Iahr 2007
Iahr 2007
Iahr 2007
ABSTRACT
The present work proposes a methodology to monitor and predict river water pollution in
regard to the requirements posed by the European Union (EU) Water Framework Directive
(WFD) 2000/60/EC and taking into consideration current peculiarities of the Greek status (water
management diffused in several authorities, multiple stakeholder conflicts, irregular and
inadequate pollution monitoring programs, low financial resources e.t.c.). Particular cost-
effective tools of quick performance for either predicting or measuring river pollution have been
identified. The Erymanthos River watershed (360 km2), which is a sub-basin of the Alfeios River
basin, Greece, was selected as a case-study. The total nitrogen and phosphorus loads of the
watershed occurring during each season of the year were estimated on the basis of typical inputs
due to municipal and agricultural land uses met in the study area for the period 1999-2001. The
river discharge was simulated using a rainfall–runoff model calibrated for the period October
1963 – September 1976 with daily rainfall data and sufficient discharge data. Simulated
discharge values in monthly basis were used to compute the necessary mean seasonal discharge,
its standard deviation, as well as the minimum and maximum values, and then the seasonal
values of the pollution loads of total phosphorus and total nitrogen were estimated. Finally, the
relevant pollution factors were calculated as the ratios of the corresponding watershed loads and
river-transported loads. During the year 2006, four expeditions (one per season) were made for
direct discharge and concentration measurements to allow direct computation of the related
pollution loads transported in the river and subsequent pollution factors. The discharge was
determined by employing quick measurement techniques combined with the logarithmic–
parabolic velocity distribution. Present findings show justifiable behaviour and could be used as
preliminary results in incoming river pollution monitoring and watershed management programs
imposed by the WFD.
1 INTRODUCTION
The Water Framework Directive (WFD) 2000/60/EC (WFD, 2000) is an important tool of
the European Union (EU) aiming to harmonize a sustainable water resources management among
2
member states by establishing a framework for the protection of inland surface waters,
transitional waters, coastal waters and groundwater. The formal compliance of Greece with the
WFD was made by adopting Law 3199/03 (Official Gazette 280A/2003) for water protection and
management. The main environmental target of the WFD is the achievement of a good ecological
potential and a good water quality within 15 years after formal compliance (in 2018 for Greece).
Therefore, it is necessary to monitor the ecological and chemical status of surface waters, the
chemical and quantitative status of groundwaters, in addition to the specifications of the protected
areas under which these areas have been established. Monitoring must cover physico-chemical,
hydro-morphological, biological and chemical parameters (Allan et al., 2006). The monitoring
programs must be introduced by December 2006 and include Surveillance, Operational and
Investigative monitoring modes (US-EPA, 2005). A simplified scheme has been given by Allan
et al. (2006), who have furthermore reviewed emerging biological and chemical monitoring tools
that may be incorporated in the techniques for water quality assessment. The decision support of
water resources in river basins following WFD must be based on the pollution loads estimations.
The several natural phenomena and anthropogenic activities occurring in a river basin
constitute pollution sources with subsequent pollutant emissions. The surface-water pollution
sources are generally distinguished in two types: (a) Point sources, where the pollution loading is
made at a well defined location or an area of limited extent compared to the watershed size or the
river length. Such sources include mostly wastewater discharges by sewerage and biological
treatment systems for domestic, municipal, industrial, agro-tourist and other installations, as well
as livestock wastewater discharges. (b) Non-point sources (NPS), where the pollution loading is
linearly or area diffused (along river banks and groundwater loadings by infiltration). NPS–
pollution loading is caused by runoff moving over and through the ground, carrying natural
(wildlife or geologic pollutants) and man-made pollutants from illicit discharges of residential
and industrial wastes, roadways, grassed areas, cropland, pastures, livestock operations and dry or
wet deposition of air pollutants and finally depositing them into surface waters. The NPS
pollution contributes mainly to nutrients, pesticides and sediments. Both point and NPS
pollutants contribute to surface-water pollution and cause eutrophication.
The present work aims to (a) propose convenient monitoring programs to be applicable in
the specific types of Greek river basins in regard to the complicated frame of authorities and
stakeholders involved, and (b) identify particular cost-effective tools of quick performance for
either predicting or measuring the river pollution. Such tools may be incorporated in river basin
management plans, as it has primarily indicated by Yannopoulos (2005). General descriptions of
pertinent management plans for river basins of particular complexity, as the Alfeios River Basin,
are given by Manariotis and Yannopoulos (2004, 2006). In order to highlight the most
appropriate tools for the aforementioned purposes, a case-study is additionally carried out herein.
