Preview document (🔗 origineel)

---

A  REVIEW ON PAST AND ONGOING RESEARCH IN THE NETHERLANDS WITH REGARD TO
THE COMPLIANCE CHECKING LEVEL OF NITRATE IN GROUNDWATER

REPORT

Introduction

In February 2008 the National Institute for Public Health and the
Environment, RIVM, invited an international review committee, consisting
of:

Dr. Jean-Claude Germon, Research Director INRA Dijon, France

Prof. Dr. Steve Jarvis, University of Exeter, United Kingdom

Dr. Jens Stockmarr, Senior Advisor GEUS, Denmark

Prof. Dr. ir. Oswald Van Cleemput, Ghent University, Belgium (Chairman)

Prof. Dr. Kristine Walraevens, Ghent University, Belgium

Dr. Frank Wendland, Forschungszentrum Juelich, Germany

to examine the scientific background of previous and currently conducted
research on the behaviour of nitrate in groundwater below farmland,
carried out by Wageningen University and Research Centre, Deltares /TNO
and the National Institute for Public Health and the Environment (RIVM).


Mariëlle van Vliet, (Royal Haskoning, The Netherlands) was secretary to
the review committee.

On April 29, 2008 the international review committee members received
all background documents and were asked to provide, before June 1, 2008,
 individually and independently a scientific answer to a number of
research questions (Annex A) related to the documents provided. These
first, independent, responses of the members of the committee were
synthesized (Annex B) and presented to the members of the review
meeting.  

The international review committee met with the research teams on June
11-12, 2008 at Bilderberg Hotel “de Klepperman” in Hoevelaken, to
discuss the research conducted and described in the following project
reports:

Broers, H.P., Griffioen, J., Willems, W.J., Fraters, B. (2004). Should
the test depth for nitrate in groundwater be changed? Background
document for evaluation of the 2004 Fertilizer Act. Nederlands Instituut
voor Toegepaste Geowetenschappen, Utrecht, TNO-rapport NITG 04-066-A. 

Fraters, B. Boumans, L.J.M., Van Elzakker, B.G., Gast, L.F.L.,
Griffioen, J., Klaver, G.T., Nelemans, J.A., Velthof, G.L., Veld, H.
(2006). A new compliance checking level for nitrate in groundwater?
Final report of the Feasibility study on monitoring the upper five
metres of groundwater, RIVM, report 680100005/2006.

In addition, the following ongoing projects were discussed:

Investigation of groundwater and sediment characteristics for shallow
subsurface, Deltares/TNO.

Modelling nitrate in groundwater with emphasis on a compliance checking
level including field study, Wageningen University and Research Centre.

The members of the international review committee appreciated the
efforts of the research teams in producing written and oral
presentations of high quality and clarity.

The reports and the presentations provided the means for a complete and
stimulating discussion on the important and difficult issue of lowering
the compliance checking level for nitrate in groundwater. 

Review conclusions

The objectives of the review were to determine if: 

the research was fit to provide answers to the research questions as
listed in Annex A, 

the questions posed were adequately answered, and

the conclusions were supported by the scientific data provided.

With regard to the past research, the review committee arrived at the
following conclusions, based on scientific considerations: 

The reports give robust and scientifically based answers and come to
sound conclusions.

The design of the studies was appropriate for the problems as defined.

The question of the most appropriate depth for sampling was examined
very carefully.

The questions with regard to the requirements of sampling for water
quality have been answered within the framework of the projects.

The review committee supports the distinction that was made between the
dry soils as compared with the neutral and wet soils. 

For the dry soils, no decrease in nitrate concentration was observed in
the first five metres below the water table. Therefore the option for
lowering the compliance checking level is not relevant.

In the neutral and wet soils, a decrease in nitrate concentration in
groundwater was observed with increasing depth. However, these soils are
drained, which will directly lead to nitrate inputs into surface water.
In order to limit this, the review committee supports the conclusions of
the research teams that the option of lowering the compliance checking
level is not justified.

The review committee is convinced that additional measuring efforts will
not lead to different conclusions. However, the review committee sees
opportunities for further research to improve confidence in, and
knowledge of, the fate of nitrate.

The review committee supports the conclusions of the above reports that,
given Dutch conditions, the present compliance checking level provides
direct feedback mechanisms for fertilizer use standards and land
management.

As well as the direct effects of nitrate pollution, the possible
problems of undesirable side-effects may be very important. The review
committee agrees that attributing concentrations of sulphate, nickel,
arsenic etc. to side-effects of denitrification will always be
difficult.

