72
A. Delcloo et al.
12.1 Introduction
12.1.1 Methods and Data
Pollen are biogenic aerosols with a diameter of typically 5–50 times larger than conventional atmospheric aerosols depending on vegetation type. Birch pollen can be
seen as a PM22 aerosol. Chemistry transport models such as SILAM (System for
Integrated modeLling of Atmospheric coMposition) are able to simulate the dispersion of pollen based on processes such as wind advection (transport with air masses),
mixing due to turbulence, gravitational settling (dry deposition), and scavenging with
precipitation (wet deposition) [6].
The simulation of birch pollen levels in the air requires the quantification of the
spatio-temporal emission sources of birch pollen at the surface in the model domain.
Stated otherwise, a map with the areal fraction of birch trees is highly necessary as
an underlying data set or input. At the European scale such a map was first compiled
by Sofiev et al. [5]. Another issue in modelling pollen levels is the evaluation of
pollen emission timing and intensity. Pollen modelling with SILAM is based on
the temperature degree days approach or the thermal time flowering model. The
parameterization of flowering follows a principle of two thresholds (start and end
of the flowering season) for the temperature sum [4], which assumes that the timing
of birch flowering is mostly driven by accumulated ambient temperature during a
certain time period. The cumulative fraction of pollen released from the beginning
of a year until a certain time is piecewise linear and proportional to the temperature
sum during the main flowering season. Short-term meteorological conditions such as
wind speed, relative humidity and precipitation rate will affect the amount of pollen
in the air. Precipitation and humidity suppress the pollen release and threshold values
are used to compute reduction factors. Typically, the lower and upper thresholds from
relative humidity are 50 and 80%. For precipitation the lower and upper thresholds
are 0 and 0.5 mm h
−1 (the grid cell average rate). At the saturation wind speed of
5 m s
−1 the pollen release rate is maximal. At a wind speed around 1 m s
−1 , no pollen
emissions occur.
12.1.2 Observational Data
ECMWF ERA-INTERIM meteorological data is used to drive the transport model.
Pollen levels simulated by SILAM are evaluated using observations of pollen data
taken from the Belgian Scientific Institute for Public Health (Sciensano). In 2008
SIPH monitored pollen concentrations at four stations (De Haan: 50.824523 N,
4.382457 E; Antwerp: 51.212683 N, 4.397888 E; Brussels: 50.824523 N, 4.382457
E; Charleroi: 50.408 N, 4.444 E) on a daily basis.
From Flemish and Walloon forest inventory data the relative diameter at breast
height (DBH) for birch trees are derived at each sampling plot (11.080 plots in
A. Delcloo et al.
12.1 Introduction
12.1.1 Methods and Data
Pollen are biogenic aerosols with a diameter of typically 5–50 times larger than conventional atmospheric aerosols depending on vegetation type. Birch pollen can be
seen as a PM22 aerosol. Chemistry transport models such as SILAM (System for
Integrated modeLling of Atmospheric coMposition) are able to simulate the dispersion of pollen based on processes such as wind advection (transport with air masses),
mixing due to turbulence, gravitational settling (dry deposition), and scavenging with
precipitation (wet deposition) [6].
The simulation of birch pollen levels in the air requires the quantification of the
spatio-temporal emission sources of birch pollen at the surface in the model domain.
Stated otherwise, a map with the areal fraction of birch trees is highly necessary as
an underlying data set or input. At the European scale such a map was first compiled
by Sofiev et al. [5]. Another issue in modelling pollen levels is the evaluation of
pollen emission timing and intensity. Pollen modelling with SILAM is based on
the temperature degree days approach or the thermal time flowering model. The
parameterization of flowering follows a principle of two thresholds (start and end
of the flowering season) for the temperature sum [4], which assumes that the timing
of birch flowering is mostly driven by accumulated ambient temperature during a
certain time period. The cumulative fraction of pollen released from the beginning
of a year until a certain time is piecewise linear and proportional to the temperature
sum during the main flowering season. Short-term meteorological conditions such as
wind speed, relative humidity and precipitation rate will affect the amount of pollen
in the air. Precipitation and humidity suppress the pollen release and threshold values
are used to compute reduction factors. Typically, the lower and upper thresholds from
relative humidity are 50 and 80%. For precipitation the lower and upper thresholds
are 0 and 0.5 mm h
−1 (the grid cell average rate). At the saturation wind speed of
5 m s
−1 the pollen release rate is maximal. At a wind speed around 1 m s
−1 , no pollen
emissions occur.
12.1.2 Observational Data
ECMWF ERA-INTERIM meteorological data is used to drive the transport model.
Pollen levels simulated by SILAM are evaluated using observations of pollen data
taken from the Belgian Scientific Institute for Public Health (Sciensano). In 2008
SIPH monitored pollen concentrations at four stations (De Haan: 50.824523 N,
4.382457 E; Antwerp: 51.212683 N, 4.397888 E; Brussels: 50.824523 N, 4.382457
E; Charleroi: 50.408 N, 4.444 E) on a daily basis.
From Flemish and Walloon forest inventory data the relative diameter at breast
height (DBH) for birch trees are derived at each sampling plot (11.080 plots in