 2022); Yanamarey was reduced by approximately 2.1 km 2 (60 %) between 1975 and 2014, fragmenting into three blocks and registering an annual average of 0.02 km 2 between 2004 and 2013 (L´ opez-Moreno et al., 2017); and Queshque lost 1.6 km 2 (approximately 36 %) between 1962 and 1999, losing approximately 85 % of its volume in the southwest-facing sector (Mark and Seltzer, 2005). Consequently, this background shows if the current rate of retreat continues, all but the glacial ice on the highest peaks in the Cordillera Blanca will be lost after 2100 (Moulton and Carey, 2023). Many previous studies have examined the controls of mass balance of these glaciers (Kaser et al., 2003; Hastenrath and Ames, 1995; 1996; Kaser et al., 2003) and explored their hydroclimatic sensitivity (Kaser et al., 2003; Vuille et al., 2008) and hydrological impact of loss (Mark and Seltzer, 2005; Baraer et al., 2012). Climate change induced glacier loss also has broader implications for a diversity of livelihoods in the Cordillera Blanca region (Mark et al., 2017; Bury et al., 2011). Hydrological concerns of glacier loss are acute with the Santa River basin because it is the main recipient of glacier meltwater, and currently supplies water to around 366,000 inhabitants and generates 10 % of the country’s hydroelectric energy (Dextre et al., 2022). Biological impacts, disease and ecological services have been implicated (Marco Zapata-Luyo, 2010) (Vuille et al., 2008). Huascar´ an National Park (HNP) is a Natural Protected Area (NPA) that conserves the varied ecosystems of the Cordillera Blanca, archaeological remains and landscapes, and is an important source of freshwater for the departments of Ancash and La Libertad (INRENA, 2003). It also contains the majority of the Cordillera Blanca glaciers, extensive high altitude wetlands and periglacial lakes, affected by recent climate change (Young et al., 2023). It is also a major tourist attraction contributing to the regional economy, receiving up to 405,588 tourists in 2019 (MINCETUR, 2021). Hazardous glacial lake outburst floods (GLOFs) are also associated with glacier melt, often with devastating impacts such as in 1941, when lake Palcacocha overflowed, causing a debris flow that destroyed a third of the city of Huaraz (Otiniano Zavala et al., 2023) and killed at least 1800 people (Emmer, 2023). While the Cordillera Blanca has thus featured much prior research on glacier changes, the projects have been limited in temporal or spatial extent or featured different methodologies. Given the antecedent results suggesting accelerating mass changes and impacts, we are motivated here to examine an up-to-date range-wide survey of glacier resilience, specifically testing the year-to-year evolution of glacier area and accumulation in the context of ongoing climate change. To this end, the objective of this research is to analyze the relationship between variations in clean glacier area and climate variability, including the relative impact of ENSO, on the glacier mass of the Cordillera Blanca between 1984 and 2023, the longest interval possible given available satellite imagery. We deploy a range-wide survey the interannual evolutionary dynamics of snow cover and the glaciers of the entire Cordillera Blanca over the last four decades up to the present to understand the impact of conditioning factors on their temporal and spatial variability. This comprehensive perspective on annual glacier resilience involves using consistent annual measures of snow cover and glacier extent will allow examination of climate controls and sensitivities, while allowing spatial comparison between different sectors and specific individual mountains understand the differences related to their extent, altitude or location. Such a comprehensive dataset could be used by institutions responsible for projecting future hydroclimatic scenarios in the framework of management of water resources for the populations dependent on these glaciers. 2. Study area The glacierized area of the Cordillera Blanca is contained within the bounds of the HNP, located between the coordinates 8◦46’11"S - 10◦3’58"S and 77◦4’58"W - 77◦49’4"W, in the department of Ancash, Peru. HNP contains 340 thousand hectares (INRENA, 2003), ranging between 2419 and 6768 m a.s.l. The NHP landscape features a rugged physiography with high relief, i.e., large altitudinal variations occur over small distances (Mergili et al., 2015). There are deeply embedded valleys formed by tectonic uplift, glacial scouring and fluvial erosion (Polk et al., 2019). Current glaciers occur normally above 5000 m a.s.l. (L´ opez Moreno et al., 2020), contained within the Cordillera region that comprises 103,696 ha (30.5 %) (Britto, 2017). This area contains the highest peaks of the range (Pulgar Vidal, 2014), whose slopes lack soil or vegetation above 4800 m a.s.l, making them prone to avalanches, landslides and rockfalls (Ingemmet, 1989). The glaciers of the HNP are distributed along a length of approximately 180 km, from the Tuco mountain in the south to the vicinity of the Champar´ a mountain in the north. The HNP has 826 periglacial lakes, with a total water surface area of 53.7 km 2 (Curo Rosales et al., 2023), and is drained by three hydrographic basins: the Santa, Mara˜ n´ on and