Trends in occurrence of glacial lake outburst floods (GLOFs) in space and time as well as frequency-magnitude relationships are typically derived from inventories of past GLOFs. Searching for geomorphic evidence (GLOF diagnostic features) in remote sensing images is a commonly used method for building GLOF inventories. In this study, I examine the longevity of geomorphic GLOF diagnostic features and thus the reliability of GLOF inventories that build on them. A set of 160 GLOFs documented from the Tropical Andes of Peru and Bolivia is analyzed, focusing on four GLOF diagnostic features: (i) breached dams; (ii) pre-GLOF water levels; (iii) outwash fans; (iv) impact areas. A total of 359 GLOF diagnostic features are identified and their evolution is analyzed from multi-decade remote sensing images, with special focus on GLOF diagnostic features that vanished in time. Building on these data, the expected longevity of GLOF diagnostic features is outlined, revealing long persistence of breached dams (>102 years), long persistence of pre-GLOF water levels (up to 102 years) and comparably more intense degradation and vanishing of outwash fans and impact areas. The main degradation processes of GLOF diagnostic features are identified, including the succession of vegetation, geomorphic reworking and anthropogenic activities. Further, the implications for building GLOF inventories are discussed, highlighting that vanishing geomorphic evidence of GLOF diagnostic features may lead to: (i) underestimated spatial extent of GLOF impact areas; and (ii) underestimated total numbers of GLOFs in GLOF inventories. Considering the number of major GLOFs as predictors of the number of minor GLOFs, an example of the Tropical Andes shows that the total number of GLOFs in the post-LIA period may be underestimated by up to 40 % (reference period 2011–2020).
Read full abstract