Abstract's details
Monitoring surface water storage in Peruvian Andean lakes and reservoirs with SWOT and Sentinel-1: Toward integrated mountain hydrology
Event: 2025 SWOT Science Team Meeting
Session: Hydrology: SWOT Lakes, Estuaries and Wetlands (SLEW)
Presentation type: Poster
Monitoring surface water storage in high-altitude regions remains challenging due to sparse in-situ data. This study evaluates a multi-sensor approach to estimate lake volume changes in three Andean water bodies: Lake Junín, Lake Sibinacocha, and the Yuracmayo reservoir by integrating surface elevations from the SWOT mission with surface areas derived from Sentinel-1 SAR imagery.
The SWOT Lake Vector product (L2_HR_LakeSP) was successfully applied to Yuracmayo, where compact morphology and well-defined shorelines enabled accurate elevation retrieval. Estimated volumes correlated well with in-situ data (R = 0.80), capturing both seasonal variation and abrupt changes from reservoir operations (9–49 hm³). In contrast, the vector product was unsuitable for Junín and Sibinacocha due to shoreline complexity: wetlands in Junín and narrow, fragmented geometry in Sibinacocha impeded consistent elevation extraction. For these sites, elevation retrieval was obtained from SWOT pixel cloud (L2_HR_PIXC). Elevation retrieval from the pixel cloud involved spatial filtering and classification to isolate valid water surfaces.
Volume dynamics observed in 2024 revealed distinct hydrological regimes across the three study sites. Lake Junín exhibited a pronounced wet-season peak, with volumes rising from 59 hm³ in January to 350 hm³ in May, then declining below 65 hm³ by November characteristic of a precipitation-driven system. Sibinacocha showed a slower, more gradual increase from 50 to 94 hm³ by June, reflecting sustained inputs from glacial melt. Yuracmayo displayed abrupt, stepwise changes in storage, increasing from 18 hm³ in December 2023 to 49 hm³ in April, then dropping to 9 hm³ by November, consistent with managed releases for water supply.
These contrasting dynamics demonstrate SWOT’s capacity to resolve diverse storage behaviors when product selection is adapted to lake morphology and observation conditions. While the Lake Vector product proved effective in Yuracmayo, accurate elevation retrieval in Junín and Sibinacocha requires processing from the pixel cloud due to shoreline complexity. The SWOT and Sentinel-1 framework thus offers a flexible and scalable solution for monitoring surface water storage in data-scarce, high-mountain basins.
Back to the list of abstractThe SWOT Lake Vector product (L2_HR_LakeSP) was successfully applied to Yuracmayo, where compact morphology and well-defined shorelines enabled accurate elevation retrieval. Estimated volumes correlated well with in-situ data (R = 0.80), capturing both seasonal variation and abrupt changes from reservoir operations (9–49 hm³). In contrast, the vector product was unsuitable for Junín and Sibinacocha due to shoreline complexity: wetlands in Junín and narrow, fragmented geometry in Sibinacocha impeded consistent elevation extraction. For these sites, elevation retrieval was obtained from SWOT pixel cloud (L2_HR_PIXC). Elevation retrieval from the pixel cloud involved spatial filtering and classification to isolate valid water surfaces.
Volume dynamics observed in 2024 revealed distinct hydrological regimes across the three study sites. Lake Junín exhibited a pronounced wet-season peak, with volumes rising from 59 hm³ in January to 350 hm³ in May, then declining below 65 hm³ by November characteristic of a precipitation-driven system. Sibinacocha showed a slower, more gradual increase from 50 to 94 hm³ by June, reflecting sustained inputs from glacial melt. Yuracmayo displayed abrupt, stepwise changes in storage, increasing from 18 hm³ in December 2023 to 49 hm³ in April, then dropping to 9 hm³ by November, consistent with managed releases for water supply.
These contrasting dynamics demonstrate SWOT’s capacity to resolve diverse storage behaviors when product selection is adapted to lake morphology and observation conditions. While the Lake Vector product proved effective in Yuracmayo, accurate elevation retrieval in Junín and Sibinacocha requires processing from the pixel cloud due to shoreline complexity. The SWOT and Sentinel-1 framework thus offers a flexible and scalable solution for monitoring surface water storage in data-scarce, high-mountain basins.