Pativilca rivers (Vilímek et al., 2016). In the case of HNP, the glaciers are one of its main tourist attractions, where the number of visitors has increased significantly in the last 30 years, exceeding 1.3 million between 2009 and 2018, benefiting the Abbreviation NDWI Normalized difference water index HNP Huascaran national park WS Weather Station ENSO El Ni˜ no – Oscilaci´ on del Sur NPA Natural protected area U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 2 population that depends directly and indirectly on this activity (Pozada-Rengifo et al., 2023). In light of this, the glacial retreat affects tourists and tour operators by altering the landscape and causing accessibility problems due to fragmentation and sliding of ice, snow and rock, and instability of moraines, which pose safety risks to tourists (WANG & ZHOU, 2019). 3. Methods Our research involves a quantitative approach that aims to explore and establish relationships between climatological and landscape variables by using computational techniques and statistical analysis. In addition, we have carried out an exhaustive data collection within the constraints of our faculty-student collaboration at our Peruvian academic institution. We have accessed prior research results using scientific articles from journals indexed in databases such as Web of Science, Science Direct, SCOPUS, Scielo, among others. We have also accessed available data from online archives and technical reports from competent scientific institutions in Peru such as the National Meteorological and Hydrological Service of Peru (SENAMHI), the Institute of the Sea of Peru (IMARPE), the Geophysical Institute of Peru (IGP), the National Institute for Research on Glaciers and Mountain Ecosystems of Peru (INAIGEM), the National Water Authority (ANA) and the Ministry of Environment of Peru. 3.1. Determination of the meteorological and climatic conditions of the study area The analysis of the meteorological conditions in the Cordillera Blanca was carried out using historical data on precipitation and temperature from seven weather stations (Fig. 1), four available on the SENAMHI web platform, as well as data from three stations obtained from the research and open access dataset of Mateo et al. (2022) (Table I). The data were systematized and processed in Microsoft Excel and Infostat/L spreadsheets and pivot tables. This was done in accordance with the World Meteorological Organization’s (WMO, 2018, 2023) recommendations for normalization, cleaning erroneous and missing data, and estimating mean and total values. These methods were Fig. 1. Variation of glacier area by decades in the HNP. U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 3 employed to analyze the behavior of rainfall and temperature over the study period. In addition, the ERA5-Land daily mean dataset was used through the Google Earth Engine web platform to obtain mean air temperature data at 2 m above the surface from 1984 to 2020 for the northern, central and southern sectors to analyze the thermal variations of glaciers in different areas of the Cordillera Blanca. Finally, annual mean temperature variation data were obtained from the National Ocean and Atmospheric Administration (NOAA) of the United States of America for the period 1984–2023. 3.2. Calculation of glacier area using satellite images The clean glacier area in the HNP was determined using multispectral satellite imagery obtained from the Glovis (Global Visulization Viewer) web platform of the United States Geological Survey (USGS). Landsat 5 TM satellite images were downloaded for the years 1984–2001 and 2003–2011, Landsat 7 ETM for 2002 and 2012, and Landsat 8 OLI/TIRS for 2013–2023. The selection criteria were to extract one image per year, corresponding to those showing the smallest glacier surface for that period, preferably between the months of August and November (end of the dry season), due to the lower cloud cover surface in the study area and during this time, the absence of precipitation and increased temperature cause temporary snow cover to disappear, leaving only the ice masses that make up the glacier. The selected scenes were processed in the open-source software QGIS and the Normalized Difference Water Index (NDWI) was calculated as a method to identify areas of snow cover and glaciers. Using the Raster Calculator, the values of the near infrared and shortwave infrared bands were contrasted using the following formula. The resulting raster was reclassified based on its values to distinguish areas covered by glaciers (0.3–1) from other surface types (−1 to 0.3). We converted the reclassified raster to a vector using the "Polygonize” (raster to vector) tool to calculate the areas of each polygon, eliminate findings smaller than 0.5 ha, discard erroneous detections if they were outside the HNP, below 4500 m above sea level, and identifying and discarding any findings corresponding to a different surface type through a visual review of each Landsat scene. Finally, we validated the results using a representative sample of all processed pixels from the 76 scenes (from 38 years) used in our analysis. 