NP Budget for the Southern North Sea

Stephen V. Smith, Paul R. Boudreau and Piet Ruardij

4 August 1997

The following budget is presented in greater detail than most others, because this budget provides a basis for comparison between the "simple" LOICZ approach to budgeting and more sophisticated modelling procedures. This topic will be explored during the "North Sea Session" of the upcoming Open Science Meeting in the Netherlands in October 1997.

The North Sea (51° to 61° N, 3° W to 9° E) is a large coastal sea bounded by the British Isles, the Scandinavian peninsula, and the mainland of northwestern Europe (Figure 1). Although this sea is often called a shelf sea, almost half the area has a depth well in excess of 100 m and is essentially oceanic in character.

Figure 1. Map of northwestern Europe, showing the Southern North Sea and surrounding countries. The two budgeted boxes (box 1, 2) used in the present analysis are illustrated.

Freshwater inflow into the Southern North Sea totals about 190 km³/yr and is dominated by the Rhine (about 65; Otto, in Anonymous, 1983). This inflow does not show strong seasonality.

The shelf portion of the North Sea is impacted by various forms of pollution and resource utilization. Approximately 150 million people live within the drainage basin for this sea. Radach and Lenhart (1995) summarize the estimated inputs of river-borne inorganic nutrients used in the budgets described here. Atmospheric nitrogen input has been estimated for the North Sea south of 56° N by Rendell et al. (1993); we have extrapolated that estimate to the area covered in this report.

As a result of these and other considerations, this area is also one of the most intensively studied shelf seas on Earth. One consequence is that developing a relatively simple budget according to the LOICZ Guidelines (Gordon et al., 1996) for this region provides a potential basis for comparing with many other studies of the region. It is, however, beyond the scope of this short write-up to review that extensive literature. Three publications which are particularly noteworthy, however, are recent issues of the Netherlands Journal of Sea Research (v. 25 (no.1/2), 1989; v. 26 (no 2/4), 1990; and v. 33 (no.3/4), 1995).

The analysis presented here is for the Southern North Sea, between the English Channel and 57° N. The English Channel is a constriction which partially isolates the North Sea from inflow from the southwest into this shelf region, while 57° N is slightly south of the shelf break and the much deeper waters of the Skagerrak and Norwegian Coastal Current—both of which are largely oceanic in character. For budgeting purposes, the system was further divided at 54° N. The southern box is the coastally dominated region receiving inputs of fresh water and nutrients from southern England and the European coast between about the Rhine River and the southern boundary of Denmark. This box has an approximate area of 90,000 km² and a volume of 3,200 km³ (yielding a mean depth of about 36 m). The northern box is more open-shelf in character, receiving inputs from Denmark, northern England, and Scotland. This latter box has an area of about 210,000 km² and a volume of 10,100 km³ (average depth = 48 m). Commonly quoted statistics for the whole of the North Sea are that it covers an area of 520,000 km² and has a volume of 41,000 km³ (average depth of 79 m). Thus, the portion we have not budgeted has a mean depth of about 126 m, emphasizing that the budget presented here largely describes the shallower (shelf) portion of the North Sea. This region is sometimes called the Southern North Sea.

Primary production varies considerably across the North Sea, with the highest values being in the coastal regions under the largest influence of terrestrial nutrient inputs (e.g., Peeters and Peperzak, 1990). At the scale of the boxes budgeted here, various estimates summarized by Varela et al (1995), Joint and Pomeroy (1992) indicate that the average primary production for the Southern North Sea is between 150 and 200 g C m^-2 yr^-1, with a strong peak (spring bloom) about April. It is difficult getting a resolution between model and observational results (e.g., Moll 1997, Baliño, 1996), or indeed to know which are the more realistic system-level values.

