Summary

The operational air quality models have received an important interest from the scientific and administration communities during the last years. OPANA is one of these air quality models, which have been developed based on academic versions (i.e. ANA model). ANA model is a mesoscale air quality modelling system (San José et al., 1997) which is composed on a non-hydrostatic mesoscale meteorological model REMEST, a chemical module CHEMA, a deposition module DEPO and an emission model EMIMA. This model is developed in FORTRAN and C languages. The operational version of this type of tools is called OPANA. DGXIII funded into the Telematics for Environment area in 1996-1998 a project called EMMA. Our laboratory was involved in this project and the operational version of ANA (OPANA) was incorporated into the EMMA project for being applied for the Madrid domain. In this contribution we will describe the EMMA application and the main results.
 
 

Aim of the research: The EMMA project

EMMA is dealing with the development, test and validation of innovative telematics systems for the monitoring and forecasting of air pollution in urban areas. Telematics components including multimedia products and services are extensively used in the project: a) specialized networks, b) GIS and multimedia archiving systems and high definition terminals. Pollution and weather data are collected through highly automated telematics networks for all fixed stations in the metropolitan areas. These data are stored in integrated relational databases and are analyzed and distributed according to EMMA architecture and presented in a user-driendly way by using graphical tools. Madrid city is one of the demo sites in EMMA project. Observations of SO₂, NO_x, O₃, CO, PM10, aromatic and radioactivity are used together with humidity, wind, temperature, and turbulence, including vertical and horizontal meteorological conditions. Air quality forecasting within a time horizon of 24-48 hours is one of the main objectives in the EMMA project.

A special feature of this architecture is the high degree of interaction between meteorology and environmental phenomena. A combination of an accurate prediction of meteorology, emission, deposition and photochemistry on fast workstation platforms is shown.

The Madrid EMMA application was developed with the tcl/tk language which is a UNIX based graphical user interface. The unique characteristics of the Tcl/tk allowed developing a software tool, which allows the user to interact easily with the complex air quality models in a comprehensive way. The operational procedure is described as follows: the user should have ready meteorological and monitoring network information in the Environmental Office. This information should be brought to the Office under quasi-real-time basis. The EMMA tool has configuration windows and a visualization window. The configuration windows deal with the proper location of initial meteorological and air quality monitoring information. The user should start the execution of the model on 12-14 hours on day 1 and he should use the meteorological and air quality monitoring data from the day 0. The simulation should run 120 hours (under typical working day conditions) and should not last more than 18-20 hours in order to have it ready for day 2, 8-9 hours.
 
 

Activities during the year

The EMMA model was installed successfully in the Madrid Community Environmental Offices. The system started to produce preliminary environmental forecasting information on July, 1998; although the project ended officially on September, 1998, the system continues to work and producing results daily. The predicted surface concentrations for different cells and times are used for warning to the population in case of concentrations above EU directive limits. Figure 1 shows results for Madrid, October, 4, 1998, day 1, 5h00 and for NO surface concentrations.
 
 
 
 

Figure 1: EMMA application over Madrid. EMMA version already installed in Madrid Community Environmental Office into the EMMA project.

The new version of EMMA is under development. The new version will include a JAVA interface for automatic access to INTERNET/WWW environmental data. Also, the new version will allow the user to select a specific domain to carry out a detailed analysis of the different emission sources and the inmission data. Figure 2 shows how this selected domain looks like.

Figure 2: Zooming capabilities for the newer versions of EMMA tool.
 

The zooming capabilities are based on the application of a Gaussian model (ISCST3, EPA, 1997) inside the gridcell (or grouped gridcells). The Gaussian model is applied by using the detailed (in time and space) meteorological information produced by EMMA tool during the specific period of interest. The Gaussian model is applied over the different emission sources of interest. The isoconcentration lines can be used for analysis and comparison studies with observed data from the air quality monitoring stations. Figure 3 shows the comparison between observed data in monitoring station "Mostoles" (426829,4464396) in the south-west of Madrid domain and modelled data by using the meteorology produced by EMMA tool as input for the thousands of Gaussian models which are applied every hour over the mobile sources (traffic mobile units such as cars, trucks, etc.). Each traffic source is considered as a point source emission and it is modelled by using a Gaussian model during one hour assuming that the meteorological conditions are taken from the EMMA (REMEST) output every hour. The results showed in Figure 3 shows that the "local" traffic emissions account for a very important part of the total air concentrations during the 24 hours of simulation (August, 31, 1998) and particularly during daytime hours the "local" traffic emissions are responsible of most of NOx air concentrations in this particular simulation. August, 31, 1998 is typically the summer holiday return day for many of Madrid citizens.
 
 

Figure 3: Comparison between observed data in "Mostoles" monitoring station and one hour accumulated simulated data by using Gaussian modelling with traffic vehicles. Meteorological input data is taken from EMMA tool.
 
 
 
 

Main conclusions

An operational version of ANA model has been built and applied into the EMMA project (DGXIII) over the Madrid domain. The tcl/tk graphical user interface has been used. The EMMA tool has been successfully installed at Madrid Community Environmental Offices. EMMA Web site shows the predicted 24-48 hours surface air concentrations on real-time. Newer versions of EMMA include the possibility of carrying out local analysis of different local emission sources and study the different impact on current air concentrations. Using the ISCST3 Gaussian model (EPA) does this and being applied over the thousands of different traffic vehicles (each vehicle is a point source and as a consequence a Gaussian model run is executed). Meteorological input data for the ISCST3 model is taken from EMMA model, so that, the quality of this information allows an excellent performance of the Gaussian approach.
 
 

References

ISCST3 (U.S. EPA) (1997) http://www.epa.gov.

San José R., Prieto J.F., Castellanos N. and Arranz J.M. Sensitivity study of dry deposition fluxes in ANA air quality model over Madrid mesoscale area. Measurements and Modelling in Environmental Pollution Ed. R. San José and C.A. Brebbia. CMP, ISBN 1 85312 461 3 (1997).