961425 - Snow Albedo Variation Due to Light Absorbing Particles from the Period 2000 to 2020 in the El Yeso Basin, Chile

Snow albedo variation can be produced by deposition of light absorbing particles (LAP) such as black carbon (BC), organic carbon (OC) and dust, and in turn, alter hydrological cycle in snow-dominated basin. We have studied these parameters in the El Yeso basin (EY), which has a height from 2,509 to 6,103 and average of 3.582 m.a.s.l., and covers an area of 349 km² in the Central Chilean Andes Mountains. The EY has a dam that forms a big reservoir with a capacity of 250 million m3, and is operated by the Aguas Andinas (AA), as the main purification and sewage treatment company in Chile. This snow-dominated basin is the main source of fresh water for the Metropolitan region and serves to 7.3 million inhabitants (40% of the national population of Chile). Therefore, we analyzed the trend of 20-year monthly data (2000-2020) via the Theil – Sen slope of snow cover (SC), snow albedo (SA) and land surface temperature (LST) from Moderate Resolution Imaging Spectroradiometer (MODIS); BC, OC and dust from Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2); and inflow to the reservoir (Qin), generated mainly by the melting of snow and glaciers within the EY and measured by AA. Results showed a high reduction rate of SA in 0.95% month-1 (p <0.05) and a LAPs increase of 0.002 mg m-2 month -1 of BC (p <0.001) and 0.015 mg m-2 month-1 of OC (p <0.05) during the months of May. In addition, the greatest variations were observed for LST (0.3 °C month-1; p <0.001), SC (-1.56% month-1; p <0.001) and Qin (-0.56 m3 s-1 month-1; p <0.001) in the months of August, November and January, respectively. A lagged correlation between LAP over Qin showed maximum values for dust (69%, month lag 0), BC (51%, month lag -2) and OC (42%, month lag -3). This would indicate that the dust has an instantaneous, while the carbonaceous particles present a positive feedback on the Qin due to the enhancement of sunlight absorption from the BC and OC particles, and the increase of snow grains size over time. Additionally, a multiple regression model (R2 = 86%) showed SA reductions were mainly due to 37% (LST), 9% (BC), 10% (OC) and 14% (dust) during the austral spring months of November, month with high statistical significance for most of the variables analyzed. Overall, BC and OC showed a positive trend that coincided with negative trends of SA in EY. Our findings would demonstrate that the presence of BC, OC and dust were reducing the SA of the EY and alter its hydrological cycle. Furthermore, these results evidenced a relationship between LAP and hydrological variables, thus they could serve as inputs for hydrological-chemical models in the studied area. Acknowledgments: Projects ID19I10359 from ANID-FONDEF and PI_M_2020_58 from UFSM.