4. Results After processing the satellite images, it was determined that the glacier area within the HNP (Fig. 3) varied between 70,017 ha (1984) and 38,912 ha (2023), showing a reduction of up to 44 % of the initial area over the period 1984–2023. There is evident interannual variability with periods of glacier contraction and expansion, but with a clear negative trend captured as decadal means (Fig. 2), accentuated in the last fifteen years (2008–2023) as seen in annual changes (Fig. 3). Our results also show that since 2012 glacier area has not exceeded 49,000 ha, whereas before that year glacier extent remained above this threshold, except in 1998, 2005, 2006, 2007 and 2010 (Fig. 3). It also shows that, prior to 2013, the average glacier extent was 54,469 ha, unlike the values recorded later, which had an average that did not exceed 42,700 ha. Results were validated with a sample of 1862 points representative of 415,375,045 pixels. Processing was done with 99 % confidence and a margin of error of 3 % in the distribution quadrant of the glacier area detected from 1984 to 2023 (8◦46’27. 7 S - 10◦1’49.3 S; 77◦7’55.2 W - 77◦46’10.9 W), resulting in a glacier area detection error of only 0.97 % (18 pixels). This error corresponds to the margins of the glacier area, which are combined and shaded in strongly hilly areas. Both the average and average maximum temperatures of the Recuay WS (Fig. 3) show a slightly negative trend, with the average temperature decreasing from a maximum value of 13.3 ◦C in 1992 and a minimum of 11.1 ◦C in 2022, and the average maximum temperature decreasing Table 1 WEATHER STATIONS USED IN RESEARCH. Weather station Latitude (◦) Longitude (◦) Altitude (m a.s.l) Variables Period Chavín a −9.586 −77.175 3140 Temperature, precipitation 2017–2023 Pachacoto a −9.852 −77.406 3733 Temperature, precipitation 2017–2023 Recuay a −9.729 −77.404 3431 Temperature, precipitation 1984–2023 Milpo a −9.951 −77.326 4400 Precipitation 1984–2014 LlanUp-4 LAS b −9.044 −77.598 3955 Temperature 2006–2015 Llanganuco PORTWX b −9.049 −77.590 4750 Temperature, precipitation 2006–2014 Cuchillacocha CUCHWX b −9.415 −77.355 4642 Temperature, precipitation 2013–2020 a SENAMHI web platform. b Open access dataset of Mateo et al. (2022). Fig. 2. Distribution of glacial areas in the HNP, recorded by decades within the analysis period. Fig. 3. Evolution of the glacier surface and temperatures of the Recuay WS. The labelled arrows indicate example years of relative increased glacier area coinciding with decreased temperatures. U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 4 from 22.6 ◦C in 1992 to 19.5 ◦C in 2012. However, the average minimum temperature shows an incremental and positive trend, fluctuating between 2.75 ◦C in 1989 and 5.32 ◦C in 2009. Notably, many specific years when the glacier area increases coincide with average minimum temperature decreases at WS, such as 1984, 1989, 1999 and 2000 (marked by arrows in Fig. 3). In the case of precipitation at the Recuay WS, a strong interannual variation is observed, with annual totals fluctuating between 455 mm and 1312 mm per year, and an overall negative trend (Fig. 4). Individual years of low precipitation align with observable decreases in glacier areas, such as 1991, 1997, 2003, 2005, 2016, 2020 and 2023 (Fig. 4). In contrast, the annual precipitation from Milpo WS depicts an increasing trend from 1984 (429 mm) to 2012 (2289 mm) (Fig. 4). It can be noted that in some years precipitation decreases coincide with the glacier area reduction (8 records) and precipitation increases coincide with the glacier area expansion (4 records); however, there are also several periods in which the trends are reversed (16 records). However, throughout the entire span of available observations, the evolutionary trends are inversed compared to the Recuay WS, showing that decreasing glacier area in the HNP is associated with increasing precipitation in the Milpo WS. As Veettil et al. (2017) As Veettil, Wang, Bremer, et al. (2017) pointed out, there has been a slight increase in precipitation in December, January, and February in the decade from 2000 to 2010, although the records are incomplete and interrupted. The HNP can be divided into three sectors (north, central and south, Fig. 1), grouping the 5 largest glaciated massifs in the northernmost (3) and central (2) zones, and the smaller glaciers in the southern zone. It was found that the largest glacier extent is in the central sector, which had 34,363 ha in 1984, but decreased to 21,827 ha by 2023. In second place is the northern sector, which had 27,399 ha of glacier in 1984, but only 14,872 ha in 2023. Meanwhile, the lowest recorded values are found in the southern zone, where the glacier coverage reached 8254 ha in 1984 and does not exceed 2214 ha in 2023 (Fig. 5). When analyzing the annual mean temperatures at 2 m altitude recorded in the ERA5-Land daily mean data set of each of the HNP sectors, consistent patterns are observed throughout the period 1984–2020 featuring: maxima in the years 1998, 2016 and 2020; strong decreases in 1984, 1993, 1999 and 2011; and an overall positive trend, with the lowest value in 1984 and the highest in 1998. Extrapolating using trend equations, it was estimated that the values reached in 2023 would be 7.74 ◦C in the northern zone, 6.92 ◦C in the central zone and 4.16 ◦C in the southern zone, which would represent a temperature