Observations of ice impurities in some Finnish lakes/Soome jarvede jaas esinevate lisandite vaatlus.
Lake ice crystal lattices consist of the pure solid phase of [H.sub.2]O. The whole sheet of lake ice contains impurities in addition to the ice crystals: gas bubbles, liquid solution, and solid particles or sediments . The impurities account for all substances other than water, including also pollutants. They are stored in the ice sheet, possibly transported due to the drifting of ice, and re-introduced into the water body when the ice melts. Impurities have been studied to some extent in the Arctic sea ice (e.g., [2, 3]), the Baltic Sea ice [4, 5], and river ice  but not much in lake ice.
In Finnish lakes the ice season normally lasts from November to May . Impurities accumulate in the ice sheet for most of the ice season and are rejected into the water body during about a one-month melting period. The available literature on these impurities is very limited. Therefore, a field programme was performed in 1997-99 to map pH, dissolved matter, and sediments of lake ice, or rather of the meltwater of the ice, including analysis of metals and nutrients. The lakes are from southern Finland representing different lake types in terms of the water quality. This paper presents the results of the field investigations.
MATERIAL AND METHODS
There are three different types of layers in the vertical stratification of a lake ice sheet: congelation ice, frazil ice, and snow-ice [8, 9]. Thin ice crystal platelets form so-called macro crystals (linear dimensions from 0.1 mm up to 1 m), which optically behave as individual crystals and are normally referred to as such. A lake ice sheet initiates with a primary ice layer. On a calm surface it is very thin (less than 1 mm). Then congelation ice grows down from the ice bottom, and through a 0-5 cm transition zone the crystals become large, often columnar, and their size further increases with depth. In turbulent water, free frazil ice crystals (size 1 mm or less) appear first, and when the buoyancy of these crystals together overcomes the turbulence, they surface and form a solid ice sheet. Frazil ice may also form later in open water spots and be deposited to the ice bottom as fine-grained frazil layers; but in Finnish lakes such layers have not been found. Snow-ice is a mixture of (a) snow and lake water, (b) snow and its meltwater, or (c) snow and liquid precipitation. It grows upward on the top of the ice due to (a) flooding of the ice cover, (b) in melt-freeze cycles, or (c) via liquid precipitation, and the resulting crystals are small as frazil ice crystals. The forms (b) and (c) are also called superimposed ice.
In southern Finland the thickness of ice is at maximum in March, with the annual mean of 50 cm and standard deviation of 10 cm . Snow-ice thicknesses are typically 5-20 cm [9, 10]. The melting of ice takes about one month in spring. The water column is stable until the solar radiation penetrating through the ice begins to heat the water significantly. How the radiative heating and the resulting convection occur depends on the quality of the water via its optical properties.
Impurities are trapped in a lake ice sheet inside macro crystals at platelet boundaries and between macro crystals. The mechanisms are the following: (1) In congelation ice growth, the separation of the freezing water from dissolved substances and suspended matter is incomplete and pockets with liquid solution and sediment particles remain within the ice; (2) In particular floating particles and gas bubbles become captured rather than rejected by a growing ice sheet; (3) In turbulent supercooled water suspended particles serve as crystallization seeds for frazil ice or may become locked within frazil slush, later to become a part of the solid ice sheet; (4) Anchor ice, formed on the bottom of the water body from frazil ice in turbulent flow, may rise and by adhering to the ice cover bring bottom material into it; (5) Snow-ice formation brings impurities into the ice sheet from the snow cover and flooded lake water; and (6) Fallout from the atmosphere accumulates on the ice/snow cover. An ice sheet may contain pockets with liquid brine where the temperature corresponds to the freezing point of the brine, as is common in sea ice . However, as the salinity of the lakes in Finland is very low (<< [10.sup.-3]), the volume of the pockets containing liquid solution is very small and does not have any significant effect on the ice properties.
To map the quality and quantity of impurities in lake ice, a specific observation programme was performed in 1997-99 in Finland (Fig. 1; Tables 1 and 2). A pilot study preceded in 1996 . The research lakes were Paajarvi, Tuusulanjarvi, Vesijarvi, and Paijanne. The Porvoonjoki River was sampled for comparison. This small river (length about 100 km) is located in southern Finland and flows into the Gulf of Finland. The sampling concerned two river sites, depth 2-3 m: No. 5 in Pukkila, upper river agricultural area, and No. 6 in Porvoo not far from the river mouth. For comparison with nearby sea ice, earlier results of Lepparanta et al.  from the Gulf of Finland were used.
[FIGURE 1 OMITTED]
The ice was sampled using a drill and an ice saw, the snow cover was carefully shovelled away from the ice. The structure and stratigraphy of the samples were recorded at the site. The snow samples were collected with plastic plates, cutting uniformly across the depth of the snow cover. Acid-washed 10-litre plastic buckets with plastic lids were used to transport and store these samples. A half-litre water sample was taken at the ice site. The samples were stored in a cold store room (+ 4 [degrees]C), the ice and water samples thus becoming melted. The meltwater volumes were 2-5 L.
