Monitoring Spatial and Temporal Differences in Andean Snow Depth Derived From Satellite Tri-Stereo Photogrammetry

Abstract

Quantifying the high elevation winter snowpack in mountain environments is crucial for lowland water supply, though it is notoriously difficult to accurately estimate due to a lack of observations and/or uncertainty in the distribution of meteorological variables in space and time. We compare high spatial resolution (3 m), satellite-derived snow depth maps for two drought years (2017 and 2019) in a high mountain catchment of the central Chilean Andes, applying a recently updated methodology for spaceborne photogrammetry. Regional weather station observations revealed an 80% reduction in precipitation for 2019 (the second driest winter since 1950) relative to 2017, though only a 10% reduction in total snow-covered area is seen in the satellite imagery. We threshold surface height changes based upon uncertainty of stable (snow-free) terrain differences for topographic characteristics of the catchment (slope, aspect, roughness etc). For a conservative analysis of change, outside of the topographically-derived confidence intervals, we calculate a mean 0.48 ± 0.28 m reduction of snow depth and a 39 ± 15% reduction in snow volume for 2019, relative to 2017 (for 23% of the total catchment area). Our findings therefore quantify, for the first time in the Andes, the relationship of high-resolution mountain snow depth observations with low elevation precipitation records and characterise its inter-annual variability over high elevation, complex terrain. The practical use of such detailed snow depth information at high elevations is of great value to lowland communities and our findings highlight the clear need to relate the high spatial (Pléiades) and temporal (in-situ) scales within the available datasets in order to improve estimates of region-wide snow volumes.

Comparison of GRACE and GNSS Seasonal Load Displacements Considering Regional Averages and Discrete Points

“…Water supply for the semiarid Chilean Andes is largely reliant on melt from seasonal snow and ice bodies from the Andean Cordillera (Azócar & Brenning, 2010;Favier et al, 2009). Shaw et al (2020) estimated the winter snow depth and the snow melt in the Andean watershed based on optical photogrammetry. The annual snowmelt reaches 3 × 10 4 m w.e..…”

“…Shaw et al. (2020) estimated the winter snow depth and the snow melt in the Andean watershed based on optical photogrammetry. The annual snowmelt reaches 3 × 10 4 m w.e..…”

Improving the Spatial Distribution of Snow Cover Simulations by Assimilation of Satellite Stereoscopic Imagery

“…However, the assimilation of data artificially degraded to lower accuracy still improved point simulation of the snowpack (Smyth et al., 2020). Very high‐resolution stereoscopic satellites such as Pléiades (0.5 m) or Worldview (0.3 m) provide HS maps with a lower accuracy (∼0.7 m) than ground, drone or airborne methods but with less logistical constraints and less cost for the end user (Deschamps‐Berger et al., 2020; Eberhard et al., 2021; Marti et al., 2016; McGrath et al., 2019; Shaw, Deschamps‐Berger, et al., 2020; Shaw, Gascoin, et al., 2020). With this method, HS maps are calculated by differencing two digital elevation models derived from stereoscopic images with and without snow.…”

Application of Fixed-Wing UAV-Based Photogrammetry Data for Snow Depth Mapping in Alpine Conditions

“…In general, there are methods based on sparse in situ measurements with various probes or geophysics measurement techniques [10,11] and methods based on remote sensing. Remote sensing methods represent unmanned aircraft vehicles (UAV) mapping [12,13], conventional airborne mapping and airborne laser-scanning [14][15][16], terrestrial laserscanning [17][18][19], or usage of satellite imagery [20][21][22]. Especially laser scanning techniques, also known as LiDAR (Light Detection and Ranging), have seen a dramatic widening of applications in the natural sciences [23].…”