With regard to the ongoing field and model studies, the review committee
came to the following conclusions:

The research approach provides a scientifically sound means of providing
answers to the overall research questions, given the current results and
the proposed follow up.

The STONE model has high potential and versatility. The first results as
demonstrated to the review committee are promising. However, the review
committee suggests that there may be some refinements that will help its
future development:

The presentation of the preliminary results should be rearranged to
improve and clarify the outcome of the model. 

The review committee was not clear to what extent the hydrodynamic
status of the simulated sediment column is taken into account (e.g.
lateral inflow, lateral outflow: flow distance to surface waters,
residence time, denitrification along lateral flow paths, etc.).

The review committee was satisfied with the extent and scope of the
field studies and considered the preliminary results to be very
promising.

With the available information the review committee was unable to judge
whether some of the specific research questions as listed in Annex A
(appendix 3) will provide definitive answers.

The review committee believes that the new information provided by both
ongoing model and field studies does not change the conclusions of the
previous studies (Broers et al., 2004 and Fraters et al., 2006) that
there are no scientific arguments for lowering the compliance checking
level.

Recommendations

1. For the requirements of the Nitrate Directive the review committee
did not find scientific arguments to change the current compliance
checking level.

2. In view of the requirements of the Water Framework Directive and
Groundwater Directive for assessing the chemical status of waters, the
review committee recommends a hydrogeological/ hydrogeochemical mapping
with respect to nitrate reduction capacity and depths. Therefore all
groundwater monitoring network data including nationwide available
‘geodata’ should be used to delineate spatial ‘geotop regions’
to indicate where and at what depth denitrification occurs. The ongoing
research is contributing to the above.

3. The review committee recommends that the STONE model should be
further developed to aid the prediction of surface water quality and
assess ecological status, as required by the Water Framework Directive.

June 17, 2008



ANNEX A

LIST OF QUESTIONS

Appendix 1: Research questions and conclusions from Broers et al (2004)

Appendix 1.1 Research questions and conclusions

General Review Question: Does the set-up of the study fit to the problem
definition?

Sub-question 1: Is it possible to identify with a sufficient degree of
accuracy areas where denitrification without adverse side-effects
occurs?

Conclusion: It appears clearly from chapter 3 that it is possible to
detect and quantify denitrification on a measuring-location scale. The
situation is however quite different on a regional or sub-regional
scale. The spatial delimitation of regions where denitrification occurs
demands a great deal of cartographic and measuring effort. 

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 2: And if so (continuing question 1), at what depths should
measurements be carried out as a basis for such identification?

Conclusion:  If the compliance checking levels are increased in selected
areas, it would be advisable to keep them in line with the measurement
depths currently used in the national and provincial measurement
networks. In view of the distribution of travel times found in Dutch
infiltration zones, it would be preferable not to choose the compliance
checking level at more than 10 metres under ground level, a depth at
which the measured concentrations relate to water than entered the soil
about 10-15 years before.

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 3:  What are the pros and cons of increasing the compliance
checking level?

Conclusions about pros and cons:

The main disadvantage of increasing the test depth to e.g. 10 metres is
that it precludes fast feedback between measuring data and fertilizer
consumption. In other words, there is then no direct relation between
the nitrate concentration at this depth and the loss standard or use
standard used in previous years. If fast feedback is required, which is
considered necessary for the monitoring objective of the setting of
fertilizer use standards, then increasing the test depth is not an
option.

For other monitoring objectives, including the evaluation of effects of
fertilizer and manure practice on the environment, an increase of the
evaluation depth is not to be recommended in well drained areas, because
the shallow groundwater has an important influence on surface-water
quality. 

In areas where there is no direct relationship between surface-water
quality and groundwater quality, monitoring of the groundwater at depths
of 10 and 25 m should be acceptable. For these situations:

Steps would also have to be taken to show with a sufficient degree of
reliability that denitrification:

is a relevant process in the area in question;

has no adverse effects;

is sustainable.

The extent to which a direct link exists between groundwater and
surface-water quality in these areas of leachable soils should also be
investigated.

A test methodology taking spatial variations in age build-up and
denitrification capacity into account needs to be worked out.

Review question: Is the research question answered and do you support
the conclusion(s)?

Appendix. 1.2 Question about the specific study ‘current knowledge of
the behavior of nitrate’

Gives Chapter 3 a correct description of the ‘current’ knowledge of
the behavior of nitrate in the subsoil (in the time of the study at
2004)?