change of more than 0.68 ◦C, 0.65 ◦C and 0.61 ◦C, respectively, compared to the base year (1984). On a smaller mountain-specific scale, glacier area changes are documented in the five main glaciated massifs called Alpamayo, Huandoy, Huascar´ an, Copa-Hualc´ an and Huatsan (Table II). Results show that those with the most preserved glacier extent throughout the period 1984–2023 are those located in the central zone, such as CopaHualc´ an (the second largest), with 70 % of its maximum record (1989) and a variation of 3756 ha, followed by Huascar´ an (the smallest), which maintains 68 % of its maximum extension, with a loss of 1,879 ha, and Huatsan (the largest), which maintains 63 % of its maximum surface, with a reduction of 7027 ha. Similarly, in the northern zone, glaciated massifs such as Huandoy (the fourth largest) and Alpamayo (the third largest), have retained about 55 % and 52 % of their maximum glacial extent, with a loss of 3918 and 5178 ha respectively. On the other hand, within the period analyzed, records of the Geophysical Institute of Peru (IGP, 2023) indicate that 16 ENSO events occurred, with moderate events in 1986–1987, 1991–1992, 1993, 2016–2017; strong events in 2015–2016 and 2023; and exceptional events in 1997–1998, where the values of the annual average Coastal Ni˜ no Index (ICEN) varied between −1.16 (2022) and 2.45 (1997) (IMARPE, 2023). It has also been observed that the years with the most pronounced glacier retreat correspond to the period after the years with the highest ICEN values, such as 1988 with a loss of 2794 ha (after 1987 which had an ICEN =1.4), 1998 with a loss of 3773 ha (after 1997 with an ICEN =2.45) and 2016 with a loss of 6066 ha (after 2015 with an ICEN =1.47). Finally, in 2023 the second highest ICEN (1.95) coincides with a strong glacier retreat (3311 ha), with only 7 months of warm-up. In order to validate the trends found when processing the data from the ERA 5 sensor, it was found that, unlike the Llanganuco_PORTWX WS, which shows a slight tendency to decrease its temperatures between 2006 and 2012, the temperatures of the LlanUp-4 WS show an increasing trend between 2006 and 2019, reaching 0.56 ◦C, as do the data of the Cuchillacocha_CUCHWX WS, with an increase of 0.19 ◦C between 2013 and 2020. When their values are compared with the annual mean temperature records at 2 m above the surface of ERA 5, they show a strong correlation (R 2 >0.7), which decreases slightly in the case of the LlanUp-4 WS (R 2 =0.59). Finally, in the case of precipitation, the two WSs for which these data are available (Llanganuco_PORTWX and Cuchillacocha_CUCHWX) show a slight downward trend, with Fig. 4. Evolution of the glacier area and annual precipitation of the Milpo WS. Fig. 5. Variation of the glacier area and average annual temperatures at 2 m high recorded by the ERA 5 reanalysis in the northern, central and southern sections of the HNP. Table 2 VARIATION OF THE SURFACE OF THE GLACIATED MASSIF. Glaciated massif Maximum area Minimum area Lost glacier area Hectares Year Hectares Year Hectares % Alpamayo 10737.0 1984 5558.9 2020 5178.1 48 % Huandoy 8797.9 1984 4879.7 2023 3918.2 45 % Huascar´ an 5953.5 1984 4074.12 2023 1879.4 32 % Copa-Hualc´ an 12636.4 1989 8880.3 2021 3756.1 30 % Huatsan 18993.4 1984 11966.6 2023 7026.8 37 % U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 5 values varying between 456 and 1501 mm per year. 5. Discussion As explained by other researchers, the current availability of multispectral satellite images and other data from remote sensors allows us to monitor glaciers over the past few decades (Veettil et al., 2017). We took advantage of this opportunity to conduct a longitudinal analysis spanning nearly 40 years, demonstrating the interannual variation of exposed glaciers and their relationship with the primary atmospheric factors impacting the region. It is important to note that the result obtained for 1984, which corresponds to the largest glacier surface identified in the entire analysis period (with more than 70,000 ha), may present a certain degree of bias, since it was estimated from the single satellite image available for that year that was taken in the month of December the beginning of the rainy season in the Peruvian Andes and part of the accumulation period (Veettil et al., 2017). Hence, this could represent an overestimate of total glacier coverage given potential misclassification of snowy areas without glacier mass to be counted as part of the glaciers for that period. An observation similar to the finding of Tang et al. (2013) of increasing minimum temperatures in mountainous areas coincident with glacier retreat, is the incremental trend shown by the annual average of minimum temperatures in the Recuay WS. Strikingly, this contrasts with the decreasing maximum temperature trend and suggests that it is one of the primary drivers of the glacier loss. This is because, unlike the maximum temperature, which depends to a greater extent on the amount of sunshine over the area on a given day, the minimum temperatures more closely align with the air mass characteristics at any time of the year. Therefore, the increase in the minimum temperature would be a manifestation of the effect of global warming on the climatic conditions of a