An inventory of tree species in Europe—An essential data input for air pollution modelling
Introduction
The biosphere act as an important source of gasses and particles to the atmosphere. Detailed data characterising the biosphere is therefore needed as input for atmospheric models in order to assess the impact of these emissions correctly. For some of these emissions, the distribution of trees down to genus or species level is a must. The release of volatile organic compounds (VOCs) (Dindorf et al., 2006, Guenther et al., 1995, Karlik et al., 2002, Karlik and Winer, 2001, Simpson et al., 1999) is an important example of such emissions, since the distribution of VOC species and the emission intensity varies strongly from one type of tree to another as well as between other groups of plants. The same picture applies e.g. to the emission of pollen. These differences are important since emissions of VOCs and pollen have great impact on both environment and health.
The biogenic VOCs are highly reactive, and together with NOx these compounds play an important role in the formation of tropospheric ozone and secondary organic aerosols (SOA) (Andreae and Crutzen, 1997, Atkinson, 2000, Gelencser et al., 2007, Kanakidou et al., 2005, Szidat et al., 2006, Went, 1960). The emission of biogenic VOCs has therefore far-reaching effects in relation to air quality and effects on human health, plants as well as the global climate. Biogenic VOC emissions depends strongly on temperature and light variations (Dindorf et al., 2006, Guenther et al., 2006). Short term changes in meteorological conditions may therefore increase or decrease emissions of biogenic VOCs in one area such as southern Europe, which in turn affect formation and transport of ozone to northern Europe. Regional tree species distributions are therefore important with respect to biogenic VOC emissions and long range transported air pollutants such as ozone and SOA. The large temporal variations in emissions suggest the use of dynamical emission models. Such models require tree species information as a main input parameter. It has been shown that dynamic emission calculations of other air pollutants such as ammonia improves the results obtained with atmospheric transport models (Ambelas Skjøth et al., 2004). It is therefore likely, that studies of air pollutants such as ozone or SOA using chemical transport models (CTMs) will benefit from dynamic emission calculations of biogenic VOCs using tree inventories on species level.
Atmospheric transport models can be extended to handle atmospheric dispersal of pollen (Helbig et al., 2004, Pasken and Pietrowicz, 2005, Sofiev et al., 2006). Similar to CTMs, pollen dispersal models are strongly dependent on the quality of the pollen emissions. As for VOCs, pollen emission changes heavily with variations in meteorology such as temperature and relative humidity (Bianchi et al., 1959, Laaidi et al., 2003, Ogden et al., 1969, Rempe, 1938). Similar to chemical air pollutants, the overall pollen concentration at a specific location may be due to local sources (Corden et al., 2000, Skjøth et al., 2008) or long range atmospheric transport (Hjelmroos, 1991, Hjelmroos, 1992, Skjøth et al., 2007, Smith et al., 2008). Furthermore, pollen production from trees such as Betula is known to vary significantly from year to year (Latalowa et al., 2002, Ranta et al., 2006, Sommer and Rasmussen, 2006). Tree pollen produced one year is released the following spring. Dynamic estimates of pollen emission should therefore be based on models that take the meteorological conditions during pollen production and release into account and a specific tree inventory.
Thus, atmospheric transport models used for chemical air pollution or pollen dispersal studies will benefit from dynamical emission models using high resolution tree inventories on species level.
Forest inventories suitable for biogenic VOC emission and pollen emission calculations are currently available at a national (Simpson et al., 1999) or sub-national level (Skjøth et al., 2007, Skjøth et al., 2008) for several European countries. These forest inventories are non-uniform as they are based on different national inventory strategies. As an example: neighbouring countries account important species such as Betula separately (e.g. Sweden or Lithuania) or in groups with two or more species (e.g. Denmark and Poland). However, most European countries have a forest inventory suitable for estimates of tree species specific distributions. Therefore a GIS framework may be used to harmonize data that is based on uneven sampling strategies or forest inventories. This harmonized data set may be redistributed to the EMEP50 grid (http://www.emep.int/grid/griddescr.html).
Unevenly distributed forest inventories and tree statistics are aggregated on national and sub-national level using a GIS framework. The inventory is redistributed into commonly applied grid definitions such as the EMEP50 grid. The final product will give a tree species distribution ready to be used as input to atmospheric transport models for biogenic VOC or pollen dispersal studies.
Section snippets
Methodology
Statistical data on tree species composition was obtained from each country within the EMEP50 area domain. Five different sources for data are considered appropriate: scientific papers, the national forest inventories, the national statistical offices, the European Forest Institute and the national reports used in the Global Forest Resources Assessment (FAO, 2006). We have selected the inventory with the largest number of species and largest number of regions within each country (Table 1). For
Results
The tree distribution inventory covers 799 regions in Europe, parts of Asia and parts of Africa. As seen on Fig. 1 the regions range from country level to municipality level. The inventory include 16 different conifer groups on genus or species level: P. sylvestris, Pinus halepensis, Pinus nigra, Pinus pinaster, Pinus sp. (undefined Pinus species), Picea abies, Picea sitchensis, Picea sp. (undefined Picea species), Pseudotsuga, Abies alba, Abies procera, Abies normanniana, Abies sp. (undefined
Uncertainty and comparison with other results
Accuracy and hence uncertainty of this dataset varies within each of the 799 regions and within the countries. The uncertainty can be grouped into three categories where the first is related to the applied inventory method, the second to the number of species categories (including the individual species and species not accounted for) and the third to the geographical size of each of the 799 regions and regions with limited or no statistics.