2.1 METHODS
Pollution load of a specified pollutant is the quantity of mass emitted per time unit from
point or non-point sources of a defined area. Each pollution load in river waters is computed by
multiplying the river discharge with the pollutant concentration measured at a prescribed river
cross-section. The methods for estimation of the NPS pollution can be classified in two main
categories: a) Direct approach and b) indirect approach. The direct approach addresses directly to
NPS pollution and the events and causes that contribute to NPS pollution are mathematically
described (Hartigan et al., 1983). On the other hand with the indirect approach, the NPS pollution
3
is correlated to water quality data available for surface bodies. Several methods have been
developed for the assessment of NPS pollution and include computer-based models for the
analysis of the water quantity and quality of river basins. Recently Geographic Information
Systems (GIS) are in use to facilitate data processing and management (Hsieh and Yang, 2006).
Integrated GIS models, which include data analysis and modelling, have been developed for the
assessment of point and NPS pollution (BASINS) (US-EPA 1997, Matejisek et al., 2003; Hsieh
and Yang, 2006). Furthermore, control strategies have been developed for the management or
nutrient pollution for watersheds and surface waters (Peters, 1973; Chapra and Tarapchak, 1976;
Haith and Dougherty, 1976; Somlyody and Wets, 1998; Yeh and Labadie, 1997; Harrell and
Ranjitham, 2003).
T1 Tj
K2 Ki
K0 K1 T
Sea
2
Tm
Kn
LEGENDE
Main watercourse
Tributary
Boundary of the river basin
Boundary of sub-basins
Watershed outflow
GIS Management
Database update
Land uses
Water uses
Statistical data treatment
Fig. 3. A general aspect of Alfeios River basin highlighting the Erymanthos watershed.
6
0 5 10 15 20 25 30
Percentage (%)
Fig. 4. Distribution of the land uses of the Erymanthos watershed for the year 2000.
Source: HEMCO (http://www.okxe.gr).
In addition, for the same period the relevant livestock categories provided are cow, sheep,
goat, swine, horse/donkey, rabbit, and poultry. Population data (census 2001) were used to
estimate wastewater loads, which contribute to total nitrogen and phosphorus in the watershed.
According to data available by Prefectures, agro-industrial activities in the area under
examination constitute a rather small number of olive-oil mils, while other industries do not exist.
Table 1. Seasonal loads (in 103 kg) of total nitrogen and phosphorus in the Erymanthos River
watershed due to anthropogenic activities
Table 2. Erymanthos river discharge and pollutant concentration measurements concerning total
phosphorus (P) and nitrate nitrogen (NO3-–N) plus ammonium nitrogen (NH4+–N)
Winter Spring Summer Autumn
Parameter
(8 Jan. 2006) (9 Apr. 2006) (24 Aug. 2006) (17 Nov. 2006)
Discharge (m3/s) 23.06 11.65 2.63 4.14
Total P (μg/l) 67 26 20 105
- +
NO3 –N plus NH4 –N (μg/l) - 190 520 531
The values in Table 2 should be considered as preliminary seasonal values of these pollutants, as
no other measurements are currently available. However, according to Kirchner et al.’s (2000)
implications, contaminants are initially flushed rapidly from the watershed, but then low-level
contamination is delivered to streams for a surprising long time. Chemical variations on
timescales that are long compared to the travel time distribution will be transmitted through the
watershed without significant attenuation. In that sense, since travel times in the Erymanthos
watershed are of order of one day and chemical-variation time scales are of order of one season
of the year, it is expected that the one day concentration measurement should represent fairly well
the average situation of an extended past period.
the routing of surface runoff at the outlet of the basin. The input data required are daily
precipitation and temperature. The calibration is performed using daily discharge measurements.
Daily precipitations data have been measured at a rain station located near the centre of the basin
(Fig. 3) and are available for the period 1955–1997. Temperature data for Erymanthos watershed
are not available. Instead, mean monthly values of temperature were used, measured through the
periods 1966–1968, 1974–1976 and 1981–1988 at Ladhon dam located in a nearby basin,
southeast of the centre of Erymanthos basin. Data preparation included completion of the
temperature time series for the precedent time period and reduction to new values, suitable for the
altitude of Erymanthos watershed’s centre. The completion of temperature time series was based
on the average monthly temperature computed from all available data from Ladhon basin. These
values were measured at an altitude of 430 m. A gradient of -0.6 oC/100m was used for the
reduction of the temperature values to the altitude of 838 m (centre of Erymanthos watershed).
The resultant monthly values were used as daily temperature input data for the model.
Measurements of daily discharge of Erymanthos River are available for sufficient intervals
of the period 1964–1976 and have been carried out 8 km upstream from the basin’s outlet (Fig.
3). Thus, the simulated sub-basin is slightly smaller with an area of 329.5 km2. The simulation
results of basin discharge have been transformed to correspond to Erymanthos’ watershed with an
area of 360 km2, using a factor f = 360/329.5 = 1.092.
