Lopez Cadierno, Javier and Borge García, Rafael and Lumbreras Martin, Julio and Rodriguez Hurtado, Encarnación and Narros Sierra, Adolfo
Study of Boundary Conditions influence on CMAQ simulations over the Iberian Peninsula.
In: "13th International Conference on Harmonization within Atmospheric Dispersion Modelling for Regulatory Purposes", 01/06/2010 - 04/06/2010, París, Francia.
Air pollution continues to be a major concern in many regions of the world, including Europe, and effective abatement measures are required to meet air quality standards in future years. The Integrated Air Quality Modelling System for the Iberian Peninsula (SIMCA) is a project that it is being implemented in Spain to support the design and evaluation of air quality strategies and abatement plans. The system relies on the WRF+SMOKE+CMAQ models. An important step in the development and implementation of SIMCA consists of understanding the uncertainties involved in modelling process. In this paper, a sensitivity analysis of the CMAQ model to the boundary conditions (BC) needed to simulate air pollution levels in the Iberian Peninsula is presented. The main objective of the study is to assess the model response to different alternatives to supply boundary conditions in the context of the Integrated Assessment Modelling (IAM) activities needed to provide an effective support to the policy-making process. Three ways to provide BC are tested: - fixed, time-independent, concentration profiles - concentrations predicted in a CMAQ mother domain (48 km, 1h resolution) - concentration values from the Geos-CHEM chemical-transport global model (2x2.5º, 3h resolution) The CMAQ model has been run in two episodes using identical configuration and input data, except for the BC. Model outputs for the main pollutants (SO2, NO2, NO, O3, PM10 and PM2.5) are then compared with observed concentration values from 165 monitoring stations. The stations selected are distributed across the Iberian Peninsula and include a balanced number of rural, industrial, urban background and traffic locations. The comparison is based on the analysis of a number of statistics commonly used for model evaluation and considers several aggregation levels so the influence of the BC can be assessed in a very detailed way. According to this study, there is no single valid method of providing boundary conditions since the performance of the alternatives tested depend on the pollutant, episode and type of location. Besides the statistical evaluation some other relevant issues in the context of IAM are discussed to gain a better insight into the suitability of the methods analysed