Discussion and conclusion
The unevenly distributed forest inventories and tree statistics from 799 geographical regions are used as the basis for a new detailed tree species inventory. We have combined and redistributed the data into the EMEP50 grid. The gridded dataset is ready to be used for input to atmospheric transport models for biogenic emissions calculations such as VOCs or pollen dispersal studies.
The tree species distribution includes important species such as Betula sp. and Alnus sp. relevant for pollen
Outlook and perspectives
The results presented here address one of the main scientific challenges described in the COST Action ES0603 (EUPOL to take place during the period 2007–2010) (http://www.cost.esf.org/index.php?id=1080). EUPOL is divided into three work packages (WP), where WP 1 concerns pollen production and release. In WP 1 a main task is the mapping of the distribution areas of the important allergenic pollen types. This inventory considers tree species distribution including important allergenic tree
Acknowledgements
This work was partly funded by the Copenhagen Global Change Initiative (www.cogci.dk) and Centre for Energy, Environment and Health (CEEH) (ref. 2104-06-0027) funded by the Danish Research Academy.
References (87)
Atmospheric chemistry of VOCs and NOx
Atmos. Environ.
(2000)- et al.
Modelling the spatial distribution of agricultural ammonia emissions in the UK
Environ. Pollut.
(1998) - et al.
Development of a high-resolution nested air pollution model—the numerical approach
J. Comput. Phys.
(2002) - et al.
Assessment of the atmospheric nitrogen and sulphur inputs into the north sea using a Lagrangian model
Phys. Chem. Earth
(2002) - et al.
Measured isoprene emission rates of plants in California landscapes: comparison to estimates from taxonomic relationships
Atmos. Environ.
(2001) - et al.
A survey of California plant species with a portable VOC analyzer for biogenic emission inventory development
Atmos. Environ.
(2002) - et al.
Ragweed in France: an invasive plant and its allergenic pollen
Ann.Allergy Asthma Immunol.
(2003) - et al.
Global air emission inventories for anthropogenic sources of NOx, NH3 and N2O in 1990
Environ. Pollut.
(1998) - et al.
Using dispersion and mesoscale meteorological models to forecast pollen concentrations
Atmos. Environ.
(2005) - et al.
Compilation of a European forest map from Portugal to the Ural mountains based on earth observation data and forest statistics
Forest Policy Econ.
(2003)
Implementing a dynamical ammonia emission parameterization in the large-scale air pollution model ACDEP
J. Geophys. Res. (Atmos.)
Atmospheric aerosols: biogeochemical sources and role in atmospheric chemistry
Science
A new SPOT4-VEGETATION derived land cover map of northern Eurasia
Int. J. Remote Sensing
Pollen release in the common ragweed (Ambrosia artemisiifolia)
Bot. Gazette
UK regional variations in Betula pollen (1993–1997)
Aerobiologia
European allergenic pollen types
Aerobiologia
Significant light and temperature dependent monoterpene emissions from European beech (Fagus sylvatica L.) and their potential impact on the European volatile organic compound budget
J. Geophys. Res. (Atmos.)
Source apportionment of PM2.5 organic aerosol over Europe: primary/secondary, natural/anthropogenic, and fossil/biogenic origin
J. Geophys. Res. Atmos.
A global-model of natural volatile organic-compound emissions
J. Geophys. Res. (Atmos.)
Estimates of global terrestrial isoprene emissions using MEGAN (model of emissions of gases and aerosols from nature)
Atmos. Chem. Phys.
Numerical modelling of pollen dispersion on the regional scale
Aerobiologia
Evidence of long-distance transport of Betula pollen
Grana
Long-distance transport of Betula pollen grains and allergic symptoms
Aerobiologia
Cited by (67)
Outdoor airborne allergens: Characterization, behavior and monitoring in Europe
2023, Science of the Total EnvironmentAerobiological modelling II: A review of long-range transport models
2022, Science of the Total EnvironmentCitation Excerpt :However; these inventories are non-uniform on the scale needed on the long-range transport modelling of trans-national scale (Sofiev et al., 2006a; Skjøth et al., 2008a). Therefore, a GIS framework may be used to harmonize data that is based on uneven forest inventories; but, for this, it is necessary to take into account the uncertainty resulted from the method, the number of species of the same taxon and the area covered by the inventory (Skjøth et al., 2008a). Usually, this type of work methodology is classified as bottom-up according to Skjøth et al. (2013).
Spatial and temporal variations in the distribution of birch trees and airborne Betula pollen in Ireland
2021, Agricultural and Forest MeteorologyTree cover mapping based on Sentinel-2 images demonstrate high thematic accuracy in Europe
2020, International Journal of Applied Earth Observation and GeoinformationOak pollen seasonality and severity across Europe and modelling the season start using a generalized phenological model
2019, Science of the Total EnvironmentA study on the spatial and temporal variability in airborne Betula pollen concentration in five cities in Poland using multivariate analyses
2019, Science of the Total Environment