One complication in budgeting this system according to the LOICZ Guidelines is that the North Sea is not "sealed" at the English Channel. That is, ocean water enters the system both along the western side of northern boundary (57° N) and through the English Channel (e.g., see Anonymous, 1983; Otto et al., 1990; Lenhart et al., 1995; Lenhart and Pohlmann, 1997). Problems associated with this ambiguity in "oceanic source waters" exchanging with the system are discussed in general in the Guidelines. The eventual solution to this problem was to assume that the English Channel flow was known, allowing calculation of both residual flow and exchange flow between boxes 1 and 2, and between 2 and the North Atlantic north of 57° . We adopt a net flow through the English Channel of 1,800 km³/yr, based on Lenhart and Pohlmann (1997).

Sensitivity of the nonconservative budgetary calculations was evaluated by altering the English Channel flow well above and well below what are generally considered to be reasonable estimates of the range of this flow.

Various budgets or partial budgets have been provided for nutrients in the North Sea. We particularly call attention to the budgets by Brockmann et al. (1990), Radach and Lenhart (1995), Kempe and Pegler (1991), and Baliño (1996), and Nelissen and Stefels (1988). Lenhart et al (in press) explore the effect of altered river flows on ecosystem dynamics. In addition to these papers generally dealing with aspects of nutrient budgets in the North Sea, Helder et al. (1996) presented a preliminary budget based on the LOICZ Guidelines as laid out by Gordon et al. (1996). The following budget is a refinement of the Helder et al. budget, based on more comprehensive data. One reason to undertake the simple LOICZ budgeting approach in such a well-studied system as the North Sea is to compare the results of the budget with the large amount of detailed information for the system.

Data used in this analysis are from NOWESP (North-West European Shelf Programme) (NOWESP, 1996;Radach et al., 1996; Radach et al., 1997; Radach & Gekeler, 1997). The data have been collated into grids cells which are 12’N-S by 20’E-W (i.e., areas of approximately 500 km² in area), and further divided into 12 equal time intervals over an annual cycle ("months"). Monthly average data are calculated only for those variables with a minimum of 3 years’ data during the period between 1960 and 1990. It is noteworthy that, over the period of this data record, there is no strong evidence of large compositional changes in nutrient concentrations within the North Sea. These data were, in turn averaged into four groups: (southern box--51-54N, 3W-9E; northern box—54-57N, 3W-9E; English Channel—48.5-51N, 4W-2E; North Atlantic (NA)—57-58 N, 1W-3E). Initially the data were averaged by months; eventually the monthly data were collapsed into annual averages. Properties included in the following analysis are salinity (S), dissolved inorganic phosphorus (DIP), nitrate, nitrite (eventually combined and labeled "NO₃"), and ammonium (NH₄).

The annual trends for the data are shown in Figure 2. As shown in the figure, there is a clear spatial progression of salinity from values of about 33.8 psu in the southern box, to about 34.3 in the northern box. Both the North Atlantic water and the English Channel water have salinity near 35 psu. Note, further, that neither the two North Sea boxes nor the two oceanic source waters show a significant seasonality. This, combined with the fact that the estimate of freshwater inflow (Anonymous, 1983) reports insignificant seasonality in freshwater inflow, led us to develop the water and salt budget (Figure 3) as a steady state budget based on the annual average data. Precipitation in excess of evaporation, is adapted from Anonymous (1983).

Figure 2. Monthly variation of water composition in the boxes used for the North Sea budget calculations.

Figure 3. Steady-state 2-box model for water and salt exchange between the southern (V₁) and northern (V₂) boxes of the Southern North Sea. The green arrows indicate known (or, in the case of the English Channel [EC] assumed) fluxes. The red arrows indicate calculated fluxes.

