Estuarine and Coastal Marine Science (1973) I,
Chemical Evidence for the Dispersal of River Mersey run-off in Liverpool Bay
M. I. Abdullah
and L. G. Royle
Department of Oceanography, University of Liverpool, Liverpool, England and Yorkshire River Authority, Leeds, Yorkshire, England Received I3 September 1973
Analysis of chemical data (nutrients and some trace metals) for surface water obtained in December 1970 and March 1971 from Liverpool Bay shows that the run-off from river Mersey can be identified in the Bay. The data also suggest that, at least, the Mersey system in Halliwell’s (1972) studies can be confirmed. The R. Mersey influence area extends in north westerly direction and
supports the view that the residual circulation
is according to Ramster &
Hill’s (1969) statement. The extent of the area and the rate of dispersal of R. Mersey effluent was found to be mainly controlled by the turbulent mixing caused by meteorological conditions as well as tidal effects.
Introduction The dispersal of run-off in coastal waters depends upon the residual circulation, the tidal currents and the associated turbulence. The dispersal of run-off in Liverpool Bay is a particularly important issue in view of the nature and magnitude of waste discharged in the area. This has induced a number of workers to study the hydrography of the area, the physical and chemical properties of the water and has resulted in the setting up of a Working Party to investigate the effects of sludge disposal in the area. Salinity and temperature studies (Bowden, 1955) and current measurements (Ramster & Hill, 1969; Ramster, 1972) suggested that the residual surface currents in the winter are north-westerly and the sea bed flow is inshore. This observation was supported by evidence furnished from drifter release (Halliwell, 1972, 1973). 1n d irect evidence based on the distribution of plankton (Khan & Williams, 1970), however, suggested a clockwise gyre which is contrary to the Ramster-Hill type of circulation postulated. The evidence based on the distribution of nutrients (Jones & Folkard, 1971) and some trace metals (Abdullah et al., 1972) in Liverpool Bay suggests an anticlockwise circulation. It has recently been suggested, therefore, that the surface residual currents in the winter months in Liverpool Bay are northerly to north-westerly in an anticlockwise gyre and are predominantly wind driven (Ramster, 1973), being different from those encountered in the summer. The hydrodynamical theory of the density induced circulation in the area made by Heaps (1972) suggests, nevertheless, that the surface residual currents would be NNW and the bottom currents inshore, thus supporting Halliwell’s results (1972). A study of the distribution of some trace metals (e.g. zinc) in Liverpool Bay (Abdullah et al., 1972) suggested that a ‘plume’ was associated with the R. Mersey run-off; other ‘plumes’ from different sources were also observed. The existence of river ‘plumes’,
M. I. Abdullah
& L. G. Royle
however, depends on the rate of mixing in the area and on the type of circulation prevalent. Studies of these processes in Liverpool Bay have been made by employing techniques such as dye diffusion (Talbot, 1972), radioactive labelling (Barrett et al., 1972), drifter release as well as by current measurements. These techniques with the exception of drifter release and arrayed current meter moorings, give limited information of time and spatial scales of dispersive mechanisms. An alternative approach is examined here and is based on the inherent and quite often unique chemical character of a particular run-off which may be used as a label. The dispersal of the R. Mersey effluent is important from both pollution and ecological considerations and because of this water’s characteristic composition, its rate of dispersal may be studied by considering the distribution of various chemical parameters in the Liverpool Bay area.