The ice and snow meltwater and water samples were analysed according to the Finnish standards. The amount of sediments was determined by evaporation; the organic matter in the sediments was estimated by loss on ignition (LOI) after filtering (pore size 0.3-0.6 [micro]m) according to [12, 13]. Electric conductivity and pH were measured from unfiltered samples according to [14, 15], and the amount of dissolved matter was measured from filtered sub-samples . Concentrations of certain chemical elements (Al, Ca, Cd, Cu, Fe, K, Mg, Mn, N[O.sub.2]N, N[O.sub.3]-N, tot-N, Na, Ni, P[O.sub.4]-P, tot-P, Pb, and Zn) were determined from ice- and snow-bound sediments. Metal concentrations were measured as acid soluble after wet digestion according to  using flame [16-20] and flameless [21, 22] atomic absorption with a Varian SpetrAA A10+ atomic absorption spectrophotometer and Varian GTA-95 graphite furnace. The concentrations of P[O.sub.4] and total phosphorus were measured photometrically by means of the molybdenum blue method with ascorbic acid reduction . The sum of N[O.sub.2] and N[O.sub.3] and total nitrogen, after wet digestion , were measured according to . The sulphate concentrations were measured according to . All non-metal analyses were made with a DR Lange LS 500 UV/VIS spectrometer.
Ice thickness and structure
The ice sheet was stable fast ice in all sites. However, in 1998 ridge formation took place in the southern basin (Asikkalanselka) of Lake Paijanne. Our sampling site was on the northern side of Pulkkilanharju esker. On the southern side of this esker, there was one floating ice ridge with sail height of 20-30 cm (Fig. 2). The ice sheet was likely weaker there because of water currents. Also on the shores of this basin ice pile-up was noted in many places. Ridging is very rare in the lakes in Finland because of limited fetch lengths. If more common, it would change the question of ice impurities by capturing water inside consolidating ridges and by bottom scouring.
[FIGURE 2 OMITTED]
The thermodynamic growth of lake ice thickness h is described in the first order by a modified Stefan's law (e.g. ):
h(t) = [lambda]a[square root of S(t), (1)
where [lambda] is a reduction factor due to snow cover (0.5 < [lambda] < 1), [approximately equal to] a 3.3 cm/[([degrees]C day).sup.1/2] is the growth factor for bare ice, and S is the sum of negative degree-days or the time integral of the below zero temperatures. The thickness of ice in the lakes was 32-40 cm in 1997, 52-63 cm in 1998, and 42-59 cm in 1999 (Table 2). The averages were 45-50 cm for the individual lakes and 37-56 cm for the individual years. The mean winter (December-February) temperature was -5.8 [degrees]C in 1997, -4.6 [degrees]C in 1998, and - 6.7 [degrees]C in 1999, all a little above the normal (-7.4 [degrees]C), and the ice thickness was within one standard deviation from the average. In the normal case the freezing-degree-days sum to 700-800 [degrees]C day, which gives [lambda] ~ 2/3.
The snow-ice thickness was 1-23 cm, 10% of the total ice thickness in 1997, 15% in 1998, and 10-40% in 1999. In 1999 the snow thickness on land was exceptionally large, and the snow-ice thickness was 23 cm in sites Nos. 1/99 and 4/99, but only 3 cm in sites Nos. 2/99 and 3/99 where the ice sheet could resist the snow load. In fact, due to the pressure in the water, there was a strong flooding event when the drill hole was made; as a result, in site No. 2/99 the water level became 12 cm above the ice surface.
The ice was bare in 1997, in 1998 the snow thickness was 5 cm in all sites, and there was 5-31 cm snow in 1999. The snow density was 0.22-0.34 g/[cm.sup.3] in 1998 while in 1999 very large values resulted as much of the snow was wet slush. River ice samples were thinner, 21-46 cm, and they showed a considerably larger portion of opaque top layer.
Dissolved matter, sediments, and pH
A property of ice, [P.sub.I], such as concentration of dissolved matter, is a combination of contributions from congelation ice (CI) and snow-ice, which further consists of snow (S) and lake water (W):
[P.sub.I] = [gamma] [P.sub.CI] + (1 - [gamma])[[alpha] [P.sub.S] + (1 - [alpha])[P.sub.W]], (2)
where [gamma] is the proportion of congelation ice and a is the proportion of snow in snow-ice. Here the former is observed and the latter is assumed equal to 0.5 according to . In an average sense, [gamma] [approximately equal to] 0.8, and then an ice sample consists of 80% of congelation ice formed from the lake water with rejecting most impurities, 10% of snow, and 10% of lake water with its impurities captured. Thus snowfall and snow-ice formation add much impurities into the ice, and with the given [gamma] and [alpha] applied for the present measurements, the contribution would be of the order of 50%. Dividing Eq. 2 with [P.sub.W] and [P.sub.S], we obtain the concentration ratios
[P.sub.I]/[P.sub.W] = (1 - [gamma])(1 - [alpha]) + [gamma] [P.sub.CI]/[P.sub.W] + (1 - [gamma]) [alpha] [P.sub.S]/[P.sub.W] (3a)
[P.sub.I]/[P.sub.S] = (1 - [gamma]) [alpha] + (1 - [alpha]) (1- [gamma] [P.sub.W]/[P.sub.S] + [gamma] [P.sub.CI]/[P.sub.S] (3b)
The first terms in the right-hand side are [approximately equal to] 0.1, which is a lower limit of the ratios. The values on snow vary according to the atmospheric conditions during snowfall.
Table 3 shows the statistics for the bulk properties of the ice, snow, and water samples in the lakes. They include the electric conductivity, the concentration of dissolved matter, pH, and the concentration of sediments with the organic proportion specified. Variations exist between lakes. This is expected since they represent different lake types. There are also large interannual variations for each individual lake. The ice sediments were sometimes seen well by bare eyes as shown in Fig. 3 as an example.
