A TIME-EVOLVING INVENTORY OF TROPICAL GLACIERS IN BOLIVIA AND NORTHERN CHILE

Glaciers are often described as the water the towers of the world and are particularly important resources in regions with large precipitation seasonality such as the tropical Andes. Many communities throughout this region are experiences challenges and concerns regarding contemporary rapid tropical glacier loss. Inventories of glaciers in the region are vital for informed water management polices and improved forecasts of continued retreat. Databases, such as the Randolph Glacier Inventory (RGI), provide invaluable information about glacier location and size, but they only provide a snapshot of this rapidly changing resource. We use Landsat images and the digital elevation model (DEM) from the Shuttle Radar Topography Mission to inventory glaciers in parts of Bolivia and Northern Chile for three dry seasons spanning 16 years. Using normalized difference water indices, normalized difference snow indices and a semi-automated algorithm, we classify glacier cover for the years 2000, 2008, and 2016.

Our inventory region accounts for ~15% of tropical Andean glaciers by areal coverage. For the year 2000, we find a total areal coverage of 336.3 km² +/- 47.5 km, which is slightly less but within uncertainty of the areal converge for this region in the RGI. Over the 16-year study period, the region experiences ~44% (148 km²) area loss. The highest rates of loss are in the arid Western Cordillera. In the wet Eastern Cordilleras, the rates of loss are slightly higher than previously published values. Our 2008 inventory is ~72% the areal coverage in the RGI, and in 2016 it is ~54% the coverage. The glacier hypsometry of our inventories differs from that in the RGI, with the RGI having more glacial area at particularly low elevations and less glacial area at particularly high elevations as compared to our inventory. Discrepancies in elevation may reflect varying DEMs between inventories. Our study provides a new inventory of glaciers and glacier change for the rapidly evolving southern tropical Andes. It also highlights potential limitations of current glacier databases. As alpine landscapes continue to rapidly change, time-evolving inventories of vital human resources are essential for informed policy making and improved forecasts.