Air quality forecast

Atmospheric pollution and their modelling

Concentrations of pollutants in the atmosphere are controlled by four processes. The first is emission that is the introduction of pollution into the air from anthropogenic and natural sources. These compounds undergo chemical transformation in the atmosphere, among others, under the influence of solar radiation, water and other air components. Pollutions are also transported by wind at different distances from the source and then deposited on the Earth's surface by gravitational precipitation and absorption on the surface or eluviation by atmospheric precipitation. In order to be able to accurately forecast concentrations of pollutants in the air, each of these processes should be included in a chemical model. In order to achieve this, information about the emission of both anthropogenic pollutants and natural sources into the atmosphere as well as meteorological conditions at a given time is required. This information provides input information for atmospheric chemistry models.

There are two types of chemical models: Euler's and Lagrange's. In Langrange's model, a portion of air of constant mass is observed and it moves in a model area called a domain, after a certain trajectory. The modelling result is the sum of the trajectories of all portions. In Eulerian's approach, there is a focus on a small part of the space through which air flows freely and all parts of that space are modelled simultaneously. At present, Eulerian models are most often used in regional models and the forecasting of pollutant concentrations. In regional air quality forecasting, Eulerian models are most often used, among others, the WRF-Chem model used in the LIFE-APIS / PL project. It uses information about the state of the atmosphere (meteorological conditions and pollutant emissions) at the start and on this basis calculates how pollutants move and which chemical reactions they undergo in the atmosphere.

WRF-Chem model

Our forecasts are prepared using the Weather Research and Forecasting with Chemistry (WRF-Chem) model. This is Eulerian (see previous section) meteorological model integrated with the pollutant dispersion model developed by NOAA/ESRL (National Oceanic and Atmospheric Administration – Earth System Research Laboratory) in collaboration with many scientific institutions around the world.

One of the advantages of the WRF-Chem model is its large configuration possibilities that allow it to be tailored to your needs. Simulations using WRF and WRF-Chem models are performed all over the world for a variety of temporal and spatial scales as well as for many applications - for both past episodes analysis and short-term weather and air quality forecasting as well as long-term research.

The WRF-Chem model is integrated online, i.e. meteorological parameters are modelled simultaneously with pollutant dispersion. This means that in our simulations we can take into account the impact that pollutants exert on the weather, e.g. dust can affect the formation of clouds and the scattering of solar radiation. This approach often improves weather forecasts and pollutant concentrations.

Input data

In order to run the WRF-Chem model, the state of the atmosphere is required at the initial moment of the model working - inside the domain, i.e. so-called. initial conditions and its borders, i.e. boundary conditions. On this basis, the model predicts further evolution of the system. Meteorological information is obtained from the global meteorological forecast -GFS (Global Forecasting Model).

Except for meteorological conditions, information on emission and concentrations of pollutants is also required. Anthropogenic emission data are described in the previous sections - for modelling, they are prepared in a suitable format, taking into account temporal variability and different emitter heights. Chemical boundary conditions for the external domain are prepared using the global transport model (MOZART), and for the internal domain they are inherited from the external domain.

Elaborated by K.Wałaszek