explanations

Background

The creation and installation of this operational phenological model was motivated by the need of the pollen forecast, to have some information of the current developmental stage of the pollen producing plants over a larger area. Back trajectories (path of an air parcel towards the receptor point) were added to the maps (current day and next day) in order to help the forecaster to assess the transport probability air masses from areas, where flowering and thus pollen release is just occurring.

Apart from the pollenforecast the maps might be of interest for the general public also.

The following 14 phases were selected:

 

Number

Scientific name

Common name

Phase

Allergenic relevance

1

Acer platanoides

Norway mapple

First flowers open

low

2

Aesculus hippocastanum

Horse chestnut

First flowers open

low

3

Alnus glutinosa

Black alder

Beginning of flowering

yes

4

Alopecurus pratensis

Foxtail grass

Beginning of flowering

yes

5

Artemisia vulgaris

Mugwort

Beginning of flowering

yes

6

Betula pendula

Silver Birch

Beginning of flowering

yes

7

Corylus avellana

Common hazel

Beginning of flowering

yes

8

Forsythia suspensa

Forsythia

First flowers open

 

9

Fraxinus excelsior

Common ash

First flowers open

yes

10

Salix caprea

Goat willow

First flowers open

yes

11

Sambucus nigra

Elder

First flowers open

low

12

Syringa vulgaris

Lilac

First flowers open

 

13

Tilia cordata

Small leaved lime

First flowers open

low

14

Secale cereale (winter)

Rye

First flowers open

yes

 

The phenological model

A 3 parameter phenological temperature sum model (e.g. Chuine at al., 2003) was fitted on basis of historical phenological observations collected during COST725 (European Cooperation in Science and Technology, Action 725) and the ECSN (European Climate Support Network) time series of daily mean temperatures on a 0.25° grid.

For the daily operational procedure the 10 day temperature forecast of the ECMWF (European Centre for Medium-Range Forecasts) is used as input for the phenological model. The calculation of the entry dates is done on a grid with a 0.25° resolution and the results are then interpolated to a 1 km grid.

Following restrictions have to be considered:

  • The natural distribution of the species has not been taken in to account. There might be areas indicating entry dates, where the plants do not occur naturally.
  • The phenological entry dates are calculated for areas with elevations below 1500 m.
  • At the moment no real time observations of phenological phases are taken into account. There exist a few real time observing networks in Europe, which provide such data (for instance http://zacost.zamg.ac.at/phaeno_portal/, http://www.natuurkalender.nl or http://www.naturescalendar.org.uk/ ).

Description of the phenological maps

There are 5 maps displayed for each flowering phase. The first map in the first column to the left shows the absolute entry date in yeardays (counting starts with 1st January with day 1 and ends with day 365 resp. 366 on 31st December), calculated 9 days ahead. The following 4 maps display the probability of entry dates having occurred in each grid element in 5 percent intervals, 0 – 25%, 25-50%, 50 – 75%, 75 – 100% and 100% for the current day and as 24, 48 and 72 hour forecast. The probabilistic representation has been chosen because of the large variation of the observed phenological entry dates within a small area. Differences in genetic disposition, variations in slope and exposition, differences in soil features, observer errors and other factors (which are not considered in the phenlogical model) contribute to that large variation of the observed phenological entry dates.

 

Any suggestions, observations and comments are welcome (Helfried.Scheifinger(at)zamg.ac.at)

 

Chuine, I., K. Kramer and H. Hänninen (2003): Plant development models. In: Schwartz, M. D., (Ed., 2003): PHENOLOGY: An Integrative Environmental Science. Kluwer Academic Publishers, Dordrecht/Boston/London, 564 pp.