FORUMThe Information Cycle as a Framework for DefiningInformation Goals for Water-Quality MonitoringJOS G. TIMMERMAN*JANNEKE J. OTTENSInstitute for Inland Water Management and Waste Water

Treatment RIZAP.O. Box 178200 AA Lelystad, The Netherlands

ROBERT C. WARDColorado Water Resources Research Institute410 N. University Services CenterColorado State UniversityFort Collins, Colorado 80523, USA

ABSTRACT / The necessity to tailor information becomesincreasingly urgent as the information revolution continues togenerate ever-increasing flows of data and so-called infor-mation. From European experiences, a new approach formonitoring system design is suggested in this paper. In this

approach, careful and detailed specification of informationneeds is a major contributing factor to the effectiveness ofinformation products. To develop better specifications forinformation products, the process of collecting and trans-forming data into useful information requires careful thoughtand guidance. A dialogue between information users on onehand and information producers on the other is essential.This dialogue can be based on the information cycle, de-scribing the continuous process from specifying informationneeds for water management and a strategy to collect infor-mation through data collection and data analysis up to utili-zation of information by water management. By following therespective steps in the information cycle, the process of in-formation gathering can be completed. The cyclic characterprovides a quantitative means of connecting monitoring sys-tem design and operations with the information expectationsand/or products required by management.

Over the last two decades, water management haschanged from a focus on distinct sectors, dealing withone or two interests at a time, to integrated watermanagement, interrelating different aspects of the wa-ter system like physical planning, ecology and emis-sions. In the future, comprehensive water managementwill integrate various water management functions intoan interactive framework of ecology, sociology, andeconomy (De Jong and others 1994, Raskin and others1997). In addition, increasing knowledge of the com-plexity of processes in water systems has led to a growingdemand for information. Water system problems arebecoming more complex and diverse, and require morespecific knowledge and integration across various disci-plines, sectors, countries, and societies (Somlyody 1994).This growing need to integrate disciplines also bringsabout the need to integrate large sets of data andinformation. Moreover, computers provide numerouspossibilities to store, retrieve, and analyze data, thusenlarging the availability of data and information.Nowadays, policy-makers and water managers are over-whelmed with data and information that may or maynot be of use to them. Ward and others (1986) de-

scribed this ‘‘data-rich but information-poor syn-drome.’’ Therefore, there is a call today for less quantityof information and more targeted, tailor-made, informa-tion (see, e.g., Ten Brink and Woudstra 1991, Adriaanseand others 1995, Ward 1995, Adriaanse and Broseliske1998, Ongley 1998).

Information is defined as the communication orreception of knowledge or intelligence in Webster’sNinth New Collegiate Dictionary published in 1983. Asin the above definition, the word ‘‘information’’ in thispaper is used in a broad sense of exchanging materialthat is more than mere data or facts. Production ofinformation in this paper therefore implies collectingand processing data in such a way that the recipient ofthe information is able to draw a conclusion from it.

In this paper, a new way of approaching informationcollection is suggested, specifically emphasizing waterquality monitoring system design. This approach hasevolved from two international conferences on monitor-ing (Monitoring Tailor-made, 1994 and 1996), fromexperiences in drawing guidelines for water-qualitymonitoring and assessment of transboundary rivers(Timmerman and others 1997), and from Dutch experi-ence in designing and optimizing monitoring networks(e.g., Adriaanse 1993). The paper presents and dis-cusses a framework that assists information producers indeveloping tailor-made information that is sized to fit

KEY WORDS: Information needs; Water-quality monitoring; Monitor-ing systems design

*Author to whom correspondence should be addressed.

DOI: 10.1007/s002679910018

Environmental Management Vol. 25, No. 3, pp. 229–239 r 2000 Springer-Verlag New York Inc.

the needs of the information users. The next twosections will continue with considerations on difficultiesin formulating questions and devising strategies toobtain answers. In the next section some examples ofmonitoring system design strategies from literature arediscussed, and the proposed framework is presented.The following sections go into detail on application ofthe framework and its respective steps.

What is the Question?