The comparison between measured and simulated discharge for the model calibration has
been performed using the correlation coefficient of linear regression between measured and
simulated values of river discharge at the outlet point and the root mean square error of simulated
values. In addition, annual values of actual evapotranspiration were calculated using the Turc
method and compared to the simulated values, serving as a third criterion. Figure 5 shows the
measured values of river discharge in comparison with those resulted from the calibrated model.
The correlation coefficient of linear regression between measured and simulated values is 0.88,
while the root mean square error 3.22 m3/s. The average annual actual evapotranspiration
computed with the Turc method, for the 13-year period from 01/10/1963 to 30/09/1976, is 578
mm/year, while the corresponding value resulted from simulation is 521 mm/year. Figure 6
shows the seasonal values of discharge calculated from the results of the simulation for the period
1963–1996.
35
30
Measured values
Daily discharge (m3/s)
25 Simulated values
20
15
10
0
21/03/1964
21/09/1964
21/03/1965
21/09/1965
21/03/1966
21/09/1966
21/03/1967
21/09/1967
21/03/1968
21/09/1968
21/03/1969
21/09/1969
21/03/1970
21/09/1970
21/03/1971
21/09/1971
21/03/1972
21/09/1972
21/03/1973
21/09/1973
21/03/1974
21/09/1974
21/03/1975
21/09/1975
21/03/1976
Fig. 5 Measured values of river discharge versus those resulted from the simulation.
9
25
Winter
Spring
20
Summer
Autumn
Discharge (m3/s)
15
10
0
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
Fig. 6 Seasonal values of discharge, calculated from the results of the simulation for the period
1963–1996.
The mean seasonal values of the river discharge for this period, their standard deviation, as
well as the minimum and maximum values, are given in Table 3. The comparison with the values
measured (Table 2) shows that the latter are between the minimum and maximum values
predicted and in the range of two standard deviations of the simulation results.
Table 3. Simulated mean seasonal values and standard deviations of river discharge for the
period 1963–1996
Parameter Winter Spring Summer Autumn
3
Mean discharge (m /s) 10.02 8.88 1.42 2.35
Minimum/maximum (m3/s) 0.63/23.76 0.15/18.90 0.15/3.55 0.32/8.74
3
Standard deviation (m /s) 6.29 4.72 0.66 1.86
10
0.5
CP computed
0.4 CP measured
CN computed
C P, C N 0.3 CN measured
0.2
0.1
0.0
Winter Spring Summer Autumn Year
Fig. 7. Annual variation of pollution factors CP and CN, regarding the contribution of the
Erymanthos River watershed to the river waters. Computed values come from modelled
runoff using precipitation data from 1963 to 1996, while measured values come from
direct measurements of the river discharge, one time within each corresponding period of
the year 2006. Upper and lower bars indicate the calculated extreme values of factors
corresponding to minimum and maximum discharge values predicted.
The values of pollution factor regarding N, CN, show considerably lower rates of
contribution during spring and autumn compared to CP, while comparable rates occur during
summer. Unfortunately, the river water sampled in the winter expedition had not been analysed
for N content and therefore no value can be estimated for N contribution during the winter period.
The approximately comparable rates of pollution factors, CP and CN, during summer may be
attributed to the low precipitation heights, which cause very limited runoff. The P and N
contribution to the river water originates from groundwater, which is less contaminated by these
constituents that runoff. The observed lower values of CP compared to CN may be attributed
mainly to the different retention rates in the watershed. An indication may be taken by Kronvang
et al. (1999), who measured that the organic N retention in stream bed sediments and riparian
zones during a low-flow period was approximately 2.6 times higher than the corresponding P
retention. In addition the N content of soil in subsurface samples (20-25 cm depth) was about 20
times higher than P content (Craft and Chiang, 2002).
4 CONCLUSIONS
The Erymanthos watershed case-study showed that the proposed methodology provides
reasonable results for the pollution factors estimations. Due to the iterative procedure proposed
between pollution monitoring and assessment of pollution factors, the methodology gains the
advantage of the gradual improvement of the results. Authors believe that it could be
incorporated in the incoming monitoring and management plans required by WFD.
ACKNOWLEDGMENTS
We thank the European Social Fund (ESF), Operational Program for Educational and
Vocational Training II (EPEAEK II), and particularly the Program PYTHAGORAS II, for
funding this work. Thanks to Dr. G. Ch. Miliaresis for providing the digital terrain model of the
11
Erymanthos basin, as well as to Hellenic Public Power Corporation for providing part of the
meteorological data used.
REFERENCES
Allan, I.J., Vrana, B., Greenwood, R., Mills, G.A., Roig, B., and Gonzalez, C. (2006), A
“toolbox” for biological and chemical monitoring requirements for the European Union’s
Water Framework Directive, Talanta 69, 302-322.