locality, an effect that in this case is not reflected in the maximum and average temperatures. Such associated decreases in daily ranges of temperatures have been shown to be a feature of global warming on a global scale (e.g., Thorne et al., 2016). Although the inverse relationship between precipitation in the Milpo WS and the glacier zone would imply that the greater the amount of rainfall, the faster the glacier retreat, on the contrary, precipitation in the Recuay WS showed a direct relationship with the glacier zone. When comparing the two variables in both cases, a low correlation was found between them (R 2 =0.11 p-value<0.0001 and R 2 =0.09 p-value<0.0001 respectively). This would indicate that the amount of precipitation that has occurred in the area has not defined the increase or decrease of the glacial mass, a result that does not agree with that proposed by Calizaya Llatasi and Mejía Marcacuzco (2018), when they state that there is a strong correlation between precipitation and the glacier surface area of Artesonraju. This difference would be due to the fact that their analysis corresponds to a shorter period (2005–2009) that had stable atmospheric conditions, free of intense ENSO events, and was carried out looking at the monthly glacier and snow cover variation. These limitations would generate a stronger relationship between precipitation and the area with snow cover, different from the interannual dynamics of the glacier, which is more subject to climatic temperature variations. However, the association of precipitation with higher temperatures in recent years and during El Ni˜ no-Southern Oscillation (ENSO) events implies a higher percentage of liquid precipitation rather than snow (Veettil et al., 2017). This would accelerate the melting of ice masses. In relation to the loss of glacier area by sector, it can be seen that the smaller, lower elevation glacierized mountains to the south of the HNP have a greater proportion of lost area (equivalent to 73 % of their original extent) than the values shown within the central and northern sectors (36 % and 46 %, respectively). This is consistent with what was reported by Viera Casta˜ neda et al. (2023), in that the extent and volume of the ice mass is inversely proportional to its degree of influence by external temperature changes. Therefore, smaller and lower elevation glaciers have a greater proportion of mass exposed to the environment, making them visibly more affected by surface temperature increases (Rabatel et al., 2013). Additional observations of radiative feedback have been noted as likely causes of accelerating glacier loss in the HNP. When analyzing the case of Nevado Huascar´ an, the highest tropical mountain on Earth, Thompson et al. (2021) report that the retreat of glaciers exposes rocky surfaces. Exposed rocks peripheral to the glaciers above Cuchillacocha above Huaraz in the HNP have been observed with infrared sensors and shown to have a greater capacity for energy absorption and emission of longwave flux (Aubry-Wake et al., 2018). This points, similar to the finding reported by (Hinzmann et al., 2024) in the graces of East Africa, shows an additional positive feedback contributing to the warming of the mountain range and explaining the acceleration of the melting of its glaciers, since with the increase in exposed rock surface, the absorption of solar energy and therefore the irradiation of heat in the same snow area increases. This can result in a temperature difference of 11.7 ◦C between the exposed rocky debris surface and the proximal glacier ice during the peak radiation period at midday (Aubry-Wake et al., 2015). In particular, this effect would be more pronounced in smaller glaciers, such as those located in the southern sector of the HNP, as reported by Viera Casta˜ neda et al. (2023), since the increase in temperature generated by the rock surfaces surrounding them will affect a greater proportion of their glacier mass, as opposed to more extensive snowfalls, which will have a smaller proportion of their mass exposed to this type of effect. Looking at the variation in glacier area in the five biggest glaciated massifs of the HNP, it is striking that the glacierized peaks with the highest percentage of ice loss are those located north of the Cordillera Blanca, coinciding with the record of the highest surface temperatures according to the ERA 5 data. However, the Huascar´ an massif, which is the smallest in surface area and located in the northern area (although closer to the central area), features the most preserved glacier area, retaining more ice than Huantsan, which is the largest and most southern of the blocks. Therefore, when the area of each block was compared with the perimeter that had the greatest extension in the year, it was found that the Alpamayo and Huandoy blocks had a greater distance to the perimeter per hectare (48 and 47 m respectively), followed by Huantsan (43 m), the Copa-Hualc´ an block (40 m) and finally Huascar´ an (28 m). Therefore, this relationship would be an indicator of the degree of exposure of its glacier mass, which would explain the results of a smaller retreat of the ice sheet in the Huascar´ an and CopaHualc´ an blocks, since it is less exposed, as well as the greater impact on the