HIRLAM Birch Pollen Modelling for Denmark · 2016. 5. 26. · Enviro‐HIRLAM Birch Pollen...
1
Enviro‐HIRLAM Birch Pollen Modelling for Denmark ALEXANDER KURGANSKIY 1,2* , ROMAN NUTERMAN 1,3 , ALEXANDER MAHURA 4 , ALEXANDER BAKLANOV 4,5 , ALIX RASMUSSEN 4 , SERGEJ SMYSHLYAEV 2 , EIGIL KAAS 1 , BENT H. SASS 4 1 Niels Bohr Institute, University of Copenhagen (NBI UC), Juliane Maries Vej 30, DK‐2100, Copenhagen, Denmark. 2 Russian State Hydrometeorological University (RSHU), Malookhtinsky prospect 98, St. Petersburg, Russia. 3 Tomsk State University (TSU), Lenin Ave., 36, 634050, Tomsk, Russia. 4 Danish Meteorological Institute (DMI), Lyngbyvej 100, DK‐2100, Copenhagen, Denmark. 5 World Meteorological Organization (WMO), 7 bis, Avenue de la Paix, 1211 Geneva 2, Switzerland. ENVIRO‐HIRLAM‐POLLEN Birch pollen emission is strongly meteorology dependent, and it is essential to forecast pollen episodes by coupled meteorology ‐ air pollution models. The Enviro‐HIRLAM (Environment ‐ High Resolution Limited Area Model; Korsholm, 2009, Baklanov et al., 2008, Nuterman et al., 2013) online‐coupled meteorology‐chemistry model was applied in the study. The main model components are presented in Fig. 2. ACKNOWLEDGEMENTS The authors are thankful to Dr. Carsten Ambelas Skjøth (University of Worcester, National Pollen and Aerobiological Research Unit; UK) for providing tree species inventory data; European Forest Institute (EFI) ‐ for broadleaved forest data; Danish Asthma Allergy Association for birch pollen observation data, Dr. Suleiman Mostamandy (RSHU, Saint Petersburg) for help in getting meteorological IC/BC data; COST Action ES1004 for financial support. MODEL VS. OBSERVATIONS REGION OF INTEREST AND INPUT DATA The model runs were performed to simulate birch pollen concentrations for European domain (154 x 148 grid horizontal points; 40 vertical hybrid levels) at 0.15° horizontal resolution with focus on Denmark (shown in Fig. 1a). In order to assess pollen sources the birch forest map (see Fig. 1b) has been derived for the selected modelling domain. The map is represented by birch forest fraction in each model grid cell. Three GIS based databases were used in deriving procedure: 1) Global Land Cover (http://landcover.usgs.gov/glcc/), 2) European Forest Institute (Päivinen et al. 2001) and 3) Tree Species Inventory (Skjøth et al., 2008). CONCLUDING REMARKS Preliminary evaluation for both modelled and observed birch pollen concentrations (see Fig. 4) showed high values (daily averages more than 100 grains/m 3 ) for Copenhagen and Viborg stations, and especially during 2‐12 May 2006 episode. Phenological evidence in combination with observed concentrations for both Danish measurement sites for the period of 15 – 30 April 2006 showed that long range transport did dominate the pollen concentrations. This is supported by the modelling. Underestimation of modelled concentrations, especially for the occurrence of maximum concentration for Copenhagen measurement site, was caused by underestimation of modelled air temperature and overestimation of modelled relative humidity at the beginning of the pollen season. It is suggested that further improvement of modelled pollen concentrations can be achieved through improvement of modelled meteorological parameters using Enviro‐ HIRLAM version taking into account urban effects. Figure 4. Birch pollen concentrations observed (red) vs. modelled (green) at Danish measurement sites: Copenhagen (55°43‘ N and 12°34‘ E) (fig. 4a) and