The mean conductivity of the lake ice samples was 7-15 [micro]S/cm, with the range of 4-28 [micro]S/cm (at the reference temperature of 25 [degrees]C). The conductivity of snow averaged 9.5-28 [micro]S/cm, being less than the conductivity of ice only in hypereutrophic Lake Tuusulanjarvi. The year 1998 gave the lowest values for the ice (4-7 [micro]S/cm) and snow (10-12 [micro]S/cm) in all lakes. The difference in the conductivity of ice and water was one order of magnitude, the water samples showing average conductivities of 79-208 [micro]S/cm.
The concentration of dissolved matter behaved much like the conductivity. It was on average 14.3 mg/L in Lake Paajarvi, 11.3 mg/L in Lake Paijanne, 12.7 mg/L in Lake Vesijarvi, and 17.3 mg/L in Lake Tuusulanjarvi. The minimum was 5 mg/L and the maximum was 28 mg/L, both in Lake Tuusulanjarvi. The average values in snow were a little higher except in Lake Tuusulanjarvi, being respectively, 15.0, 20.0, 23.5, and 17.0 mg/L for these lakes. The concentrations in water samples averaged 31-143 mg/L. For the water samples the concentration of dissolved matter (in mg/L) was about 0.5 times the conductivity (in [micro]S/cm) as it should be but for ice and snow meltwater the ratios were larger, for ice always more than one. This fact reflects changes in the dissolved matter-suspended matter proportions in the melt-freeze cycles (e.g. ).
The ice was thus much cleaner than the water: the ratio of the average dissolved matter (DM) in ice to average dissolved matter in water was 0.22 in Lake Paajarvi, 0.36 in Lake Paijanne, 0.24 in Lake Vesijarvi, and 0.12 in Lake Tuusulanjarvi. The coefficient of variation was 0.3-0.6 for ice, 0.3-0.9 for snow, and 0.5-1.0 for water. Considering Eq. 3a, the ratio [DM.sub.CI]/[DM.sub.W] is at most ~ 0.1 since the last term is very small. On the other hand, Eq. 3b tells that [DM.sub.CI] ~ [DM.sub.S] to make the ratio [DM.sub.I]/[DM.sub.S] close to one.
[FIGURE 3 OMITTED]
Data for pH were obtained only in 1997 and 1999. The averages of ice and water pH were 6.6-7.0, ice pH being greater only in Lake Paajarvi. In 1997 the values ranged from 5.9 to 6.8, and in 1999 the level was much higher, 6.7-7.3. The snow data are not given in Table 3 since snow data are only from 1999; then the snow pH was 0.2 units lower than the pH of water and ice. In 1976-81 snow samples (286) were studied in Finland and lower pH values were obtained: range 3.7-6.2 with a median of 4.6 .
The sediment content or suspended matter (SM) was roughly the same in ice and water. The average levels in ice were 2.1 mg/L in Lake Paajarvi, 2.1 mg/L in Lake Paijanne, 2.0 mg/L in Lake Vesijarvi, and 12.6 mg/L in Lake Tuusulanjarvi. The minimum was 0.7 mg/L (Lake Paajarvi 1997) and the maximum 20.8 mg/L (Lake Tuusulanjarvi 1998). The water samples showed averages of 3.7, 1.4, 1.1, and 11.5 mg/L. In the averages, the ratios of the sediments in ice to the suspended matter in water were in these lakes 0.57, 1.50, 1.82, and 1.10, respectively. The snow cover had higher sediment concentration levels in Lake Tuusulanjarvi where the level was 11.6 mg/L, just 1 mg/L lower than in ice. The coefficient of variation was 0.1-0.7 for ice, 0.8-1.0 for snow, and 0.4-1.0 for water. The mean proportion of organic matter in the ice sediments was 23-36%. In the snow the range was 38-58% and in the water it was 22-50%. Considering Eqs. 3, we must have [SM.sub.CI] ~ [SM.sub.W] and [SM.sub.CI] < [SM.sub.S]. Consequently, suspended matter is captured from the lake water much more efficiently than dissolved matter.
The river ice samples showed higher levels of impurities. The mean conductivity for the ice was 17.0 [micro]S/cm in the upper river, and 19.0 [micro]S/cm in Porvoo, and the mean dissolved matter contents were 24.0 and 19.0 mg/L, respectively. Both values are greater than in the lake data, and in 1999 the values were much higher than in the other years. The mean conductivity was about an order of magnitude larger in the river water. The ice-water ratio for the average dissolved matter content was 0.19 in the upper river and 0.14 in Porvoo. The pH was lower in the river ice samples than in the lake water samples in 1997, but in 1999 the opposite result was obtained. The sediment content was the highest in the ice. The maximum (526.0 mg/L) was registered in the upper river in 1997. The mean levels were much larger than in the lake data.
In lakes Paajarvi, Vesijarvi, and Paijanne the metal concentration levels in the ice were around 100 [micro]g/L for aluminium and iron, 30 [micro]g/L for zinc, 6-9 [micro]g/L for copper, and 1-7 [micro]g/L for manganese and nickel (Table 4). The cadmium level was high (0.5 [micro]g/L) in Lake Paajarvi. Hypereutrophic Lake Tuusulanjarvi showed much higher levels for aluminium and iron (nearly 1000 [micro]g/L) than the other lakes. River ice possessed much higher metal concentrations, as expected, but still comparable with the levels in the water. Exceptionally high values in ice relative to water were obtained for copper in both river sites.