The major challenge in monitoring and assessmentis making the information obtained fit the informationneeded (Adriaanse and others 1995). ‘‘All too often,monitoring projects are initiated with a minimum offorethought, and result in a collection of poorly docu-mented data which are never analyzed, provide little orany feedback to resource managers, and contributelittle or nothing to our understanding of the systemsbeing monitored’’ (MacDonald 1994). MacDonald statesthat formulating the specific objectives of the monitor-ing project is the most important and most difficult stepin the entire monitoring process. In this paper wesuggest that the information need is the crucial step inthe information-producing process and the determin-ing factor to specify monitoring objectives. In theNetherlands, in redesigning the national monitoringnetwork, monitoring objectives had been defined, with-out specifying information needs first (Adriaanse 1993).Although attempts have been made since then toimprove this situation, it is still felt that the monitoringnetwork does not fully meet the information needed.Defining what information should be produced, what isreally the question or the problem, is not easy to do,however. How a problem is defined determines theunderstanding and approach to the problem and thusthe possible solutions to the problem (Bardwell 1991).

The information objectives should be identified bymanagers in consultation with technical staff (MacDon-ald 1994), but management issues tend to be vague andloosely specified. Policy-makers, politicians, the public,and other information users tend to pose questions ofthe form: Is this country safe against flooding? or Whatwill be the consequences of dry years for agriculture?(Hofstra 1995). These questions are not easily or simplyanswered, and they even raise new questions. The firstquestion above, for example, generates the questions:What is ‘‘safe’’? Maybe a statistical chance of one in10,000 years is safe? Does this count for the whole of theNetherlands or just for one area? And against what cost?

On the other hand, experts (researchers, scientists,and other information producers that generate informa-tion) tend to provide answers like: The maximum water

level is 34.6 m above mean sea level; or pH is 7.8. Manyinformation users may find it difficult to relate thoseanswers to their questions (Timmerman and Ottens1997). Lack of interpretation against a clearly under-stood information need prevents the data from becom-ing tailor-made information.

These questions and answers reflect different worldsof thinking. Information producers often do not speakthe same language as information users (Helsel 1997).Information users tend to oversimplify and have unreal-istic expectations, whereas information producers failto address the management needs (MacDonald 1994).These two worlds need to be linked together throughcommunication. A dialogue is needed between informa-tion users and information producers to develop theconnecting questions—questions that are clearly articu-lated and understood by both information producersand users.

In addition to formulating these questions, it isimportant to determine how the answer should bepresented and the level of detail and precision to beincluded in the answer. Given the question on safety, itmay be agreed that safety will be expressed as astatistical chance. In this way, safety is defined as afeature that may be calculated from existing data.Another option for expressing safety is to present anumber of scenarios with consequences that are con-nected economically (e.g., costs of measures and im-pacts), sociologically (e.g., half of the village will disap-pear in building a dike of several meters high), andecologically (e.g., valuable nature reserves will disap-pear).

The way a question is formulated and stated dependson a number of factors, such as the legal/managementcontext, environmental conditions, and, also very impor-tant, the latest concepts in water management. A goodexample of changing concepts may be found in thedevelopment of water management policy in the Neth-erlands. In the first national policy document (pub-lished in 1968), water quantity was the central issue. Thesecond national policy document (published in 1984)added a relation with economy and brought aboutattention to water-quality aspects. The third nationalpolicy document (published in 1989) portrayed wateras a system and ecology was valued next to the economy(Ministry of Transport, Public Works and Water Manage-ment 1989, Wisserhof 1994, Blom 1995). The trend inwater management today is toward integrated watermanagement—a balance between economics, ecologyand sociology (e.g., Raskin 1997, Van Rooy and De Jong1995). These aspects are integrated in the main aim ofthe fourth national policy document: ‘‘To have andmaintain a safe and habitable country and to develop

J. G. Timmerman and others230

and maintain healthy and resilient water systems whichwill continue to guarantee sustained use’’ (Ministry ofTransport, Public Works and Water Management 1997).

Thus, today economics, ecology, and sociology to-gether are key to forming and stating the questions tobe answered, but only the right question will lead to theright information to make decisions (Adriaanse 1997).

How to Get the Answer

If the question is clear and well understood, the nextstep is to look for ways to find the answer. This shapingof the answer must also be part of the dialogue betweeninformation users and information producers. Theremay be different strategies to collect information. Forexample, in our flooding example, water levels may bemonitored and the height of dikes and dams can bemeasured precisely. If the level of public support mustalso be taken into account in a decision relating toheight of dikes and dams, one method to measure socialsupport could be a questionnaire. The measurementsystems combine physical and social variables to pro-duce information.