Chemical observations in Liverpool Bay Surveys were made during September 1970, December 1970, March 1971 and August 1971. During each of these salinity and temperature were measured, silicate and nitrates (46 stations) and some trace metals (23 stations) were determined on surface water samples from stations on a grid covering the area east of an Anglesey-Isle of Man-Morcambe Bay boundary. Since the distribution of nutrient salts and, to some extent, trace metals are likely to be modified by plankton utilization in the summer and autumn, only the results obtained in December 1970 and March 1971 were considered. The level of nutrients found in the summer and autumn was very low and comparable to that reported by other workers (Jones & Folkard, 1971; Spencer, 1972), who attributed the low level to plankton utilization. The distribution of salinity, silicate, nitrates, zinc and nickel are shown in Figure I. The highest values were found in the near-shore water around the R. Mersey outfall and these values decrease seawards. The lines of equal concentrations show the characteristic bulge indicating a ‘plume’ off the R. Mersey. In a two-component system of fresh water-sea water mixture, in inshore waters such as the Bristol Channel and Southampton waters (Abdullah et al., 1973; Banoub & Burton, 1968), the chemical species when these are contributed mainly by the run-off, bear an inverse linear relationship with the salinity. Thus, for example, the silicate, nitrate and phosphate levels show this inverse relationship over the whole of the Bristol Channel. The bulk of the run-off is added to the eastern part of the Channel and the Si/S%, ratio follows the theoretical mixing relationship. Because the silicon is primarily land derived, the mixing and dispersion of run-off is best considered by examining the silicon in the water. The silicon oersus salinity plots in Liverpool Bay for December 1970 and March 1971 are shown in Figure 2. An inverse relationship can be observed but instead of a single relationship as found in the Bristol Channel and other areas, two lines can be drawn in both instances (December and March) distinguished as A, low silicate content, and B, high silicate content relative to the salinity. Silicon removal by the reconstitution of alumina-silicates has been postulated in estuarine mixture (Bien et al., 1958), however, it was found that this is more likely to take place at relatively low salinities (Banoub & Burton, 1968). This removal is unlikely to be the cause of the low silicate in relations A since the salinity range for the two lines (A and B) overlap considerably and the lowest salinity is about 30%~. It is, therefore, likely that the two lines (Figure 2) represent two discrete water massesproduced by the mixing of chemically different run-offs with sea water. The existence of these two water mixtures is further supported by the examination of other chemical parameters such as nitrates, zinc and nickel (Figure 2)
& %I! I
-0 a 2 I
$ M 1
-0 6 B I
.// . /
where a plot of the relationship of these salts with the salinity exhibits two lines as shown by the Si W~YSUS S%,,plots in both December 1970 and March 1971. The Si-S%, relationship in a simple fresh water-sea water mix may be defined by the equation: Si=a-k
where a is the silicon content of the fresh water, k is a constant and is a function of the salinity of the base sea water, its silicon content and the silicon content of the fresh water. The silicon (and other salts) content of the run-off responsible for the two relationships in both December and March can be extrapolated to S=o%, (see Table I). The difference in the run-off composition is quite marked. Calculating the theoretical mixing relation for the two relationships in December and March assuming run-offs of the compositions listed in Table I is mixed with open sea water of S%,=33*6%, and Si content of 6.5 pg at./litre in December and S%,=33.9$‘& and Si content of 5.9 pg at. litre in March, a good fit between the observed and calculated Si/S%, ratios were obtained for both lines A and B in March but only for line A in December. The discrepancies found for the line B in December, however, are due to the assumption that this particular run-off is being diluted with open sea water. This assumption is not entirely valid since the Si versus S%, lines for December do not intersect at the open sea water salinity (Figure 2). Recalculating the theoretical mixing relationship for the B run-off in December with sea water of the same composition as that at the intersection of lines A and B (Figure 2, Si, December) then the observed Si/S%, ratio in the samples defining line B, show a good agreement with the calculated values, as seen in Figure 3, and no significant anomalies are found. Thus the influence of more than one runoff source can be identified in Liverpool Bay. TABLE I. Extrapolate
and measured composition December A
Run-off Extrapolated composition Si (ug at./l) NO,-N (ng at./l) Ni (w/l) Zn (dl)
Measured composition Si (ng at./l) R. Dee at Iron Bridge 65 NOS
169 350 14
(Dee and Cwlyd
R. Mersey at Widnes Manchester Ship Canal R. Weaver at Frodsham R. Dee at Iron Bridge
R. Dee at Iron Bridge R. Mersey
107 (Mersey and Weaver River Authority 1971) 602 (Mersey and Weaver River Authority 1971) 430 (Mersey and Weaver River Authority 1971) 80 (Dee and Clwyd River Authority 1971)
I IO (30-380) (Dee and Clwyd 160” (O’SUlhan, 1972)
“As zinc in waste discharged into the tidal section of the Mersey estuary. The actual value will be higher due to input of zinc from sources other than waste discharge.
