Changes in Andes snow cover from MODIS data, 2000–2016

Freddy A. Saavedra, Stephanie K. Kampf, Steven R. Fassnacht, Jason S. Sibold
2018 The Cryosphere  
<p><strong>Abstract.</strong> The Andes span a length of 7000<span class="thinspace"></span>km and are important for sustaining regional water supplies. Snow variability across this region has not been studied in detail due to sparse and unevenly distributed instrumental climate data. We calculated snow persistence (SP) as the fraction of time with snow cover for each year between 2000 and 2016 from Moderate Resolution Imaging Spectroradiometer (MODIS) satellite sensors (500<span
more » ... span class="thinspace"></span>m, 8-day maximum snow cover extent). This analysis is conducted between 8 and 36°<span class="thinspace"></span>S due to high frequency of cloud (&amp;gt; <span class="thinspace"></span>30<span class="thinspace"></span>% of the time) south and north of this range. We ran Mann–Kendall and Theil–Sens analyses to identify areas with significant changes in SP and snowline (the line at lower elevation where SP<span class="thinspace"></span> = <span class="thinspace"></span>20<span class="thinspace"></span>%). We evaluated how these trends relate to temperature and precipitation from Modern-Era Retrospective Analysis for Research and Applications-2 (MERRA2) and University of Delaware datasets and climate indices as El Niño–Southern Oscillation (ENSO), Southern Annular Mode (SAM), and Pacific Decadal Oscillation (PDO). Areas north of 29°<span class="thinspace"></span>S have limited snow cover, and few trends in snow persistence were detected. A large area (34<span class="thinspace"></span>370<span class="thinspace"></span>km<sup>2</sup>) with persistent snow cover between 29 and 36°<span class="thinspace"></span>S experienced a significant loss of snow cover (2–5 fewer days of snow year<sup>−1</sup>). Snow loss was more pronounced (62<span class="thinspace"></span>% of the area with significant trends) on the east side of the Andes. We also found a significant increase in the elevation of the snowline at 10–30<span class="thinspace"></span>m<span class="thinspace"></span>year<sup>−1</sup> south of 29–30°<span class="thinspace"></span>S. Decreasing SP correlates with decreasing precipitation and increasing temperature, and the magnitudes of these correlations vary with latitude and elevation. ENSO climate indices better predicted SP conditions north of 31°<span class="thinspace"></span>S, whereas the SAM better predicted SP south of 31°<span class="thinspace"></span>S.</p>
doi:10.5194/tc-12-1027-2018 fatcat:jn5vqbbvvbcobpw27njdf6h4dy