Evaluation with throughfall data

In order to evaluate the model, generated deposition esfimates are compared to deposition estimates derived from throughfall and stemflow measurements.

Throughfall and stemflow measurements have been found useful for independent evaluation of deposition modeis (e.g. Erisman and Draaijers, 1995; Spranger et ui, 1994) though a major limitation is that information to distinguish between in canopy and anthropogenic sources of chemical compounds often is lacking, thereby hampering accurate deposition estimation.Änother probiem when using throughfall and stemflow measurements for model evaluation is related to the spatial scale of the estimates. Throughfall and stemflow deposition estimates usually refer to a parficular forest stand, whereas model deposition estimates hold for an area approximately 10x20 km wide. For this reason the EDÄCS-model was run using base cafion precipitation concentrations measured at the sites and local information on z.

Results of throughfall, stemflow and bulk precipitation measurements were obtained from national organisations responsible for throughfall and stemflow monitoring in their countries (Table 4,1), In total, results from 174 throughfall measurement sites were gathered. The measurement sites are reasonably well distributed over Europe although sites in southern Europe are almost lacking.

The majority of the sites (71%) is situated in coniferous forest stands; only 17% in deciduous forest stands and 12% in mixed forest stands. Monitoring took place between 1987 and 1994 for a period of at least one year. Data for 1989 were used but, if not available, the average for the whole measurement period was computed and used for model evaluation. At most sites stemflow was not measured as it generally contributes only a small fraction of the total flux to the forest fioor. In that case, the stemflow fluxwascomputed as a percentage of the throughfall flux using a parametrisation on free species and stand age described by Ivens (1990).

Wet deposition was estimated from bulk precipitation measurements by applying bulk to wet-only correction factors presented by Van Leeuwen et ui (1995).

The Enneh Envwonment 27

0

Table 4.! Per country the number of throughfaN monitoring Iocations in coniferous, deciduous and mixed forest stands, respectively, used for model evaluation. The measurement period 15 also presented.

Country No. of sites Measurement period

Coniferous Deciduous MixeU

CzechRepubhc’)t) 6 990-1994

Denmark2 992

Finland2) 9 1989-1992

France3 18 8 1993

Germany4)2) 63 2 2 1989

Hungary1) 1990-1993

Netherlands5 22 2 4 1987-1991

Norwayt) 2 1990-1993

PoIand2 1993-1994

Russian Federation1 2 1990-1993

Sweden2 4 1989-1993

SwitzerIand6 12 5 1987-1992

data provided by the Czech Geologica] Surey, Prague T Paces andj.^Cern)

data from the CP-Integrated Monitonng Programme provded by the Finnish Envwonrnent nstisute, Helanki (S.

Kleemola and M. Forsius)

3) data provided by the Office National des Forts, Fontainebleau tE. Ulnch)

4) data provided by the University of Stuttgart QTh. Gauger, R. Köble and G. Smotek)

5) data availahle at the National Instftute of Publjc Heafth and the Environment, B lthoven (G. Draaijers and jW Ensman)

6) data provded by the nstitut fiir Angewandte Pflanzenbioogy, Schönenbuch (S. Braun)

A canopy budget model developed by Ulrich (1983) was used to estimate the impact of canopy leaching on throughfall and stemflow fluxes of Mg2, Ca2 and K. An extensive description and uncertainty analysis of the model is presented by Draaijers and Erisman (1995). In the model, Na is assumed not to be influenced by canopy exchange. Therefore dry deposition of Na can be calculated by subtracting wet deposition from the throughfall ÷ stemflow flux, Particles containing Mg2, Ca2 and K are assumed to have the same mass median diameter as Na containing particles. Dry deposition of Mg2, Ca2 and K can subsequently be calculated according to:

DD (TF+ SFN- BPN)/BP ^*BP [1]

where DD, TF, SF, and BP represent dry deposition, throughfall, stemflow and bulk precipitation flux, respectively and x denotes Mg2, Ca2 or K. In equation [11 (TFN,+SFN;BPN)/BPN, represents the so-called ‘dry deposition factor’. In this study annual mean throughfall, stemflow and bulk precipitation fluxes are used through which differences in dry deposition factor caused by seasonal changes in pollution climate and canopy characteristics are neglected. Ädditional error will arise by using bulk precipitation data instead of wet-only deposition. The assumption that Mg2, Ca2 and K containing particles are deposited with equal efficiency as Na* containing particles infroduces an error as the particle size distribution of these constituents is not necessarily the same (Milford and Davidson, 1985). There will tie a shift in the size disfribution towards particles with smaller radii with increasing distance to source areas and/or lower relative humidity (fitzgerald, 1975). Sulphur was assumed to behave inert in the canopy.

