During summer, the city of Grenoble, which is located in a mountainous region, and its surroundings are subjected to periods of summer smog. This smog results from complex interactions of meteorological, chemical, biological and anthropogenic processes. To judge their relative importance for the creation and destruction potential of species like ozone, computer models were found to be very important. This project has been organised and supported by the agency of air quality of Grenoble ASCOPARG and the EPFL.
The French Alps city of Grenoble and its surroundings, with more than 400 thousand inhabitants, experiences periods of high ozone concentrations in summer time. Furthermore, Grenoble is located in a Y-shaped convergence of three deep valleys with surrounding mountains as high as 3000 m above sea level (asl), which leads to a very complex wind pattern. In summer 1999, a major field measurement campaign, involving standard ground-based stations, two LIDAR systems, airborne measurements, DOAS, microwave and SODAR wind profilers, was performed in order to obtain a suitable measurement database required for the modelling study (Quaglia, 2000). The first Intensive Observation Period (IOP), from 25th to 27th July, was characterized by the presence of a stable Azores's anticyclone over the Alps. The second IOP came on 2nd to 4th August. On July the 27th a strong photochemical plume with ozone concentrations up to 95 ppb at 20 km south of Grenoble (the highest O3 concentration measured during GRENOPHOT) was measured at ground level and sounded. This first IOP was selected as simulation base case for sunny summer periods leading to high photo-oxidant levels in the southern part of Grenoble. METPHOMOD, a photochemical airshed model (Perego, 1999), developed at the University of Bern and the Swiss Federal Institute of Technology in Lausanne, is used to investigate the air pollution dynamics in the Grenoble area during the first IOP.
With the horizontal one, the great field of simulation is a square of 240 km sides, directed North-South. It contains 40 X 40 cells of 6 X 6 km. This field covers the whole of the Rhone-Alps area and its principal cities. When with the small field of simulation, it covers a rectangle of 68 by 78 km on sides, centered on Grenoble and directed northern south, with 34 X 39 cells of 2 X 2 km. Figure 2 shows the fields of simulation and the iso ligne of ground. With the vertical, we took a system of Cartesian coordinates, identical for the large one and the small field, with 26 levels having an increasing interval from 50 to 8000 m at the node of the model.
Metphomod calculates the meteorological (wind, temperature, humidity, among many others) and air quality dynamics based on stated-of-the-art meteorological and atmospheric chemistry modules. Furthermore it computes transport, diffusion and chemical phenomena. Two of the most important components of this model are the transport algorithm and the chemical mechanism. The chemical mechanism used is the so-called Reactive Atmospheric Chemical Mechanism (RACM) lumped species mechanism (Stockwell, 1997). A large amount of data and boundary conditions are required to run METPHOMOD. In the Grenoble case, the one-way nesting technique is used. The nesting domain is a 198 km square with a grid resolution of 6 x 6 km. The small one is a 78 x 68 km rectangle with a grid resolution of 2 x 2 km. Each grid has 24 vertical levels up to 8000 m asl.
Concerning the meteorological data, we consider the boundary wind direction and speed, temperature, air humidity and the upper layer pressure. These data, provided and calculated by the Swiss Institute of Meteorology with their proprietary synoptic scale model, enable us to initialize the boundary condition on the nesting model. The Swiss Model (SM) is the hydrostatic meso-beta scale numerical weather prediction model (grid mesh 14 km) developed in a joint effort by the Deutscher Wetterdienst (DWD) and MeteoSwiss as a high-resolution version of the Europa Modell of the DWD [Majewski, 1991]. It was the operational model at MeteoSwiss from 1994 until March 2001.
Two emission inventories were prepared by the Centre Interprofessionel d'Etude de la Pollution de l'Air (CITEPA). The first on for the large grid resolution (6 x 6 km), and the second for the town of Grenoble zooms with a resolution of 1 x 1 km. These inventories distinguish among surface (9 sectors), linear (motorways and highways) and point emissions, and are space and time-resolved (distinction is made between weekday and Sunday). The inventories include the emissions of NOx, CO, CH4, and 23 non-methane hydrocarbon species (NMHC). These emissions are lumped to the 32 classes of VOC as required for the RACM (Stockwell 1997) mechanism. The highest emission densities are located to the cities of Lyon and Grenoble, and along the major highways form them. The biogenic emissions are estimated within the surface emission inventory, forest sector.
The model was set up to simulate the meteorology and the temporal and spatial evolution during the first IOP from 25th to 27th July. The focus was set on the region in the south of Grenoble where the O3 plume reached the highest concentration. In the following comparisons, the dots depict model results and the solid line represents measurements.
Description and location of ground measurement and LIDAR stations for comparison with model
During the observation period the synoptic weather in France was dominated initially by north winds and then by south windsnorth winds. This period generates peak ozone concentrations from 80 to 95 ppb. The models results show a complex interaction between the synoptic weather situation and local wind systems, developed in the complex Grenoble topography. All these wind systems are generally well reproduced by Metphomod.
Comparison between ground measurements and model calculations from 25th to 27th July for wind direction (°) and wind speed (m/s). Solid blue lines are measurements and black dots are model results.
Comparison between ground measurements and model calculations from 25th to 27th July for O3 [ppb]. Solid blue lines are measurements and black dots are model results.
Modeled O3 values [ppb] on the surface on 25th July at (a) 16h00, (b) 20h00 and 27th July at (c) 9h00, and (d) 17h00 LT.
Measured and simulated vertical O3 values [ppb] between the 25th and the 27th July. Measured vertical O3 values [ppb] was performed by the EPFL mobile tropospheric ozone UV DIAL lidar.
To forecast smog episod over the the Grenoble region, the LEGI at the University of Grenoble in France and the the LPAS at the EPFL in Switzerland have collaborated to build a mesoscale model chain using MM5 (Gregg and all.,1 1994), CHIMERE (Vautard and all., 2001)and METPHOMOD (Perego, 1999). At this time, you can have a look to air pollution forecast for Grenoble region on the following link .
![[LPAS]](/PAS/MOD//lpas_back.gif){kind=small}
![[DGR]](/icon_gif/back_dgr.gif){kind=icon}
![[EPFL]](/icon_gif/back_epfl.gif){kind=icon}
![[MAIL]](/icon_gif/mail.gif){kind=icon}
Auteur: Olivier Couach. Rédacteur:
Olivier Couach