The concentrations of metals in the ice relative to those in the water were within 0.5-2.0. Exceptions were cadmium (> 5) and manganese (0.2) in Lake Paajarvi, nickel (0.25) in Lake Vesijarvi, and manganese (0.2) in Lake Tuusulanjarvi. Comparing snow with ice, a low cadmium level was seen again in Lake Paajarvi. In Lake Vesijarvi high aluminium, iron, manganese, and nickel concentrations were observed in the snow. In Lake Tuusulanjarvi the snow was poor in manganese. Snowfall is therefore a significant source for many metals. The maximum was found usually in the water except cadmium and copper for which the highest levels appeared in the ice.
The concentrations of heavy metals, except iron, in snow, ice, and lake water were quite small compared to concentrations in river and ground waters [30, 31]. Iron concentrations were high and variation between sampling sites was large. The reason for this can be redox conditions during winter and support from the catchment. Samples from Lake Tuusulanjarvi and Pukkila clearly show the effects of agriculture on lake and river water. Aluminium concentrations are also high in the samples from Lake Tuusulanjarvi and Pukkila. These values reflect the same source as for iron.
Except in Lake Tuusulanjarvi, nutrient levels in lake ice were usually the following (Table 5): total phosphorus 2 [micro]g/L, total nitrogen 200 [micro]g/L, sulphates 1 mg/L, chloride 1 mg/L, calcium 2 mg/L, magnesium 0.1-0.9 mg/L, potassium 0.5 mg/L, and sodium 1 mg/L. In hypereutrophic Lake Tuusulanjarvi the phosphorus and nitrogen levels were somewhat larger. The levels of nutrients in ice were much lower (by one order of magnitude) than in water. Exceptions were found only in Lake Paijanne, an oligotrophic water basin: the levels of phosphorus, calcium, magnesium, potassium, and sodium were about the same in the ice and the water. The concentrations in snow were usually close to the ice values. Nitrogen showed much higher snow levels in lakes Paajarvi, Paijanne, and Vesijarvi, while the levels of phosphorus, sulphates, and magnesium in the snow were higher in Lake Vesijarvi.
According to different classifications [32, 33] the total phosphorus and total nitrogen values from Lake Paajarvi and Lake Paijanne reflect mainly oligotrophy and mesotrophy. Some of the total nitrogen values are typical of eutrophy. The proportion of soluble N[O.sub.2] and N[O.sub.3] varies from 2% to 68% of the total nitrogen. Lakes Vesijarvi and Tuusulanjarvi and the Porvoonjoki River are clearly eutrophic. Some values from the Porvoonjoki can be regarded as indicating hypereutrophic conditions. Total nitrogen concentrations also reach values characteristic of hypereutrophic conditions. Sulphate and chloride concentrations are in the range typical of river waters .
Information on the vertical stratification is available for lakes Paajarvi and Tuusulanjarvi from an optics field campaign on 26-27 March 1996. The measurements were made of conductivity, concentration of yellow substance, chlorophyll a, and sediments for 10-cm vertical resolution, the volume of each sample being about one litre. Yellow substance, a concept used in optics of natural water bodies, consists of dissolved organic matter, which absorbs light in short wavelengths. The concentration of yellow substance was obtained from light beam attenuation measurements, that of sediments by filtering (pore size 0.45 [micro]m), and of chlorophyll a by the filter pad technique.
The vertical stratification data are shown in Fig. 4. The profiles are quite different. Lake Tuusulanjarvi shows higher levels for all three columns. Paajarvi ice is poor in dissolved matter except for the top layer, which was snow-ice. Chlorophyll a level is 6 mg/[m.sup.3] and then decreases with depth in Tuusulanjarvi. In Paajarvi the value is very low. The ratio of the concentrations in ice and water was in round numbers 0.1 for the yellow substance and 0.5 for the sediment content; for the chlorophyll a the ratio was 2.5 in Tuusulanjarvi and below 0.16 in Paajarvi. Ignoring the snow-ice layer, the yellow substance ratio goes down to 0.01 in Paajarvi but does not change much in hypereutrophic Lake Tuusulanjarvi. The levels in the water were: yellow substance 41.7 and 26.0 mg/L, chlorophyll a 1.2 and 3.1 mg/[m.sup.3], and suspended matter 11.9 and 13.0 mg/L in lakes Tuusulanjarvi and Paajarvi, respectively.
[FIGURE 4 OMITTED]
From Lake Paajarvi another dataset is available for winter 1999. Five layers were sampled separately: new snow, old snow, snow-ice, congelation ice, and water (Table 6). The results confirmed the deductions from the bulk data. The highest levels occurred usually in water; exceptions were copper and zinc in snow and nickel in congelation ice. Except for copper the lowest concentrations occurred in snow and ice, and about equally often in both.
In general, the results for the snow are close to those for the ice. This is likely due to the major role of atmospheric fallout as the source of impurities. The fallout contributes to the ice when the ice is bare and via the formation of snow-ice. In our earlier study in 1996, pH values below 4 were observed in ice and snow but the minimum was 5.9 in the present work. However, in general snowfall brings acid deposits down from the atmosphere and these deposits accumulate into the snow layer during the whole ice season. In 1976-81 the snow samples (286) studied in Finland had a pH range of 3.7-6.2 with a median of 4.6 .
Ice, snow, and water samples were taken from four Finnish lakes in winters 1997-99 for determining the quality and quantity of impurities (for ice and snow from the meltwater). The data include pH, electric conductivity, and dissolved matter concentration, and for the suspended matter within the ice include the total mass, organic proportion, and concentrations of chemical elements. One river was examined for comparisons.