Plotting the boundary between the stations defining the two lines A and B (Figure 2) for December and March produce a well-defined area (Figure I) where the Si-S%, relationship is defined by line B and outside of which the Si-S%, relationship is given by line A. The configuration and location of this area suggests clearly that it is influenced entirely by the
M. I. Abdullah &f L. G. Royle
output from R. Mersey. Defining the area of the R. Mersey influence using other chemical parameters such as nitrate and zinc, the extent and configuration obtained is almost identical to that obtained using silicon data which strongly suggests that the R. Merseyrun-off contains different (high) levels of these chemical salts and that the dispersal of this run-off is rather restricted, being controlled by the mixing processes prevalent in the Liverpool Bay area. Very little is known of the chemical composition of the R. Mersey water. However, it is assumed to contain high levels of nitrate and some trace metals because of the waste discharge into the river. The extrapolated values for the R. Mersey (relation B, Figure 2) and the rest of run-off (relation A, Figure 2) compare reasonably well with the limited data available (Mersey and Weaver River Authority, Dee and Clwyd River Authority; O’Sullivan, 1972), see Table I. The extrapolated silicon value from line A is of the same order as that reported for R. Dee and similarly the extrapolated value of nitrate from line B is comparable to values reported for the R. Mersey. December
h5% I 0%
Figure 3. The relationship between silicon-salinity ratio (pg at./l/%J and salinity (X) in the surface water of Liverpool Bay. Superimposed on these are the theoretical mixing relationship between fresh water and sea water (seetext).
The enclosed area as defined above is, therefore, designated here as being the area of sea where (in the surface layer) the effect of the R. Mersey run-off is identifiable by the present analytical means. Its extent or the position of its outer boundaries are approximate and are only accurate within both the analytical efficiency and the spacing of the sampling stations. The existence of such a well defined area indicates that mixing along the shore between North Wales coast and Lancashire coast is not complete and that some restriction must be present. In the sea bed drifter release study in Liverpool Bay, Halliwell (1972, 1973) observed the existence of two boundaries, one between R. Dee and R. Mersey and the other between R. Mersey and R. Ribble, which drifters never crossed and thus divided the area into three systems: the north Lancashire, the Mersey and the Dee, and north Wales systems. These boundaries were observed over a period of twelve months and although they changed position with season, the data presented suggests that, at least near shore, the change was not signiticant. The position of the ‘boundaries’ reported by Halliwell lies very close to the lateral boundaries of the area of the R. Mersey influence derived here. However, since Halliwell’s studies relate to bottom waters, the significance of this coincidence is not immediately clear in the present state of knowledge.
River Mersey run-off in Liverpool Bay
Mlxlng and dispersal of R. Mersey run-off in Liverpool Bay The volume of R. Mersey water in the enclosed areas (Figure I) is estimated by considering both the salinity and the silicon distribution and the total volume of sea in the area. All calculations were made graphically by drawing z-km sections across the off-shore direction of the plume. The values for the volume of fresh water present in the areas calculated from the salinity and silicon are listed in Table 2 and show good agreement in both December 1970 and March 1971. It is assumed that because of the strong turbulent mixing found in Liverpool Bay (Heaps, 1972), top to bottom differences are small enought not to affect the calculations unduly. The time required to replace the fresh water in the area is calculated by dividing the fresh water volume by the rate of R. Mersey flow. This flow rate for the periods October-November 1970 and November r97o-February 1971, corrected for ungauged area is 103 and 87 cumecs, respectively (data supplied by Mersey and Weaver River Authority). The replacement times for the Mersey influence area in December and March calculated from salinity values (Table 2) are 37 and 124 days, respectively, which agree with those calculated from the silicon data (32 and I 17 days, respectively). As the volume (or area) of sea involved in the calculation of the replacement time is different, the replacement time is not necessarily significant as an indicative parameter of water replacement or the rate of mixing. In this respect, it is the area that is significant and replacement time is related to this and the rate of run-off. TABLE 2. Some calculated pool Bay
