The Snowline and 0°C Isotherm Altitudes During Precipitation Events in the Dry Subtropical Chilean Andes as Seen by Citizen Science, Surface Stations, and ERA5 Reanalysis Data

Abstract

Understanding the variability of the snowline and the closely related 0°C isotherm during infrequent precipitation events in the dry Andes in Chile is fundamental for precipitation, snow cover, and discharge predictions. For instance, it is known that on the windward side of mountains, the 0°C isotherm can be several hundreds of meters lower than on the free air upwind counterpart, but little is understood about such effects in the Andes due to missing in situ evidence on the precipitation phase. To bridge this gap, 111 photographs of the snowline after precipitation events between 2011 and 2021 were gathered in the frame of a citizen science programme to estimate the snowline altitude. Since photographs of the mountain snowline are in good agreement with Sentinel-2 imagery, they have great potential to validate empirical snowline estimations. Using the snowline altitude from the photos, we evaluated different methods to estimate the snowline and 0°C isotherm altitude during precipitation events based on surface meteorological observations and ERA5 reanalysis data. We found a high correlation between the observed snowline altitude and the extrapolated 0°C isotherm based on constant lapse rates (−5.5 to −6.5°C km−1) applied to air temperature from single, near stations. However, uncertainty increases for distances >10 km. The results also indicate that the linear regression method is a good option to estimate ZSL, but the results strongly depend on the availability of high-elevation station datasets. During half of the precipitation events, the 0°C isotherm lies between ∼1,800 and ∼2,400–2,500 m asl. in winter, and the snowline is on average ∼280 m below this altitude. Our results indicate the presence of a mesoscale lowering of the 0°C isotherm over the windward slopes compared to the free-air upwind value during precipitation events and a possible isothermal layer of near-freezing air temperatures comparable to other mountain ranges. Due to this mesoscale and local behavior, ERA5 data generally overestimate the snow–rain transition in high-elevation areas, especially for relatively intense events. On the other hand, the 0°C isotherm altitude is underestimated if only low-elevation valley stations are considered, highlighting the importance of high-altitude meteorological stations in the network.