The average concentration of dissolved matter was 11-17 mg/L in ice, 15-23 mg/L in snow, and 31-143 mg/L in water, i.e. [DM.sub.I] ~ [DM.sub.S] << [DM.sub.W]. The sediment content was in these media, respectively, 2-12 mg/L, 4-11 mg/L, and 1-11 mg/L, i.e. [SM.sub.I] ~ [SM.sub.S] ~ [SM.sub.W]; in cleaner waters the level was the highest in the snow. The organic proportion was 5-85% in the sediments. In general, the results for the snow are close to those in the ice, which is likely due to the major role of atmospheric fallout as the source of impurities. The fallout contributes to the ice when the ice is bare and via the formation of snow-ice. The river ice sites showed sediment content higher by one order of magnitude than in lake ice.
The concentration of dissolved matter in the lake ice amounted to the fraction of 0.1-0.4 of that in the lake water. The level is close to that found for the brackish waters on the coast of the Gulf of Finland . But the level should be still much less if the dissolved matter originated from the lake water in congelation ice growth (e.g., ). This suggests that the dissolved impurities observed were mainly from the snow-ice growth and atmospheric fallout. The observed levels in the snow cover and the more detailed vertical stratification also support this suggestion.
Metal concentrations in the ice were usually around 100 [micro]g/L for aluminium and iron; 30 [micro]g/L for zinc; 1-10 [micro]g/L for copper, manganese, and nickel; and below 0.1 [micro]g/L for cadmium. Exceptions were high levels of aluminium and iron in Lake Tuusulanjarvi and a high level of cadmium in Lake Paajarvi. The concentrations in the ice relative to those in the water were within 0.5-2.0. Snowfall is a significant source for many metals. The maximum was usually found in the water except for cadmium and copper for which the highest levels appeared in the ice. Nutrient levels in the ice were usually the following: total phosphorus 2 [micro]g/L, total nitrogen 200 [micro]g/L, sulphates 1 mg/L, chloride 1 mg/L, calcium 2 mg/L, magnesium 0.1-0.9 mg/L, potassium 0.5 mg/L, and sodium 1 mg/L. The levels were much lower (by one order of magnitude) than in the water. Nitrogen showed much higher levels in the snow in lakes Paajarvi, Paijanne, and Vesijarvi; also phosphorus, sulphates, and magnesium levels were high in the snow in Lake Vesijarvi.
The storage and later release of sediments may become an important ecological factor especially as the load of harmful substances to natural water bodies increases. The accumulation takes place during the whole ice season (3-7 months in Finnish lakes), while the load is released to lake water within one month. Part of the ice sediments sinks contributing to the bottom sediments, while the rest remains in the surface layer. The effect of the meltwater load depends on the stratification of the water column. Under lake ice there is normally a thin (one metre or so) stable layer, and then the meltwater load may be heavy.
Thanks go to Ms. Sirje Maekivi for providing vertical structure data for 1996. Financial support was received from the Sohlberg Delegation of the Finnish Society of Sciences.
Received 1 October 2002, in revised form 18 February 2003
[1.] Shumskiy, P. A. Principles of Structural Glaciology. English translation by Dover Publications, Inc., 1964.
[2.] Reimnitz, E., Kempema, E. W. & Barnes, P. W. Anchor ice, seabed freezing and sediment dynamics in shallow Arctic seas. J. Geophys. Res., 1987, 92(C2), 14671-14678.
[3.] Pfirman, S. L., Eicken, H., Bauch, D. & Weeks, W. F. The potential transport of pollutants by Arctic sea ice. Sci. Total Environ., 1995, 159, 129-146.
[4.] Granskog, M. Baltic Sea ice as a medium for sediment, nutrient and pollution storage and transport - some preliminary results. In Proceedings of the XX Nordic Hydrological Conference, Vol. 1. Helsinki, 1998, 21-31.
[5.] Lepparanta, M., Tikkanen, M. & Shemeikka, P. Observations of ice and its sediments on the Baltic Sea coast. Nord. Hydrol., 1998, 29, 199-220.
[6.] Kaartinen, M.-M. Jaan ainespitoisuudet Vantaanjoen alajuoksulla ja jokisuun edustan merialueella talvella 1999-2000. MSc thesis. Department of Geography, University of Helsinki, 2002.
[7.] Kuusisto, E. The thickness and volume of lake ice in Finland in 1961-90. Publ. Water Environ. Res. Inst., 1994, 17, 27-36.
[8.] Gow, A. J. & Govoni, J. W. Ice growth on Post Pond, 1973-1982. CRREL Report 83-4. U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, N.H., 1983.
[9.] Lepparanta, M. & Kosloff, P. The structure and thickness of Lake Paajarvi ice. Geophysica, 2000, 36, 233-248.
[10.] Kuusisto, E. Jaalla on lunta ja vetta. Kalamies, 1979, 3, 1 and 3.
[11.] Weeks, W. F. Growth conditions and the structure and properties of sea ice. In Physics of Ice-Covered Seas (Lepparanta, M., ed.), Vol. 1. University of Helsinki Press, Helsinki, 1998, 25-104.
[12.] SFS-EN 872. Water quality. Determination of suspended solids. Method by filtration through glass fibre filters. Suomen standardisoimisliitto SFS, 1996.
[13.] SFS 3008. Determination of total residue and total fixed residue in water, sludge and sediment. Suomen standardisoimisliitto SFS, 1990.
[14.] SFS-EN 27888. Water quality. Determination of electrical conductivity. Suomen standardisoimisliitto SFS, 1994.