of the area of the Mersey influence December
Volume of F.W. in the area (from salinity) Volume of F.W. in the area (from Si) Replacement time (from salinity) Replacement
time (from Si)
Area of influence sq. km Horizontal dispersal rate (from salinity) Horizontal dispersal rate (from Si)
331 X IOOms 283 X 10~ m3
37 days (322 x 106 s) 32 days g5 x IO6 4
993 X 10~ ms 882x xosma 124 days (IO.74 x 106 s) I 17 days (10.16 x 10~ s) 952
175 m*/s 204 ms/s
89 ml/s 100 mP/s
From the area over which the R. Mersey water is identifiable and the replacement time it is possible to calculate the horizontal dispersal rate in the surface layer which, from dimensional consideration, is the area of influence divided by the replacement time. This was found to be 174 m2/s in December and 98 m2/s in March as calculated from the salinity values and 203 m2/s in December and IOO m2/s in March as calculated from the silicon data. It is necessary to point out that by virtue of the method of calculations, the calculated dispersal rates represent the total overall mixing and dispersal process averaged over a long period of time ranging between one and three months. Since these processes are likely to be affected by meteorological and environmental conditions and since the R. Mersey run-off is discharged into an area, as some evidence (Halliwell, 1972) suggests, of restricted exchange with surrounding water, the significance of the dispersal rate is in indicating the rate at which the R. Mersey run-off is being mixed so as to become unidentifiable. As horizontal diffusion due to turbulence is component process, the dispersal rate, calculated here, cannot be compared with coefficient obtained by such techniques as dye or radioactive tracer diffusion studies (Talbot, 1972, Barrett et al., 1972). However, calculation of the effective horizontal diffusion
M. I. Abdullah
& L. G. Royle
coefficient made by Bowden & Sharaf El Din (1966), taking into consideration the mean transport of salt, shear and tidal effects, give a value of 160 m2/s (longitudinal, Ky) which is of the same order of magnitude as the dispersal rates calculated here. Conclusions From the results and calculations presented above, a number of general observations can be made. I. The isohalines and lines of equal concentration of silicate nitrates and zinc etc., show no symmetrical arrangement within the area of the R. Mersey influence. These lines are mainly governed by the volume of fresh water in the mixture, the magnitude and direction of the surface currents. The area, therefore, signifies the extent of seacontaining detectable amounts of Mersey water. Also, the isohalines and isochemical lines do not necessarily identify a plume precisely. 2. The extension of the area seawards from the mouth of the R. Mersey is different in December and March. The length in December 1970 is 32 km and in March 48 km approximately in a north-westerly direction. The greater extent in March does not necessarily mean that there is a higher proportion of Mersey water present but rather that it is being mixed less rapidly with the surrounding water. Considering the isohaline distribution, the dispersion of R. Mersey run-off is in a north westerly direction and is in agreement with Ramster’s (1973) findings for the winter season in Liverpool Bay. The area of the Mersey influence will depend on the hydrography of Liverpool Bay, the rate of mixing and the volume of run-off. From the data obtained, it appears that these conditions were different in December 1970 from March 1971. For example the average discharge from R. Mersey for the three months prior to each survey is greater in the early winter than late winter, i.e. run-off for the periods October-December, 1970 is 94 m3/s and January-March, 1971 is 60 m3/s and for the periods November-December 1970 is 115 m3/s and February-March 1971 is 57 m3/s. Thus the area of the Mersey influence (560 km2 for December 1970 and 952 km2 for March 1971) does not appear to follow the volume of runoff. It is more likely that turbulence and wind-induced mixing play a very important role in the dispersal of R. Mersey water, thus masking the effect, the volume of run-off has on the size of the area of Mersey influence. Turbulence and mixing due to meteorological conditions may be examined by considering the speed and direction of the wind. The average wind speed was calculated from the meteorological records for stations around Liverpool Bay (Valley, Anglesey; Ronaldsway, Isle of Man; and Squires Gate, Lancashire). The wind direction for the periods November 1970 and December r970-February 1971 was predominantly SW-SSW and the average wind speed for these periods was 29.4 km/h and 23.1 km/h, respectively. The area of the Mersey influence plotted from the Si/S%, ratio for December 1970 and March 1971 surveys is found to be approximately inversely proportional to the square of the wind speed (II’) ; the dispersal rate in the surface layer compared graphically is found to be approximately directly proportional to W3 (W is the wind speed, km/h). 