Abstract

Near Santiago de Chile and towards the Andes Mountains there are several areas sensitive to contamination, such as, the Natural Sanctuary Yerba Loca, National Park Rio Clarillo, several glaciers and the Farellones ski resort. The health of the glaciers is very important, because a large fraction of the water supply of Santiago is provided by them. To study the influence of Santiago's contamination on the glaciers, a black carbon monitoring campaign was performed from December 2014 (summer) until July 2015 (winter). Four monitors were placed between the city and the glaciers in the mountain along with meteorological stations. An analysis of the measurements indicates a direct transport of black carbon between Santiago de Chile and the Andes Mountains with a travel time of about 11 h in summer and 9 h in winter. Black carbon concentration at the mountain (La Parva) is higher in summer and lower in winter, a trend that is opposite to any other city in Chile. This is the only town in Chile in which transport, not local emissions, is mostly responsible for the BC observed. The fraction of BC in La Parva compared to Las Condes in the eastern part of Santiago changes from 14% in December to 2% in July. Model simulations of black carbon using high resolution meteorological data generated by a regional climate model confirms the difference in summertime and wintertime transport of black carbon from Santiago to La Parva. In December (summer) emissions from Santiago are estimated to contribute with 51% to the concentrations in La Parva while nearby sources from mining activities and long distance transport from other sources in central Chile contributes with 37 and 12% respectively. In July (winter) simulated contributions are dominated by long range transport from other sources with a 50% contribution while simulated contributions from sources in Santiago and from mining are both 25%. The contribution from mining activities is uncertain due to inaccuracies in simulated wind speed and direction in the mountains and too coarse model resolution and therefore warrants further investigation.