Changes of the tropical glaciers throughout Peru between 2000 and 2016 – mass balance and area fluctuations

Glaciers in tropical regions are very sensitive to climatic variations and thus strongly affected by climate change. The majority of the tropical glaciers worldwide are located in the Peruvian Andes, which have shown significant ice loss in the last century. Here, we present the first multi-temporal, region-wide survey of geodetic mass balances and glacier area fluctuations throughout Peru covering the period 2000–2016. Glacier extents are derived from Landsat imagery by performing automatic glacier delineation based on a combination of the NDSI and band ratio method and final manual inspection and correction. The mapping of debris-covered glacier extents is supported by synthetic aperture radar (SAR) coherence information. A total glacier area loss of inline-formula $M1inlinescrollmathml-\mathrm{normal\; 548.5}\pm \mathrm{normal\; 65.7}$ 70pt10ptsvg-formulamathimg4a56d867710ad1d0e6b59b40efc7543a tc-13-2537-2019-ie00001.svg70pt10pttc-13-2537-2019-ie00001.png  kminline-formula² (inline-formula−29 %, inline-formula−34.3 kminline-formula² ainline-formula⁻¹) is obtained for the study period. Using interferometric satellite SAR acquisitions, bi-temporal geodetic mass balances are derived. An average specific mass balance of inline-formula $M7inlinescrollmathml-\mathrm{normal\; 296}\pm \mathrm{normal\; 41}$ 52pt10ptsvg-formulamathimgdaccc0027f423fe62c08bfd3497e240a tc-13-2537-2019-ie00002.svg52pt10pttc-13-2537-2019-ie00002.png  kg minline-formula⁻² ainline-formula⁻¹ is found throughout Peru for the period 2000–2016. However, there are strong regional and temporal differences in the mass budgets ranging from inline-formula45±97 to inline-formula $M11inlinescrollmathml-\mathrm{normal\; 752}\pm \mathrm{normal\; 452}$ 58pt10ptsvg-formulamathimgfcf2e06ca9213df129dd310b27a4dc03 tc-13-2537-2019-ie00003.svg58pt10pttc-13-2537-2019-ie00003.png  kg minline-formula⁻² ainline-formula⁻¹. The ice loss increased towards the end of the observation period. Between 2013 and 2016, a retreat of the glacierized area of inline-formula $M14inlinescrollmathml-\mathrm{normal\; 203.8}\pm \mathrm{normal\; 65.7}$ 70pt10ptsvg-formulamathimgea393321be0a3323d1a14edbade0f209 tc-13-2537-2019-ie00004.svg70pt10pttc-13-2537-2019-ie00004.png  kminline-formula² (inline-formula−16 %, inline-formula−101.9 kminline-formula² ainline-formula⁻¹) is mapped and the average mass budget amounts to inline-formula $M20inlinescrollmathml-\mathrm{normal\; 660}\pm \mathrm{normal\; 178}$ 58pt10ptsvg-formulamathimg4d564a6ec6bec32c37de7ae43c3acb5d tc-13-2537-2019-ie00005.svg58pt10pttc-13-2537-2019-ie00005.png  kg minline-formula⁻² ainline-formula⁻¹. The glacier changes revealed can be attributed to changes in the climatic settings in the study region, derived from ERA-Interim reanalysis data and the Oceanic Nino Index. The intense El Niño activities in 2015/16 are most likely the trigger for the increased change rates in the time interval 2013–2016. Our observations provide fundamental information on the current dramatic glacier changes for local authorities and for the calibration and validation of glacier change projections.