[15.] SFS 3021. Determination of pH-value of water. Suomen standardisoimisliitto SFS, 1979.
[16.] SFS 3044. Metal content of water, sludge and sediment determined by atomic absorption spectroscopy, atomization in flame. General principles and guidelines. Suomen standardisoimisliitto SFS, 1980.
[17.] SFS 3018. Metal content of water determined by atomic absorption spectroscopy, atomization in flame. Special guidelines for calcium and magnesium. Suomen standardisoimisliitto SFS, 1982.
[18.] SFS 3046. Metal content of water, sludge and sediment determined by atomic absorption spectroscopy, atomization in flame. Special guidelines for aluminium. Suomen standardisoimisliitto SFS, 1982.
[19.] SFS 3047. Metal content of water, sludge and sediment determined by atomic absorption spectroscopy, atomization in flame. Special guidelines for lead, iron, cadmium, cobalt, copper, nickel and zinc. Suomen standardisoimisliitto SFS, 1980.
[20.] SFS 3048. Metal content of water, sludge and sediment determined by atomic absorption spectroscopy, atomization in flame. Special guidelines for manganese. Suomen standardisoimisliitto SFS, 1982.
[21.] SFS 5074. Metal content of water, sludge and sediment determined by atomic absorption spectroscopy. Atomization in a graphite furnace. General principles and guidelines. Suomen standardisoimisliitto SFS, 1990.
[22.] SFS 5502. Metal content of water, sludge and sediment determined by flameless absorption spectrometry. Atomization in a graphite furnace. Special guidelines for aluminium, cadmium, cobalt, chromium, copper, lead, manganese, nickel and iron. Suomen standardisoimisliitto SFS, 1990.
[23.] SFS-EN 1189. Water quality. Determination of phosphorus. Ammonium molybdate spectrometric method. Suomen standardisoimisliitto SFS, 1997.
[24.] Valderrama, J. The simultaneous analysis of total nitrogen and total phosphorus in natural waters. Mar. Chem., 1981, 10, 109-122.
[25.] SFS 3030. Determination of the sum of nitrite and nitrate in water. Suomen standardisoimisliitto SFS, 1990.
[26.] SFS 5738. Determination of sulphate in water. Nephelometric method. Suomen standardisoimisliitto SFS, 1992.
[27.] Lepparanta, M. A review of analytic modeling of sea ice growth. Atmosphere-Ocean, 1993, 31, 123-138.
[28.] Lepparanta, M., Reinart, A., Arst, H., Erm, A., Sipelgas, L. & Hussainov, M. Investigation of ice and water properties and under-ice light fields in fresh and brackish water bodies. Nord. Hydrol., 2003, 34, in press.
[29.] Soveri, J. Influence of meltwater on the amount and composition of groundwater in Quaternary deposits in Finland. Publ. Water Res. Inst., 1985, 67, 1-63.
[30.] Lahermo, P., Ilmasti, M., Juntunen, R. & Taka, M. The Geochemical Atlas of Finland, Part 1. The Hydrogeochemical Mapping of Finnish Groundwater. Geological Survey of Finland, Espoo, 1990.
[31.] Lahermo, P., Vaananen, P., Tarvainen, T. & Salminen, R. Geochemical Atlas of Finland, Part 3. Environmental Geochemistry--Stream Waters and Sediments. Geological Survey of Finland, Espoo, 1996.
[32.] Hakanson, L. An ecological risk index for aquatic pollution control--a sedimentological approach. Water Res., 1980, 14, 975-1101.
[33.] Lehmusluoto, P. Veden pieneliotoiminnoista ja niiden mittaamisesta radioaktiivisen hiilen avulla. Vesianalyyttisia menetelmia. Kemistien 14. taydennyskurssin moniste. Finland, 1969.
Matti Lepparanta (a) *, Matti Tikkanen (b), and Juhani Virkanen (b)
(a) Division of Geophysics, University of Helsinki, P.O. Box 64 (Gustaf Hallstrominkatu 2), FIN-00014 Helsinki, Finland