Appendix 2: Research questions and conclusions from Fraters et al (2006)

Appendix 2.1 Research questions and conclusions

General Review Question: Does the set-up of the study fit to the problem
definition?

Sub-question 1: Does the nitrate concentration decrease with depth in
agricultural lands in the sandy regions? If so, what is the extent of
this decrease and are there differences between regions?

Conclusion: There is no indication that nitrate levels decrease with
depth in the dry sand regions (see for example figure S1). Data from
studies performed in 2004 and 2005 support this conclusion. In a number
of wells (mainly in dry sand) the nitrate concentration increased with
depth. In the neutral and the wet sandy regions, the nitrate
concentration did decrease on average within the first five metres of
the groundwater column. The extent of the decrease varied between 15 and
40 % in the neutral areas and between 30 and 100% in the wet areas.  

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 2: If the nitrate concentration decreases within the upper
five metres, can this be attributed to denitrification, or are there
other causes for this decrease?

Conclusion: In neutral and wet sandy soils, the nitrate concentration
does drop between the first and fifth metre of the groundwater column.
This decrease is probably partly due to denitrification. However, we
expect that a number of other causes also contribute to this decline,
such as hydrological aspects (interfering soil layers), regional upward
seepage from underlying water layers, and drainage of part of the
precipitation surplus to nearby surface waters.

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 3: If denitrification occurs, to what extent does this
process lead to adverse environmental effects such as an increase in the
levels of sulphate or heavy metals, or increasing hardness of the water?

Conclusion: It was not possible to quantify the increase in other
substances as a result of denitrification. Because of the heterogeneity
of the subsoils in the Dutch sandy regions, there is considerable
spatial variation in the rate of denitrification and in the effects that
denitrification has on other aspects of groundwater quality.  

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 4: In the clay and peat regions, lowering the compliance
checking level in order to be able to apply less strict application
standards would lead to insufficient reduction of the nitrate load to
surface waters. To what extent does this apply to the sandy soil areas?

Conclusion: If the compliance checking level for sandy soils is lowered
then concentrations exceeding 50 mg l-1 will be accepted in the first
meter. This implies allowing higher concentrations in the drain water.
This will lead to nitrate concentrations in tile drain water exceeding
11.3 mg l-1 nitrogen (the objective used to support the derogation).
Surface water concentrations may be three or four times as high as the
target value of 2.2 mg l-1 nitrogen.

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 5: ‘Regardless of how complex, is it feasible to measure
the nitrate concentration at a depth of five metres in the saturated
zone, or can the average nitrate concentration in the first five metres
be determined?’

Conclusion: If the compliance checking level is lowered, monitoring data
on nitrate will also be needed to check compliance. It is possible to
routinely measure nitrate concentrations at a number of depths in the
first five metres of the saturated zone. Such measurements cost more
than those in the top level of groundwater.

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 6: Regardless of how complex, is it feasible to measure the
extent to which denitrification occurs, whether it is sustainable, and
whether the process produces undesirable by-products?

Conclusion: It is not feasible to measure the extent to which
denitrification occurs, the sustainability or whether undesirable
by-products (e.g. sulphates, heavy metals) result using routine
measurements such as those used in monitoring programs or networks.
Using a combination of existing non-routine methods it is possible to
determine the extent of denitrification and the presence of undesirable
by-products at specific locations. One unavoidable uncertainty is that
other soil processes may produce the same undesirable compounds, so that
the extent to which denitrification leads to a deterioration in
groundwater quality cannot be accurately established.

Review question: Is the research question answered and do you support
the conclusion(s)?

Sub-question 7: Regardless of how complex, is it feasible to measure
whether the groundwater at five metres depth has the same origins
(infiltrated at the same place) as the groundwater in the first metre? 

Conclusion: It is not feasible to measure whether groundwater at five
metres depth has the same origins as groundwater in the first metre, or
at least not using routine measurements in a monitoring program or
network. By using a combination of methods currently not used in routine
programs it is possible to determine whether groundwater at five metres
depth infiltrated at the same place as groundwater in the upper metre of
groundwater. 

Review question: Is the research question answered and do you support
the conclusion(s)?

Appendix. 2.2 Specific questions for the field and laboratory studies

A field study at the LMG multi-screen wells was carried out by RIVM in
December 2004. Samples were taken to obtain additional information on
the development of the nitrate concentration with depth (see chapter 2).