glaciated massifs located further north, since a greater proportion of its mass is exposed to the thermal conditions of its surroundings. Another reason that may explain this difference is that, being the highest mountain in the Cordillera Blanca and its physiography, the Huascar´ an glacier has a proportionally larger area in the accumulation zone and a smaller extension in the ablation zone in relation to the other snowcapped mountains, so it would be visibly less affected by glacial retreat, unlike the lower and smaller glaciers, which, as Rabatel et al. (2013) points out, do not have a permanent accumulation zone and would be the first to disappear. After analyzing the variables studied, the strong inverse correlation between the glacier area within the HNP and the variations in global average temperature reported by NASA stands out (r = − 0.864; p-value <0.0001), which would mean that one of the main factors driving the retreat of the glaciers in the Cordillera Blanca is the increase in local temperatures caused by global warming (Fig. 6). Likewise, it was found that in this case, global warming is better reflected in the aggregate temperatures of the mountain range (ERA 5 records), where an increase in temperature is observed, as opposed to the local conditions at the ground level weather stations, such as the valley (Recuay WS), where local factors (e.g., cloudiness) prevail. As shown, the maximum temperature at Recuay has decreased slightly (non-significantly), while minimum temperatures capture stronger upward trends reflective of U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 6 global warming. Similarly, the ERA 5 temperatures show a warming of the atmosphere over the glaciers of the Cordillera Blanca (Fig. 7), increasing between 0.61 ◦C (southern sector) and 0.68 ◦C (northern sector) from 1984 to 2023. Furthermore, it was observed that these records have a stronger inverse correlation with the variation of the glacier area (0.453 <R 2 <0.519; p-value <0.0001), compared to the correlations obtained with the Recuay WS data (0.003 <R 2 <0.2), with a greater coincidence in years where there is an increase in temperature and a decrease in glacier area, as well as in years where there is a decrease in temperature and an increase in glacier area. This difference could be due to the fact that the Recuay WS, although relatively close to the glaciers of the Cordillera Blanca (16 km), is located at a different altitude (Quechua region), at the bottom of the Santa River valley. Therefore, its equipment measures parameters of local air masses that are different from those that determine the meteorological conditions at the top of the snowcapped mountains. On the other hand, reanalysis data (ERA5) aggregates various measures including satellite sources, so their variations will have a greater correlation with the effects they have on the ice masses. Although the influence of ENSO on tropical glacier variation is controversial (Veettil et al., 2017), our results support this argument. Despite the low correlation obtained when comparing the annual average of ICEN values and the glacier area (R 2 =0.022), greater distinctions are apparent when looking at how the retreat of the glaciers occurred over decades (Fig. 8). Iit was found that the greatest decadal differences occur in the values of the average, minimum and median area of the period 1991–2000 in relation to the period 1984–1990, followed by the difference between the decade 2011–2020 with the decade 2001–2010, which would be due to the fact that in these periods the event occurred extraordinary ENSO of 1997–1998 (ICENmax =4. 08) and the moderate ENSO of 2015–2016 (ICENmax =2.2), with a high correlation between the maximum values of the annual mean ICEN and the difference in the mean glacier area for each decade (R 2 =0.873; p-value =0.066). Likewise, it can be observed that the evolution of the glacier area in the HNP and the variation of temperatures within the analysis period correspond to what was proposed by Díaz Aguilar et al. (2017) in their evaluation of the glaciers of the Carabaya mountain range in Puno, Peru. They identified three distinct periods: (1) the last two decades of the 20th century with an increase in temperature due to the strong manifestations of ENSO, causing a large loss of glacier mass; (2) the period between 2000 and 2015, where no clear warming is observed due to the absence of ENSO and lack of significant glacier impact; and (3) the period after 2015, featuring a new period of frequent manifestations of ENSO that has warmed the atmosphere and caused the loss of glacier area. Similarly, our results show that the maximum percentage of losses occurred in the years 1985 and 1998, which is consistent with a greater presence of very strong ENSO events. This finding aligns with Georges (2004) statement that the reduction of glaciers intensified in the 1990s. Then, from 1999 to 2012, we observe a fluctuation in glaciers and temperatures without a clear trend of increase or decrease, with only the occurrence of two weak ENSO events, but from 2012 onwards we observe the opposite trends of a marked glacier retreat and increase in temperatures, which would be due to the manifestation of two strong and one moderate ENSO. One of the most