(b) Department of Geography, University of Helsinki, P.O. Box 64 (Gustaf Hallstrominkatu 2), FIN-00014 Helsinki, Finland; firstname.lastname@example.org, email@example.com
* Corresponding author, firstname.lastname@example.org
Table 1. The studied lakes Lake Area, Depth, m [km.sup.2] Mean Paajarvi 13 15.3 Paijanne, southern basin 70 15.0 Vesijarvi (Lahti) 112 6.6 Tuusulanjarvi 6 3.1 Lake Depth, m Eutrophication At site Paajarvi 10 Meso-oligotrophic Paijanne, southern basin 25 Oligotrophic Vesijarvi (Lahti) 7 Eutrophic Tuusulanjarvi 2 Hypereutrophic Table 2. Samples 1996-99 (1-4 for lakes, 5-6 for the river); x--no data. Layer thicknesses and snow density are also given No. Site Date Ice, cm 1/97 Paajarvi March 19, 1997 37 2/97 Paijanne March 21, 1997 40 3/97 Vesijarvi (Lahti) March 21, 1997 39 4/97 Tuusulanjarvi March 23, 1997 32 5/97 Porvoonjoki (Pukkila) March 21, 1997 21 6/97 Porvoonjoki (Porvoo) March 21, 1997 28 1/98 Paajarvi March 27, 1998 52 2/98 Paijanne March 27, 1998 63 3/98 Vesijarvi (Lahti) March 27, 1998 55 4/98 Tuusulanjarvi March 29, 1998 53 5/98 Porvoonjoki (Pukkila) March 27, 1998 46 6/98 Porvoonjoki (Porvoo) March 27, 1998 40 1/99 Paajarvi March 23, 1999 53 2/99 Paijanne March 23, 1999 43 3/99 Vesijarvi (Lahti) March 23, 1999 42 4/99 Tuusulanjarvi March 28, 1999 59 5/99 Porvoonjoki (Pukkila) March 23, 1999 45 6/99 Porvoonjoki (Porvoo) March 23, 1999 53 No. Opaque top Snow layer (1), cm cm g/[cm.sup.3] 1/97 11 0 -- 2/97 1 0 -- 3/97 4 0 -- 4/97 2 0 -- 5/97 14 4 x 6/97 13 0 -- 1/98 10 5 0.33 2/98 8 5 0.22 3/98 7 5 0.26 4/98 7 5 0.34 5/98 11 5 0.32 6/98 14 5 0.34 1/99 23 18 0.25 2/99 3 24 (0.5) (2) 3/99 3 31 (0.5) (2) 4/99 23 5 0.55 5/99 31 5 0.32 6/99 36 5 0.37 (1) snow-ice in the lakes, snow-ice or frazil ice in the river; (2) wet snow and lower 12 cm slush, 0.5 g/[cm.sup.3] is a rough average density Table 3. Bulk properties of the melted ice and snow samples and the water samples, average [+ or -] standard deviation; x--no data Water body Conductivity, Dissolved pH sample [micro]S/ matter, mg/L cm [25 [degree]C] Lake Paajarvi Ice 13.0 [+ or -] 5.6 14.3 [+ or -] 5.5 6.7 Snow 16.5 [+ or -] 7.8 15.0 [+ or -] 5.7 x Water 108.0 [+ or -] 5.2 64.0 [+ or -] 30.6 6.6 Lake Paijanne Ice 7.7 [+ or -] 3.5 11.3 [+ or -] 4.0 6.6 Snow 19.0 [+ or -] 8.5 20.0 [+ or -] 9.9 x Water 78.7 [+ or -] 5.5 31.3 [+ or -] 24.4 7 Lake Vesijarvi Ice 7.0 [+ or -] 4.4 12.7 [+ or -] 5.5 6.6 Snow 28.0 [+ or -] 22.6 23.5 [+ or -] 17.7 x Water 128.3 [+ or -] 3.2 52.3 [+ or -] 30.2 7 Lake Tuusulanjarvi Ice 15.0 [+ or -] 11.8 17.3 [+ or -] 11.7 6.6 Snow 9.5 [+ or -] 0.7 17.0 [+ or -] 15.6 x Water 208.3 [+ or -] 170.3 143.0 [+ or -] 152.2 6.8 Porvoonjoki River Pukkila Ice 17.0 [+ or -] 11.8 24.0 [+ or -] 13.7 6.6 Snow 28.0 [+ or -] 19.8 25.0 [+ or -] 17.0 x Water 239.0 [+ or -] 64.3 127.3 [+ or -] 80.1 6.9 Porvoo Ice 19.0 [+ or -] 14.2 19.0 [+ or -] 14.9 6.6 Snow 14.5 [+ or -] 0.8 15.5 [+ or -] 4.9 x Water 239.7 [+ or -] 74.3 135.0 [+ or -] 90.2 6.9 Water body Sediment content sample mg/L organic, % Lake Paajarvi Ice 2.1 [+ or -] 0.5 36 Snow 4.2 [+ or -] 3.4 38 Water 3.7 [+ or -] 4.2 40 Lake Paijanne Ice 2.1 [+ or -] 1.4 23 Snow 7.0 [+ or -] 5.9 50 Water 1.4 [+ or -] 0.6 49 Lake Vesijarvi Ice 2.0 [+ or -] 0.2 33 Snow 9.9 [+ or -] 9.5 54 Water 1.1 [+ or -] 0.4 50 Lake Tuusulanjarvi Ice 12.6 [+ or -] 5.0 24 Snow 11.6 [+ or -] 9.8 58 Water 11.5 [+ or -] 8.3 22 Porvoonjoki River Pukkila Ice 190.0 [+ or -] 291.1 31 Snow 64.2 [+ or -] 78.3 80 Water 16.0 [+ or -] 4.8 19 Porvoo Ice 24.5 [+ or -] 13.7 32 Snow 14.9 [+ or -] 7.8 56 Water 17.0 [+ or -] 5.2 19 Table 4. Average concentrations ([micro]g/L) of