3. The width of the area of the Mersey influence (the longshore dimension) in both surveys is found to be approximately the same (285 I km). The width of the area generally signifies the lateral spread across the main direction of the residual current. This lateral spread is most likely to have been produced by tidal effect in the Bay as the direction of the tidal floods (Bowden, 1955; Admiralty Charts) is almost perpendicular to the off-shore orientation of the influence area. The tidal excursion for the near-shore region off R. Mersey was estimated to be IO km (Heaps, 1972). The lateral spread of run-off from a point source due to
tidal movement is, therefore, twice the tidal excursion, i.e. 20 km for the area off the R. Mersey mouth. The observed spread ranged between 27 and 30 km. This difference suggests that although the tidal effect is the most important factor, it is not the only one in the lateral spreading of run-off in near shore waters. Acknowledgements The authors wish to acknowledge the helpful advice and encouragement given by Professor K. F. Bowden and his critical reading of the manuscript. The assistance of the Master and Officers and Crew of R.V. Prince Madog is also acknowledged. References Abdullah, M. I., Dunlop, H. M. & Gardener, D. 1973 Chemical and hydrographic observations in the Bristol Channel during April and June 1971. Journal of the Marine Biological Association of the United Kingdom 53, 299-3 I 9. Abdullah, M. I., Royle, L. G. & Morris, A. W. 1972 Heavy metal concentration in coastal waters. Nature 235, 158-160. Banoub, M. W. & Burton, J. D. 1968 The winter distribution of silicate in Southampton water. Journal du Conseil, Conseil permanent international pour l’exploration de la Mer 32, 201-208. Barrett, M. J., Munro, D. & White, K. E. 1972 Sludge dispersion in sea water. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay, London H.M.S.O. Vol. II, 145-169.
Bien, G. S., Contois, D. E. & Thomas, W. H. 1958 The removal of soluble silica from fresh water entering the sea. Geochimica et Cosmochimica Acta 14 35-54. Bowden, K. F. 1955 Physical oceanography of the Irish Sea. Fishery Investigations, Ministry of Agriculture, Food and Fisheries Series 2, 18, No. 8, 67. Bowden, K. F. & Sharaf El Din, S. H. 1966 Circulation and mixing processes in the Liverpool Bay area of the Irish Sea. Geophysical Journal, Royal Astronomical Society II, 279-292. Dee and Clwyd River Board, Annual Report year ending 1971. Halliwell, A. R. 1972 Sea bed drifter studies. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay, London H.M.S.O. Vol. II, 81-130. Halliwell, A. R. 1973 Residual drift near the sea bed in Liverpool Bay: an observational study. Geophysical Journal, Royal Astronomical Society 30, 439-458. Heaps, N. 1972 Estimation of density currents in the Liverpool Bay area of the Irish Sea. Geophysical Journal, Royal Astronomical Society 30, 415-432. Khan, M. A. & Williams, D. I. 1970 Seasonal changes in the distribution of Chaetognatha and other plankton in the eastern Irish Sea. Journal of Experimental Marine Biology and Ecology 5, 285-303. Jones, P. G. W. & Folkard, A. R. 1971 Hydrographic observations in the eastern Irish Sea with particular reference to the distribution of nutrient salts. Jotrmal of the Marine Biological Association of the United Kingdom 51, 159-182. Mersey and Weaver River Authority 6th Annual Report, year 1970-1971. O’Sullivan, A. J, 1972 Discharges to Liverpool Bay. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay. London, H.M.S.O. Vol. II, 9-15. Ramster, J. W. 1972 Current Measurements. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay. London, H.M.S.O. Vol. II, 57-79. Ramster, J. W. 1973 The residual circulation of the northern Irish Sea with particular reference to Liverpool Bay. Ministry of Agriculture, Fisheries and Food, Fisheries Laboratory Technical Report Series, No. 5, z I. Ramster, J. W. & Hill, H. W. 1969 Current systems in the northern Irish Sea. Nature 224 59-61. Spencer, C. P. 1972 Plant nutrient and productivity study. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay. London, H.M.S.O. Vol. II, 357-401. Talbot, J. W. 1972 Transport and dispersion of soluble material. Department of the Environment Working Party Report on Sludge disposal in Liverpool Bay. London, H.M.S.O. Vol. II, 209-271.