striking results is the decrease in the average extent of the glaciers between the years 1984–2012 (54,469 ha) and the last decade from 2013 to 2023 (42,633 ha), with a reduction of more than 22 % of their extent. (11,836 ha), highlighting that in this last period when the glacier surface within the HNP does not exceed 50,000 ha, a figure that was exceeded 70 % of the years analyzed before 2013. For example, the maximum reduction in glacier area per decade shows a greater variation between the periods 2011–2020 and 2021–2023, with at least 13 thousand hectares of loss between its maximum values, although there have been years with lower temperatures (according to EM Recuay) due to an intense La Ni˜ na event in 2021–2022, where the glaciers of the Cordillera Blanca have not expanded as they did in previous decades. Together with the greater retreat caused by the 2015–2016 ENSO compared to the significantly more intense and warmer 1997–1998 event, which would mean that it is losing the recovery capacity of its glacier mass. The disappearance of tropical glaciers is a process that increases the vulnerability of the people living around them and those who depend on the meltwater for their intense agricultural activity in the Callej´ on de Huaylas during the dry season (Kaser et al., 2003), taking into Fig. 6. Correlation between glacier area and the increase in global average temperature according to NASA. Fig. 7. Correlation between the glacier surface and the average annual ERA5 temperatures at 2 m above the surface. Fig. 8. Variation of glacier area between decades. U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 7 consideration that the supply of fresh water provided by mountain glaciers to tropical regions with water stress can amount to more than 80 % (Vuille et al., 2008). As deglaciation continues, dry season flows will gradually decrease, a process that has long been predicted to lead to drastic water shortages if glacial headwaters disappear (Braun et al., 2000). Over the past decades assimilation of observations and modeling of glacier fed tributaries in the Cordillera Blanca have shown many having already passed “peak water” (Baraer et al., 2012). Others have shown that hydrological changes will likely complicate vulnerabilities and water resource management. As stream water availability decreases and demand increases, competition for the resource between economic sectors, political jurisdictions and users of high and low areas will likely increase, leading to conflict, food shortages and increasing poverty (Lynch, 2012). It is therefore necessary to develop the capacities of these populations and alternatives for sustainable resource management that will allow them to adapt to future scenarios and mitigate the effects of climate change, through projects such as water planting, payment for environmental services, reforestation, conservation of high Andean wetlands, among others. 6. Conclusions The Cordillera Blanca in the HNP had a reduction of 44 % of its initial extension in 1984, related to an increase in temperatures in its highest area, generated by global warming, as well as the occurrence of recent ENSO events generate more intense effects, since they have strong intensity that feature a high incidence in the loss of glacial area due to the increase in temperatures and liquid precipitation, which accelerate melting. Likewise, it was determined that glaciers with a smaller accumulation area and a greater proportion of their extent exposed to other types of surfaces, will have a greater loss of coverage, because rocky or liquid water surfaces absorb more solar energy and increase their temperatures, warming the nearby air and ice, which in the last decade would be causing the glaciers of the Cordillera Blanca to be losing their recovery capacity, even in years with better atmospheric conditions. CRediT authorship contribution statement Ulises Francisco Giraldo Malca: Writing – review & editing, Writing – original draft, Validation, Supervision, Software, Methodology, Investigation, Formal analysis, Conceptualization. Lilian Netsy Yauri Solano: Writing – original draft, Supervision, Software, Resources, Investigation, Formal analysis, Data curation. Sofia Valeria Choroco Carranza: Writing – review & editing, Validation, Supervision, Software, Investigation, Formal analysis, Data curation. Daniela Geraldine Camacho Alvarez: Writing – original draft, Software, Resources, Investigation, Data curation, Conceptualization. Fernanda Cryztal Quispe Quispe: Writing – original draft, Validation, Resources, Investigation, Data curation. Johann Alexis Ch´ avez García: Writing – original draft, Software, Resources, Data curation. Bryan G. Mark: Writing – review & editing, Visualization, Supervision, Formal analysis. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This research has been supported by the Faculty of Engineering of the Universidad Peruana de Ciencias Aplicadas – UPC, Perú. Data availability Data will be made available on request. References Aguilar-Lome, J., Espinoza-Villar, R., Espinoza, J.-C., Rojas-Acu˜ na, J., Willems, B.L., Leyva-Molina, W.-M., 2019. Elevation-dependent warming of land surface temperatures in the andes assessed using MODIS LST time series (2000–2017). Int. J. Appl. Earth Obs. Geoinf. 