metals from winters 1998-99 Water body A1 Cd Cu Fe Lake Paajarvi Ice 102 0.5 6 149 Snow 216 <0.1 5 127 Water 99 <0.1 5 310 Lake Paijanne Ice 114 0.1 8 73 Snow 140 <0.1 5 110 Water 75 <0.1 6 87 Lake Vesijarvi Ice 94 <0.1 9 62 Snow 517 <0.1 5 303 Water 55 <0.1 5 40 Lake Tuusulanjarvi Ice 867 <0.1 5 965 Snow 416 <0.1 8 277 Water 2195 0.1 11 1710 Porvoonjoki River Pukkila Ice 1706 0.1 27 1278 Snow 1916 0.1 8 1676 Water 1680 0.1 7 1775 Porvoo Ice 1445 0.1 16 1094 Snow 626 0.2 6 707 Water 1795 <0.1 6 1545 Water body Mn Ni Zn Lake Paajarvi Ice 4 7 28 Snow 3 5 27 Water 22 7 27 Lake Paijanne Ice 1 4 33 Snow 3 4 25 Water 2 3 56 Lake Vesijarvi Ice 2 1 26 Snow 7 6 36 Water 2 4 55 Lake Tuusulanjarvi Ice 9 4 29 Snow 5 4 18 Water 53 9 50 Porvoonjoki River Pukkila Ice 21 5 32 Snow 39 10 21 Water 76 8 26 Porvoo Ice 17 6 32 Snow 9 6 26 63 12 48 Table 5. Nutrient concentrations in winter 1999 Water body P N P-P Tot N-N Tot [O.sub.4] [O.sub.x] [micro]g/L Lake Pajarvi Ice 2 3 95 260 Snow <2 3 592 978 Water 5 14 867 1257 Lake Paijanne Ice <2 2 11 102 Snow <2 4 390 590 Water <2 4 225 487 Lake Vesijarvi Ice <2 <2 <5 204 Snow 5 10 232 755 Water 22 28 185 567 Lake Tuusulanjarvi Ice 8 18 213 487 Snow 5 6 130 318 Water 40 85 3216 4799 Porvoonjoki River Pukkila Ice 40 61 555 687 Snow 14 24 853 1394 Water 82 121 4870 6068 Porvoo Ice 38 64 664 848 Snow 5 16 431 933 Water 74 118 3920 4910 Water body S[O.sub.4] Cl Ca mg/L Lake Paajarvi Ice 1.5 0.9 2.5 Snow 2.8 0.9 1.6 Water 18.0 7.5 8.6 Lake Paijanne Ice <1 0.3 3.2 Snow 2.4 1.8 1.8 Water 9.4 6.3 2.6 Lake Vesijarvi Ice <1 0.4 1.2 Snow 4.5 3.0 2.4 Water 14.0 9.5 10.4 Lake Tuusulanjarvi Ice 2.1 1.7 2.4 Snow <1 0.4 1.3 Water 39.0 31.0 16.9 Porvoonjoki River Pukkila Ice 2.5 1.7 2.1 Snow 4.4 2.4 3.6 Water 28.0 21.0 15.7 Porvoo Ice 2.7 2.4 1.8 Snow 1.1 1.0 1.2 Water 29.0 23.0 14.7 Water body Mg K Na mg/L Lake Paajarvi Ice 0.2 0.3 0.4 Snow 0.2 0.3 0.5 Water 2.7 2.4 4.2 Lake Paijanne Ice 0.9 1.0 3.8 Snow 0.3 0.4 1.0 Water 0.7 0.9 3.1 Lake Vesijarvi Ice 0.1 0.3 0.3 Snow 0.6 0.8 1.2 Water 3.7 3.1 6.1 Lake Tuusulanjarvi Ice 0.5 0.5 0.8 Snow 0.1 0.3 0.3 Water 10.0 4.8 14.4 Porvoonjoki River Pukkila Ice 0.6 1.3 1.1 Snow 1.0 0.7 1.3 Water 6.0 9.3 20.5 Porvoo Ice 0.6 0.8 1.3 Snow 0.2 0.4 0.6 Water 7.0 7.7 21.3 Table 6. Results for vertical stratification in Lake Paajarvi, 1999 Parameter New snow Old snow Snow-ice Conductivity, S/cm 12.7 25.7 23.6 Dissolved matter, mg/L 20 19 16 Sediments, mg/L 1.2 2.1 3.3 pH 6.99 6.59 7.17 Aluminium, [micro]g/L 62 51 56 Cadmium, [micro]g/L <0.1 <0.1 <0.1 Copper, [micro]g/L 4.2 6.9 5.5 Iron, [micro]g/L 32 46 56 Manganese, [micro]g/L 0.3 4.9 5.1 Nickel, [micro]g/L 2 3 5 Zinc, [micro]g/L 46.2 6.4 26.3 Calcium, mg/L 1.8 2.2 5.9 Magnesium, mg/L <0.1 0.5 0.5 Potassium, mg/L 0.1 0.4 0.4 Sodium, mg/L 0.2 0.8 0.7 P[O.sub.4]-phosphorus, <2 <2 3 [micro]g/L Total phosphorus, <2 4 5 [micro]g/L N[O.sub.3]- & 485 699 176 N[O.sub.2]-nitrogen, Total nitrogen, g/L 933 1023 407 Sulphate, mg/L 1.2 4.4 2.6 Chloride, mg/L 0.3 1.4 1.3 Parameter Congelation Water ice Conductivity, S/cm 7.8 105 Dissolved matter, mg/L 18 32 Sediments, mg/L 1.6 3.6 pH 7.32 6.97 Aluminium, [micro]g/L 41 98 Cadmium, [micro]g/L <0.1 <0.1 Copper, [micro]g/L 5 3.4 Iron, [micro]g/L 41 383 Manganese, [micro]g/L 0.4 21.6 Nickel, [micro]g/L 9 7 Zinc, [micro]g/L 29.9 16.8 Calcium, mg/L 1.8 8.5 Magnesium, mg/L <0.1 2.6 Potassium, mg/L 0.1 2.2 Sodium, mg/L 0.2 4.1 P[O.sub.4]-phosphorus, <2 5 [micro]g/L Total phosphorus, <2 14 [micro]g/L N[O.sub.3]- & 13 867 N[O.sub.2]-nitrogen, Total nitrogen, g/L 113 1257 Sulphate, mg/L <1 18 Chloride, mg/L 0.4 7.5