We test the sensitivity of the budgetary calculations to estimated strength of the flow through the English Channel (Figure 4), using a range of flows between 1,500 and 5,000 km³/yr; based on Anonymous (1983), Otto et al. (1990), and Brockmann et al. (1990), Lenhart et al. (1995), and Lenhart and Pohlmann (1997). Because the salinity of the North Atlantic water and the English Channel water masses are close to one another, the choice of this flow has relatively little effect on the total exchange of North Sea water with respect to the "ocean" (i.e., the combination of the English Channel and North Sea). This choice does, of course, affect the exchange between the southern and northern boxes, because this choice "directs" the partitioning of oceanic inflow into the two boxes. With 1,800 km³/yr as the nominal English Channel inflow, the calculated water exchange time for the two boxes combined is estimated by this approach to be about one year (370 days).

Figure 4. Changing water exchange as a function of the strength of English Channel flow. Note that both residual flows and south-north exchange are more sensitive to the English Channel flow than is the exchange flow between the northern box and the North Atlantic.

Lenhart and Pohlmann (1997) report a flushing time for the whole of the North Sea (to 61° N) of 167 days. Discussions with Lenhart (personal communication) suggest that part of the discrepancy may lie with a systematic underestimate of water exchange by consideration of the North Sea water column as vertically well mixed and with single-layer flow. A second part of the discrepancy may lie with comparing the whole of the North Sea with the Southern North Sea calculations presented here. The two-layer versus one-layer discrepancy cannot be resolved with data in Lenhart and Pohlmann (1997), but data in Lenhart et al. (1995) suggest this error to be of the order of 40% for the whole of the North Sea. In any case, calculation of two layer (net) exchange for the Southern North Sea Lenhart et al. (1995), give results much more comparable to ours: 353 days. We conclude that the discrepancy between the two models for the water exchange time of the Southern North Sea lies somewhere between 5 and 40%.

The nutrients, especially DIP and NO₃, show strong seasonality, with only slight differences among the water masses. NH₄ shows somewhat weaker seasonality and more differences between water masses. To the extent that there are water mass differences in nutrient composition, they progress from higher concentrations in the southern box, lower in the northern box, and roughly comparable values for the two oceanic water masses. Figures 5-7 present the DIP, NO₃ (+NO₂), and NH₄ budgets for the North Sea. In this preliminary analysis, terrigenous input of N is all attributed to NO₃. This is apparently a reasonable approximation of present river nutrient loadings.

Figure 5. Steady-state 2-box model for DIP flux in the North Sea. This is based on an assumed annual flow of 1,800 km³ of water from the English Channel into the southern box. This assumed flow affects DDIP in the southern box but does not affect the summed DDIP of the southern and northern boxes very much.

Ultimately, this does not matter, because the nonconservative fluxes of the nitrogen species are grouped into the term "DDIN" (Table 1).

Figure 6. Steady-state 2-box model for NO₃ flux in the North Sea. This is based on an assumed annual flow of 1,800 km³ of water from the English Channel into the southern box. This assumed flow does not have much affect on DNO₃ in either the southern or northern boxes. Note that all of the terrigenous input of DIN is attributed to NO₃.

Table 1. Nonconservative fluxes for dissolved inorganic phosphorus (DDIP) and dissolved inorganic nitrogen (DDIN) in the two boxes of the Southern North Sea.

Figure 7. Steady-state 2-box model for NH₄ flux in the North Sea. This is based on an assumed annual flow of 1,800 km³ of water from the English Channel into the southern box. This assumed flow does not have much affect on DNH₄ in either the southern or northern boxes. Note that all of the terrigenous input of DIN is attributed to NO₃.

Input of DIP from land into the two boxes of the North Sea is about 1.4 x 10⁹ mol/yr (1.1, Southern box; 0.3 Northern box; Figure 5). The nonconservative flux in the southern box is a small portion of the terrigenous input and varies strongly as a function of the estimated English Channel flow. Therefore this nonconservative flux is not significant. However, with respect to the whole of the North Sea, about 70% of the DIP input from land is taken up. This calculation is not sensitive to the English Channel flow, because of similar DIP concentrations in the English Channel and North Atlantic (Figure 2).