Viborg (56°27‘ N and 9°24‘ E) (fig. 4b). REFERENCES Baklanov, A., Korsholm, U., Mahura, A., Petersen, C., and Gross, A.: ENVIRO‐HIRLAM: on‐line coupled modelling of urban meteorology and air pollution, Adv. Sci. Res., 2, 41‐46, 2008. Korsholm U.S.: Integrated modelling of aerosol indirect effects ‐ development and application of a chemical weather model, PhD thesis University of Copenhagen, Niels Bohr Institute and DMI, Research Department; http://www.dmi.dk/dmi/sr09‐01.pdf , 2009. Nuterman, R., Korsholm, U., Zakey, A., Nielsen, K. P., Sørensen, B., Mahura, A., Rasmussen, A., Mažeikis, A., Gonzalez‐Aparicio, I., Morozova, E., Sass, B. H., Kaas, E., and Baklanov, A.: New developments in Enviro‐HIRLAM online integrated modelling system, Geophysical Research Abstracts, Vol. 15, EGU2013‐12520‐1, 2013. Päivinen, R.,Lehikoinen, M., Schuck, A., Häme, T., Väätäinen, S., Kennedy, P., and Folving, S.: Combining Earth Observation Data and Forest Statistics, Research Report, 14, 101 p., EFI, Joensuu and Joint Research Centre / European Commission, 2001. Skjøth, C. A., Geels, C., Hvidberg, M., Hertel, O., Brandt, J., Frohn, L. M., Hansen, K. M., Hedegaard, G. B., Christensen, J., and Moseholm, L.: An inventory of tree species in Europe ‐ an essential data input for air pollution modelling, Ecol. Model., 217, 292‐304, 2008. Figure 2. Overall structure of Enviro‐HIRLAM birch pollen forecasting system. Figure 1. Modelling domain (1a) and spatial distribution of birch forest fraction (1b). *CONTACT Alexander Kurganskiy, email: [email protected] 4a 4b 1a 1b MODELLED POLLEN CONCENTRATIONS 3b 3a 3c 3d Figure 3. Birch pollen concentration (in grains/m 3 ) simulated by Enviro‐HIRLAM on 7 th May, 2006 for 06 UTC (3a), 12 UTC (3b), 18 UTC (3c) and 8 th May, 2006 for 0 UTC (3d). Forecast length – 6h with surface meteorological data assimilation only.
Transcript of HIRLAM Birch Pollen Modelling for Denmark · 2016. 5. 26. · Enviro‐HIRLAM Birch Pollen...
Enviro‐HIRLAM Birch Pollen Modelling for DenmarkALEXANDER KURGANSKIY1,2*, ROMAN NUTERMAN1,3, ALEXANDER MAHURA4, ALEXANDER BAKLANOV4,5,
ALIX RASMUSSEN4, SERGEJ SMYSHLYAEV2, EIGIL KAAS1, BENT H. SASS 4
1 Niels Bohr Institute, University of Copenhagen (NBI UC), Juliane Maries Vej 30, DK‐2100, Copenhagen, Denmark.
2 Russian State Hydrometeorological University (RSHU), Malookhtinsky prospect 98, St. Petersburg, Russia.
3 Tomsk State University (TSU), Lenin Ave., 36, 634050, Tomsk, Russia.
4 Danish Meteorological Institute (DMI), Lyngbyvej 100, DK‐2100, Copenhagen, Denmark.
5 World Meteorological Organization (WMO), 7 bis, Avenue de la Paix, 1211 Geneva 2, Switzerland.
ENVIRO‐HIRLAM‐POLLEN
Birch pollen emission is strongly meteorology dependent, and it is essential to forecastpollen episodes by coupled meteorology ‐ air pollution models. The Enviro‐HIRLAM(Environment ‐ High Resolution Limited Area Model; Korsholm, 2009, Baklanov et al.,2008, Nuterman et al., 2013) online‐coupled meteorology‐chemistry model was appliedin the study. The main model components are presented in Fig. 2.
ACKNOWLEDGEMENTS
The authors are thankful to Dr. Carsten Ambelas Skjøth (University of Worcester, NationalPollen and Aerobiological Research Unit; UK) for providing tree species inventory data;European Forest Institute (EFI) ‐ for broadleaved forest data; Danish Asthma AllergyAssociation for birch pollen observation data, Dr. Suleiman Mostamandy (RSHU, SaintPetersburg) for help in getting meteorological IC/BC data; COST Action ES1004 forfinancial support.