Canopy uptake of oxidised and reduced nitrogen compounds could not tie quantified due to lack of data. Dry deposition of sulphur and nitrogen compounds

J

he Finneh Environrnent 27

was therefore calculated according to:

DD ⁼TF+Sf-BP [2]

Dry deposition estimates derived this way were compared to modelled dry deposifion estimates for the 174 measurement sites. Moreover, a comparison is made for sites with data for 1989 and a comparison for solely ICP 1M sites. The EDACS-model was run using base cation precipitafion concentrations measured at the site and EMEP concentrations, together with local information on the roughness length(Z) for each forest stand. The latter has been found dependent on free height (h) and canopy closure (Jarvis et al, 1976). The following relationship was assumed (Wieringa, 1992):

z0=0.06*h [3]

If site-information on free height was not available, free height was indirectly derived using species-specific relationships between free height and tree age presented by Schober (1987), or taken equal to the average height of other forest stands situated in the region/country. The impact of canopy closure on the roughness length could not be taken into account due to lack of data.

Table 4.2 gives an overview of site characteristics of the ICP IM sites used for model evaluation. The average stand age equals about 80 years (and varies from 30 to over 100 years) whereas the average tree height amounts 1$.5m (and varies between 5.8 and 30.6m). Most sites are located in Northern Europe (Scandinavia)

Table 4.2 Overview of site characteristics of ICP 01 sites used in this study.

Code Location Longitude Latitude Tree species Stand age Tree height Period Cl0l Anenske Povodi 15°05’OO” E 49°35’OO” N Piceaexcelsa, Pinus sylvestris 70 24] 1990-1993 DEOI-031 Forellenbach 13°25’03” E 48°56’38” N Picea abies, tagus sylvatica 90 30.6 1991-1993 Dt0I-041 Forellenbach 13°25’03” E 48°56’38” N Fagus sylvatica, Picea abies 26.5 1991-1993 DKOI Hjerl Hede 0830’00” E 56°25’OO” N Quercus robur, Quercus petraea >50 992 F101-002 Valkea-Kotinen 25°03’41” E 6114’12” N Piceaabies, Betulaspp., Pinus sylvestris lIS 22.7 989 Fl0I-003 Valkea-Kotinen 2Y03’41” [ 6ldl4I2 N Betula spp., Picea abies, Pinus sylvestris 30.3 1989 FIOl-Ol 1 Valkea-Kotinen 25°03’41” E 61°14’12” N Picea abies, Betula spp., Populus tremula 1991-1992 F103-00I Hietajärvi 30°40’45” t 63°09’OS” N Pinus sylvestris, Betula spp., Picea abies >50 6.6 990-1992 F103-004 Hietajärvi 30°40’45” E 63°09’OS” N Pinus sylvestris, Betula spp. 23.3 1990-1992 F104-00I Pesosjärvi 29°26’32” E 6617’12” N Picea abies, Pinus sylvestris, Betula spp., >50 14.4 1990-1992

Populus tremula

F104-001 Pesosjärvi 29°26’32” E 6617’12” N Pinus sylvestris, Picea abies, Betula spp. 16.1 1990-1992 F105-002 Vuoskojärvi 26°5617” E 69°44’04” N Betula pubescens spp. tortuosa, Pinus sylvestris>50 5.8 1990-1992 F105-003 Vuoskojärui 26°56’17” E 69°44’04” N Pinus sylvestris, Betula pubescens spp, tortuosa 10,6 1990-1992 HUOI Mentelek 19’32’26” E 46°57’40” N Pinus sylvestris, Pinus nigra 30 1990-1993 N001 Brkenes 08°15’OO” £ 58°23’OO” N Picea abies e.a. 80 16.3 1990-1993 N002 Kårvatn 08D5300 £ 62°4?’OO” N Pinussylvestrise.a. >50 8.8 1990-1993 PLO4 Brenna I95600” E 49°40’OO” N Norwaysprue, Fagus sylvatica >100 1993-1994 RUO5 Juga Massif 40°28’OO” £ 43°53’OO” N Abies nor., Acer trab,, Betula lit., Salix cap., >100 15.9 990, 1991,

Sorb.auc. 1993

RUI5 Tayozhny Log 33°21’OO” £ 5?05I00 N Piceetum Oxalidosum >50 1990-1993 SEOI Tiveden 14°38’OO” £ 58°41’OO” N Betula pub., Piceaabies,Pinus sylvestris >100 18.6 1992-1993 5E02 Berg 12°48’OO” E 50°04’OO” N Fagus sylvatica, Picea abies, Pinus sylvestris, >100 26.4 1992-1993

Quercus petraea

5E03 Reivo 19°05’Ol” E 65°4?’OO” N Picea abies, Pinus sylvestris, Betula spp. >100 9 1992-1993 SEO4 Gårdsjön 11°0I’OO” E 58°03’OO” N Picea abies, Pinus sylvestris, Betula spp. 80 1? 1989

The Finnish Environment 27

0