Input of DIN from land plus the atmosphere is estimated to be about 69 x 10⁹ mol/yr. This input is about equally split between the Southern and Northern boxes, with the terrigenous input apparently dominating in the south and the atmospheric input in the north (Figures 6, 7). In both boxes, the nonconservative uptake appears significant, accounting for most of the land + atmosphere input in the southern box and an amount exceeding this land + atmospheric input in the north. Apparently no DIN entering by these pathways escapes the North Sea.

Figure 8. Seasonal variation in net ecosystem metabolism for the southern and northern boxes of the North Sea. The symbols represent monthly calculations, and the lines represent a 2-month running mean through the data.

The data in Table 1 can be used along with the stoichiometric model to calculate the rates of net ecosystem production (p-r) = -106 x D DIP. These calculations are based on the assumptions that organic matter being processed is dominated by plankton, and that nonconservative fluxes of DOP and DON are minor. The results, expressed per unit area, suggest that (p-r) for the North Sea as a whole averages about 0.4 mol m^-2 yr^-1. The primary production is assumed to be about 17 mol m^-2 yr^-1, so net ecosystem production apparently accounts for about 2% of primary production.

Figure 9. Seasonal variation in net nitrogen fixation minus denitrification for the southern and northern boxes of the North Sea. The symbols represent monthly calculations, and the lines represent a 2-month running mean through the data.

The stoichiometric model can also be used to calculate the difference between net nitrogen fixation minus denitrification (nfix-denit) = DDIN_obs – 16 x DDIP. The rate obtained is approximately –130 mmol m^-2 yr^-1 (~200, southern box; 100, northern). This is very much in line with estimates of denitrification for this region (e.g., Lohse, 1996; Lohse et al. 1996).

One further calculation is presented here for the nonconservative nutrient fluxes. Even though a steady state model was used for the water and salt budgets (Figure 3), the strong seasonalities in the nutrient concentrations lent themselves to calculating seasonal variation of the nonconservative fluxes. These are presented for each of the two boxes and are directly expressed as rates of (p-r) and (nfix-denit) per unit area (Figures 8, 9). Because the data are somewhat noisy, a 2-month running average was calculated through the data sets. Net ecosystem metabolism (p-r) does not show much difference between the two boxes, but there is a strong seasonality. The rate peaks during April-May ([p-r] ~ 1.5 mol m^-2 mo^-1), with most of the remainder of the year exhibiting (p-r) near or slightly below 0. By contrast with the strong seasonality exhibited by the net ecosystem metabolism, (nfix-denit) appears to be relatively constant (and negative) throughout the year. The lack of a seasonal pattern and general "noisiness" of the denitrification data may reflect the assumption that atmospheric inputs are distributed evenly over the year.

Some final points are worth noting. The calculated nonconservative fluxes for the Southern North Sea as a whole are not particularly sensitive to the water exchange calculations. The basis for this conclusion is that residual inflow from the English Channel, residual outflow to the North Atlantic, and mixing exchange between the northern box and the North Atlantic all involve water masses of rather similar composition. In contrast, the budgets suggest that the calculated net nonconservative fluxes are very sensitive to terrigenous or atmospheric inputs. Apparently little of the dissolved inorganic nutrient material put into this system from land leaves the system in dissolved inorganic form. Rather, most of it is converted to either dissolved organic material or (we assume more likely) particulate material. We assume that this particulate material would be dominated by organic matter. Any particulate material is either sedimented or (more likely) exported in particulate form. The terrigenous loadings of dissolved inorganic nutrients appear to be relatively well known. With respect to the nitrogen budget, it is worth pointing out that some estimates of atmospheric input to the North Sea are higher than we have used (see references in Rendell et al., 1993). Should these higher estimates prove correct, the consequences with respect to this budget would be that calculated denitrification would be elevated above what we have calculated.

More information on NOWESP data may be found at the Ecological Modelling homepage of the Institute of Oceanography, University of Hamburg

Back to [Node Introduction] [LOICZ]

You are visitor number since April 25, 1998

Last Updated 21 May 2006 by DPS