77, 119–128. https://doi.org/10.1016/j.jag.2018.12.013. ANA, 2019. Retroceso glaciar en la Cordillera Blanca 1948 - 2018. https://hdl.handle.ne t/20.500.12543/3770. Aubry-Wake, C., Z´ ephir, D., Baraer, M., McKENZIE, J.M., Mark, B.G., 2018. Importance of longwave emissions from adjacent terrain on patterns of tropical glacier melt and recession. J. Glaciol. 64 (243), 49–60. https://doi.org/10.1017/jog.2017.85. Aubry-Wake, C., Baraer, M., McKenzie, J.M., Mark, B.G., Wigmore, O., Hellstr¨ om, R.Å., Lautz, L., Somers, L., 2015. Measuring glacier surface temperatures with groundbased thermal infrared imaging. Geophys. Res. Lett. 42 (20), 8489–8497. https:// doi.org/10.1002/2015GL065321. Baraer, M., Mark, B.G., McKenzie, J.M., Condom, T., Bury, J., Huh, K.-I., Portocarrero, C., G´ omez, J., Rathay, S., 2012. Glacier recession and water resources in Peru’s Cordillera blanca. J. Glaciol. 58 (207), 134–150. https://doi.org/10.3189/ 2012JoG11J186. Britto, B., 2017. Actualizaci´ on de las Ecorregiones Terrestres de Perú propuestas en el Libro Rojo de Plantas End´ emicas del Perú. Gayana. Bot. 74 (1), 15–29. https://doi. org/10.4067/S0717-66432017005000318. Calizaya Llatasi, E., Mejía Marcacuzco, J., 2018. Modelamiento de la desglaciaci´ on de los nevados de la cuenca del Río Santa y su impacto en los recursos hídricos. Revista INGENIER´ IA UC 25 (3), 347–356. https://www.redalyc.org/journal/707/70 757670005/html/. Chevallier, P., Pouyaud, B., Suarez, W., Condom, T., 2011. Climate change threats to environment in the tropical andes: glaciers and water resources. Reg. Environ. Change 11 (S1), 179–187. https://doi.org/10.1007/s10113-010-0177-6. Curo Rosales, Y.A., Celmi Henostroza, G.T., Mejía Camones, M.D., Fern´ andez Vega, J. de D., 2023. Inventario Nacional de Glaciares y Lagunas de Origen Glaciar 2023. http s://doi.org/10.36580/inaigem.document17. Dextre, R.M., Eschenhagen, M.L., Camacho Hern´ andez, M., Rangecroft, S., Clason, C., Couldrick, L., Morera, S., 2022. Payment for ecosystem services in Peru: assessing the socio-ecological dimension of water services in the upper santa river basin. Ecosyst. Serv. 56, 101454. https://doi.org/10.1016/j.ecoser.2022.101454. Díaz Aguilar, R.D., Sanchez Larico, S.V., Lujano Laura, E., Lujano Laura, A., 2017. An´ alisis multi-temporal entre 1975 y 2015 sobre cambios de la cobertura glaciar en los nevados Allin Capac y Chichi Capac, Perú. Rev. Investig. Altoandin. - J. High Andean Res. 19 (3), 265–274. https://doi.org/10.18271/ria.2017.291. Emmer, A., 2023. Vanishing evidence? On the longevity of geomorphic GLOF diagnostic features in the tropical andes. Geomorphology 422, 108552. https://doi.org/ 10.1016/j.geomorph.2022.108552. García-Tadeo, D.A., Montoya-Zavaleta, M., Tan, Y., 2023. Understanding the susceptibility of the tropical proglacial environment in Peru using optical imagery and radon measurements. Atmosphere 14 (3), 568. https://doi.org/10.3390/ atmos14030568. Georges, C., 2004. 20th-Century glacier fluctuations in the tropical Cordillera Blanca, Perú. Arctic Antarct. Alpine Res. 36 (1), 100–107. https://doi.org/10.1657/15230430(2004)036[0100:TGFITT]2.0.CO;2. Gilberto Medina, D., Mauro Rodríguez, C., Abel Mejía, M.M., 2022. Estadística estoc´ astica multifractal aplicada al estudio del huascar´ an en la cordillera blanca. Prospectiva Universitaria 7 (1), 61–66. https://doi.org/10.26490/uncp. prospectivauniversitaria.2010.7.1151. Glasser, N.F., Clemmens, S., Schnabel, C., Fenton, C.R., McHargue, L., 2009. Tropical glacier fluctuations in the Cordillera Blanca, Peru between 12.5 and 7.6ka from cosmogenic 10Be dating. Quat. Sci. Rev. 28 (27–28), 3448–3458. https://doi.org/ 10.1016/j.quascirev.2009.10.006. Gordon, R.P., Lautz, L.K., McKenzie, J.M., Mark, B.G., Chavez, D., Baraer, M., 2015. Sources and pathways of stream generation in tropical proglacial valleys of the Cordillera Blanca, Peru. J. Hydrol. 522, 628–644. https://doi.org/10.1016/j. jhydrol.2015.01.013. Hinzmann, A., M¨ olg, T., Braun, M., Cullen, N.J., Hardy, D.R., Kaser, G., Prinz, R., 2024. Tropical glacier loss in East Africa: recent areal extents on Kilimanjaro, Mount Kenya, and in the Rwenzori Range from high-resolution remote sensing data. Environ. Res.: Climate 3 (1), 011003. https://doi.org/10.1088/2752-5295/ad1fd7. Hugonnet, R., McNabb, R., Berthier, E., Menounos, B., Nuth, C., Girod, L., Farinotti, D., Huss, M., Dussaillant, I., Brun, F., K¨ a¨ ab, A., 2021. Accelerated global glacier mass loss in the early twenty-first century. Nature 592 (7856), 726–731. https://doi.org/ 10.1038/s41586-021-03436-z. IGP, 2023. Eventos El Ni˜ no y La Ni˜ na Costero. Subdirecci´ on De Ciencias De La Atm´ osfera E Hidr´ osfera. http://met.igp.gob.pe/elnino/lista_eventos.html. IMARPE, 2023. ´ Indice Costero El Ni˜ no (ICEN). INSTITUTO DEL MAR DEL PERU. https:// www.imarpe.gob.pe/imarpe/index2.php?id_seccion=I0178090300000000000000. INAIGEM, 2017. Informe de la Situaci´ on de los Glaciares y Ecosistemas de Monta˜ na en el Perú. https://www.inaigem.gob.pe/wp-content/uploads/2019/04/Interiores-Inf orme-anual-2017.pdf. Ingemmet, 1989. Estudio Geodin´ amico de la Cuenca del Río Santa. https://repositorio. ingemmet.gob.pe/handle/20.500.12544/313. U.F. Giraldo Malca et al. Quaternary Science Advances 19 (2025) 100286 8