Chapra, S.C., and Tarapchak, S.J. (1976), A Cholorophyll-a model and its relationship to
phosphorus loading plots for lakes, Water Resources Research 12 (6), 1260-1264.
Cho, J.C., Sung, K.S., and Ha, S.R. (2004), A river water quality management model for
optimizing regional wastewater treatment using a generic algorithm, Journal of
Environmental management 73, 229-242.
Craft, C.B., and Chiang, C. (2002), Forms and amounts of soil nitrogen and phosphorus across a
longleaf pine-depressional wetland landscape, Soil Science Society of America Journal 66,
1713-1721.
Haith, D.A., and Dougherty, J.V. (1976), Nonpoint source pollution from agricultural runoff,
Journal of Environmental Engineering Division-ASCE, 102 (EE5), 1055-1069.
Harrell, L.J., and Ranjithan, S.R. (2003), Detention pond design and land use planning for
watershed management, Journal of Water Resources Planning and Management 129 (2), 98-
106.
Hartigan, J.P., Quasebarth, T.F., and Elizabeth, S. (1983), Calibration of NPS model loading
factors, Journal of Environmental Engineering Division-ASCE, 109 (6), 1258-1272.
Hsieh, C.D., and Yang, W.F. (2006), Study of Total Maximum Daily Load and Nonopoint
Source Pollution Control Strategies for Reservoir Watershed, Practice Periodical of
Hazardous, Toxic, and Radioactive Waste Management-ASCE, 10 (3), 134-143.
Kirchner, J.W., Feng, X., and Neal, C. (2000), Fractal stream chemistry and its implications for
contaminant transport in catchments, Nature 403 (3), 524-527.
Kronvang, B., Hoffmann, C.C., Svendsen, L.M., Windolf, J., Jensen, J.P., and Dørge, J. (1999),
Retention of nutrients in river basins, Aquatic Ecology 33, 29-40.
Manariotis, I.D., and Yannopoulos, P.C. (2004), Adverse effects on Alfeios River Basin and an
integrated management framework based on sustainability, Environmental Management 34
(2), 261-269.
Manariotis, I. D., and Yannopoulos, P. C. (2006), Integrated management plan for the Alfeios
River Basin, in CD-ROM Proceedings of the 2nd International Congress “AQUA2006 -
Water Science and Technology - Integrated Management of Water Resources”, Athens -
Greece, 23-26 November 2006, 6 pages.
Matejicek, L., Benesova, L., and Tonika, J. (2003), Ecological modelling of nitrate pollution in
small river basins by spreadsheets and GIS, Ecological Modelling 170, 245-263.
Nachtnebel, H.P., Lettl, W., Baumung, St., (1993), Abflußprognosemodell für das Einzugsgebiet
der Enns und der Steyr (Handbuch), Institut für Wasserwirtschaft, Hydrologie und
konstruktiven Wasserbau, Wien, Austria.
Peters, R.H. (1973), Nutrient balances for the evalution of nutrient sources in water quality
management, Water Resources Bulletin 9(1), 49-53.
Somlyody, L., and Wets, R.J.B. (1998), Stochastic optimization models for lake eutrophication
management, Operational Research 36 (5), 660-681.
WFD (Water Framework Directive) (2000), Directive 2000/60/EC of the European Parliament
and of the Council of 23 October 2000 establishing a framework for Community action in
12
the field of water policy, Official Journal of the European Communities, L327, 22.12.2000,
Vol. 43, 1-72.
US-EPA (1997), Compendium of tools for watershed assessment and TMDL development. EPA
841-B-97-006, United States Environmental Protection Agency, Washington, DC.
US-EPA (2005), Handbook for Developing Watershed Plans to Restore and Protect Our Waters.
Draft, Document No. EPA 841-B-05-005, Nonpoint Source Control Branch, Office of
Water, United States Environmental Protection Agency, Washington.
www.epa.gov/owow/nps/pubs.html
WMS (1998), Watershed Modeling System, Environmental Modeling Systems Inc., USA,
http://www.ems-i.com.
Yannopoulos, P.C. (2005), A methodology for monitoring and quickly predicting river pollution
related to 2000/60/EC Directive, Hydrotechnika, Journal of the Hellenic Hydrotechnical
Association 15 (December, Special edition), 103-112 (in Greek).
Yannopoulos, P.C., and Manariotis, I.D. (2005), Adverse effects on the hydro-morphology of the
Alfeios River by infrastructure works and gravel extractions, in Proceedings of the 5th
National Congress of the National Committee for Water Resources Management “Integrated
Water Resources Management”, Xanthi - Greece, 6-9 April 2005, pp. 275-281 (in Greek).
Yeh, C.H., and Labadie, J.W. (1997), Multiobjective watershed-level planning of stormwater
detention systems, Journal of Water Resources Planning and Management 123 (6), 336-343.