Are sufficient data available and are the data reliable enough to
substantiate the conclusions? 

Chapter 3 describes the water quality study on farms. The objective of
this field study was to examine the change in the nitrate concentration
in the upper five metres of the groundwater using measurements taken in
the plot. 

Does the set-up of the study fit the objective?

Are sufficient data available and are the data reliable enough to
substantiate the conclusions?

The objective of the considered field study (see chapter 4) was to test
methods for sampling the upper five metres of groundwater under field
conditions and to compare the methods with one another in terms of the
technical aspects of the implementation and the quality. 

Does the set-up of the study fit the objective?

Are sufficient data available and are the data reliable enough to
substantiate the conclusions about technical aspects as well as quality
aspects? 

Soil samples were also taken for the denitrification study. The
potential denitrification was determined by Alterra (see chapter 5).

Are sufficient data available and are the data reliable enough to
substantiate the conclusions about potential denitrification? 

Chapter 6 describes the study about the presence of reactive components
in the soil as a factor in the denitrification of nitrate (by TNO).

Does the set-up of the study fit the objective?

Are sufficient data available and are the data reliable enough to
substantiate the conclusions?

Appendix 3: Research questions from current denitrification study

General question:

Is the research method scientifically sound for answering research
questions, given current results and proposed follow up?

Question 1: What is the average regional change of the nitrate
concentration in groundwater between groundwater surface and five metres
below groundwater surface for agriculture areas with a constant level of
fertilization? 

Regionals are specified as North, Middle and South sand areas
distinguished by different water table classes (dry sand, medium dry
sand and wet sand). 

Please indicate:

which part of the change is due to denitrification?

wich part of the change is due to hydrological phenomena (seepage,
dilution)

the average decrease and the uncertainty in the decrease

Question 2:  On the scale of areas, what are the expected side-effects
of denitrification? Indicate the expected change in concentration for
heavy metals, sulphate and increase of hardness of water. Furthermore
specify the uncertainty in the change.

Question 3:  What is the expected regional development of
denitrification capacity in time? (Indicate at least increase, steady
state or decrease and the time frame?)

Question 4: Specify the present nitrogen load to surface waters caused
by agriculture activities on regional scale? What percentage originates
from nitrate in groundwater (beneath agricultural areas) at a depth
between the groundwater table and five metres below groundwater table?

Question 5: What is the present nitrogen concentration in groundwater
beneath agricultural areas leaching directly to surface waters? What
percentage originates from nitrate in groundwater at a depth between
groundwater table and five metres below groundwater table on
agricultural lands?

ANNEX B

MAIN LINES OF ANSWERS TO THE QUESTIONS (summarized)

Broers et al. (2004):

Report gives scientifically based answers and comes to sound
conclusions.

Question of the most appropriate depth for sampling is reviewed very
carefully.

Answers with regard to requirements of water quality sampling have been
answered within framework of the project.

Identification of delimitations of regions (nitrate-reducing areas) is
possible. Discussion about:

A lot of effort need?

Comparable general characteristics?

Bigger the scale, the more difficult?

Prefer compliance checking level not deeper than 10 m under ground
level.

If fast feedback is required, which is considered necessary for the
monitoring objective of the setting of fertilizer use standards, then
increasing the test depth is not an option.

For other monitoring objectives an increase of the evaluation depth is
not to be recommended in well drained areas, because the shallow
groundwater has an important influence on surface-water quality. 

Chapter 3 gives a correct description of the ‘current’ knowledge of
the behaviour of nitrate in the subsoil

Question(s) from reviewer(s):

Is it really efficient and necessary to develop a test methodology,
which takes spatial variations in age build-up and DN capacity in
account?

The assumption here, and in the other report, is that piston type flow
predominates and this is probably the case but even in the situation of
very well drained soils, some bypass flow sometimes occurs. Has this
ever been examined in these soils by using tracers etc? 

Fraters et al. (2006):

The set-up of the study fits to the problem definition.

Reviewers support the conclusions within the constraints of the
resources.

It would always be possible to do more with more resources.

To attribute concentrations of substances of the undesirable effects of
denitrification will always be doubtful and displays a lot of
uncertainty.

Tracers and models can give more information.

Number of data (various opinions):

Don’t expect a better result with more data

The number of multilevel wells and farms is too few to be generalized
for all sandy soil types occurring in the Netherlands.

Data collection throughout the profile will certainly give better
results than only 5 metres (cost benefits analysis)

 PAGE   1 

 PAGE   9