Characterizing Nitrogen and Sulfur Deposition in Latin America: Simulation and Evaluation

Haley Lewis1, Alexandra Ponette2, Kathleen Weathers3, Donna Schwede4 and Barron Henderson5

Population growth and economic development in Latin America is significantly increasing reactive nitrogen and sulfur emissions that acidify terrestrial and aquatic ecosystems. Multiple studies have been conducted in order to characterize nitrogen and sulfur in the United States and Europe, while few studies have been conducted within Latin America. The goal of this project is to provide a first estimate total deposition of nitrogen and sulfur in Latin America and provide observational constraints.

We use GEOS-Chem (Yantosca et al. 2015) to calculate total N and S deposition and compare to a collection of observations from previous studies accounting for observational artifacts. We collected observations of NO3N, NH4N, and SO4S wet deposition from collectors without automatic closure systems. As a result, observed “wet deposition” includes artifacts from aerosol dry deposition. We predicted deposition using GEOS-Chem v9 with standard emissions options and meteorology from GEOS-5 for all years. Predictions were post processed to create NO3N, NH4N, and SO4S from the sum of wet deposition and particulate dry deposition. We quantified deposition bias in Latin America for nitrogen and sulfur, with artifacts, at 14 sites from 2006-2010.

Normalized mean biases for total deposition of NO3N, NH4N, and NO3N+NH4N had the following means respectively: .23, -.65, and -.26. The Normalized mean biases for the total depositions of NO3N, NH4N, and NO3N+NH4N had the following ranges respectively: -3.4 in Sao Paulo, Brazil to 2.1 in Manizales, Columbia. -3.2 in Sao Paulo, Brazil to 2.0 in Rio de Janeiro, Brazil, -6.5 in Sao Paulo, Brazil to 2.2 in Pergamino, Argentina. Sao Paulo, the most populated city within Brazil, consistently produced the largest and most negative normalized mean biases for nitrogen containing species.

Results show that in general, heavily populated cities produce larger and more negative normalized mean biases which suggests that the model is under-predicting total deposition.  Total deposition for SO4S did not show a consistent relationship between modeled and predicted results. Land use from GEOS-Chem will be used in order to determine the urban fraction of the grid scales for each of the 14 sites. The normalized mean biases will be regressed against the urban fraction to identify the role of anthropogenic emissions in the deposition bias. We will determine whether GEOS-Chem is a good predictor and whether it predicts better or worse in urban areas. We will present current results as well as the results from the regression.


1University of Florida,
2University of North Texas,
3Cary Institute,
4Environmental Protection Agency,
5University of Florida,