MODEL VS. OBSERVATIONS
REGION OF INTEREST AND INPUT DATA
The model runs were performed to simulate birch pollen concentrations for Europeandomain (154 x 148 grid horizontal points; 40 vertical hybrid levels) at 0.15° horizontalresolution with focus on Denmark (shown in Fig. 1a). In order to assess pollen sourcesthe birch forest map (see Fig. 1b) has been derived for the selected modelling domain.The map is represented by birch forest fraction in each model grid cell. Three GIS baseddatabases were used in deriving procedure: 1) Global Land Cover(http://landcover.usgs.gov/glcc/), 2) European Forest Institute (Päivinen et al. 2001) and3) Tree Species Inventory (Skjøth et al., 2008).
CONCLUDING REMARKS
Preliminary evaluation for both modelled and observed birch pollen concentrations(see Fig. 4) showed high values (daily averages more than 100 grains/m3) forCopenhagen and Viborg stations, and especially during 2‐12 May 2006 episode.
Phenological evidence in combination with observed concentrations for both Danishmeasurement sites for the period of 15 – 30 April 2006 showed that long rangetransport did dominate the pollen concentrations. This is supported by the modelling.
Underestimation of modelled concentrations, especially for the occurrence ofmaximum concentration for Copenhagen measurement site, was caused byunderestimation of modelled air temperature and overestimation of modelled relativehumidity at the beginning of the pollen season.
It is suggested that further improvement of modelled pollen concentrations can beachieved through improvement of modelled meteorological parameters using Enviro‐HIRLAM version taking into account urban effects.
Figure 4. Birch pollen concentrations observed (red) vs. modelled (green) at Danishmeasurement sites: Copenhagen (55°43‘ N and 12°34‘ E) (fig. 4a) and Viborg (56°27‘ Nand 9°24‘ E) (fig. 4b).
REFERENCESBaklanov, A., Korsholm, U., Mahura, A., Petersen, C., and Gross, A.: ENVIRO‐HIRLAM: on‐line coupled modelling of urban meteorologyand air pollution, Adv. Sci. Res., 2, 41‐46, 2008.Korsholm U.S.: Integrated modelling of aerosol indirect effects ‐ development and application of a chemical weather model, PhDthesis University of Copenhagen, Niels Bohr Institute and DMI, Research Department; http://www.dmi.dk/dmi/sr09‐01.pdf, 2009.Nuterman, R., Korsholm, U., Zakey, A., Nielsen, K. P., Sørensen, B., Mahura, A., Rasmussen, A., Mažeikis, A., Gonzalez‐Aparicio, I.,Morozova, E., Sass, B. H., Kaas, E., and Baklanov, A.: New developments in Enviro‐HIRLAM online integrated modelling system,Geophysical Research Abstracts, Vol. 15, EGU2013‐12520‐1, 2013.Päivinen, R., Lehikoinen, M., Schuck, A., Häme, T., Väätäinen, S., Kennedy, P., and Folving, S.: Combining Earth Observation Data andForest Statistics, Research Report, 14, 101 p., EFI, Joensuu and Joint Research Centre / European Commission, 2001.Skjøth, C. A., Geels, C., Hvidberg, M., Hertel, O., Brandt, J., Frohn, L. M., Hansen, K. M., Hedegaard, G. B., Christensen, J., andMoseholm, L.: An inventory of tree species in Europe ‐ an essential data input for air pollution modelling, Ecol. Model., 217, 292‐304,2008.
Figure 2. Overall structure of Enviro‐HIRLAM birch pollen forecasting system.Figure 1.Modelling domain (1a) and spatial distribution of birch forest fraction (1b).
*CONTACT
Alexander Kurganskiy, email: [email protected]
4a 4b
1a 1b
MODELLED POLLEN CONCENTRATIONS
3b3a
3c 3d
Figure 3. Birch pollen concentration (in grains/m3) simulated by Enviro‐HIRLAM on 7th
May, 2006 for 06 UTC (3a), 12 UTC (3b), 18 UTC (3c) and 8th May, 2006 for 0 UTC (3d).Forecast length – 6h with surface meteorological data assimilation only.