With technical knowledge, the technical possibilitiescan be derived, and using equations and models,chances for calamities may be calculated. But then thequestion becomes: How exact should the answer be? Anexact answer will require extensive inventories of theexisting situation, collection of large amounts of datafrom different dimensions of the question, and detaileddata analysis and interpretation in order to provide asexact an answer as possible. Thus, a balance must bestruck between the level of detail of the answer and thenecessary effort. For this balancing, the term ‘‘relevantmargin’’ is important. The relevant margin may bedefined as the information margin that is of concern tothe information user. If, for example, the critical level ofa pollutant lies at 10 µg/liter, it is no use to provideinformation in nanograms per liter. The relevant mar-gin may be about 5 µg/liter in such a case (Adriaanse1995).

In determining how the answer to the question willbe presented, it is important to define the preparationand presentation of data with both information produc-ers and users. How data are analyzed and presented mayinfluence the way data are collected. Calculation oftrends, for instance, is a different monitoring objectiveand requires a different monitoring strategy than com-pliance with standards. Thus, again, the nature of thequestion being asked will result in different approachesto the design and implementation of monitoring pro-grams. Factors that are influenced by the question arethe selection of variables to be measured, the frequency

and location of measurements, the additional informa-tion needed for interpretation, and the way in whichinformation is generated and presented (Bartram andBallance 1996). In fact, these choices define the moni-toring objectives and consequently are the prerequisitesfor the collection of data.

A Common Framework

As stated before, a dialogue is needed betweeninformation users and information producers. To facili-tate this dialogue, a common basis, or framework, isneeded to ensure that all parties involved share thesame notion of the process at hand. Such a frameworkshould appeal to both information users and informa-tion producers. Since most information users are notinterested in technical details, the framework shouldpreferably be easy to understand yet describe all essen-tial steps. The concept of developing a common basisfor approaching the design of water monitoring systemsis well established in the water monitoring literature.Several of these approaches will be examined from theperspective of incorporating the desires of informationusers into the approach being proposed.

Ward and others (1990) define the monitoringsystem as containing a number of tasks to convert asample of water in the environment into accurateunderstanding of it (Figure 1). Ward (1995) stresses theneed to customize monitoring efforts. In other words,we must define what information we want and then

Figure 1. Flow of information through a monitoring system(Ward and others 1990).

Defining Information Goals for Monitoring 231

design the monitoring system to produce it. Such aprocess should be clear to both information users andinformation producers. However, in the steps in Figure1, there is no clearly defined role for information users.

MacDonald (1994) notes that little attention is givento the explicit and iterative process necessary for theefficient design and execution of monitoring projects.He suggests 12 key steps in this design and execution(Figure 2). These steps are rather descriptive and willnot easily be reproduced by information users. Neverthe-less, MacDonald (1994) emphasizes the involvement ofmanagement as information users in defining monitor-ing objectives and describes the significance of feed-back of monitoring results to management (Figure 3).In this feedback loop the individual steps become clearand also the role of management is clear.

In Chapman (1996) key elements of an assessmentprogram are described (Figure 4). The possible types ofwater-quality assessment programs are numerous andshould be designed or adopted according to objectives.These objectives should be set on the basis of environ-mental conditions, water uses (actual and future), waterlegislation, etc. (Chapman 1996). However, in this

framework the objectives are not explicitly related tothe needs of information users. The emphasis is still onthe information production side, with little concern formanagerial issues.

In this paper, the information cycle (Figure 5) ispresented as a general framework that describes theessential steps in the process of information collection.In the information cycle, information collection isinextricably bound up with water management. Theinformation cycle incorporates water management inthe frameworks described (Ward and others 1990, Ward1995, MacDonald 1994, Chapman 1996) and addsspecification of information needs and developing ofan information strategy. A monitoring plan can be theoutcome of the information strategy, but this is notinevitable. Thus, the information cycle may be used as ageneric framework for designing information collectingsystems.

In going from information required to informationobtained, the following steps can be distinguished: (1)Information users, as part of the information cycle,should, in cooperation with information producers,

Figure 2. Key steps in the design and execution of a monitoring project (MacDonald 1994).

Figure 3. The monitoring feedback loop (MacDonald 1994).Figure 4. Key elements of an assessment program (Chapman1996).

J. G. Timmerman and others232

decide upon the characteristics of the information thatis needed. (2) Information producers will, in coopera-tion with information users, decide upon the best way(i.e., strategy) to collect information. (3) The actualcollection of data is the next step in the informationcycle. (4) The collected data are analyzed and theresults interpreted relative to the information sought;information statements are made. (4) The resultinginformation is presented and transferred to the informa-tion users in a proactive manner.

Unfortunately, too often the information presentedas a result of the above actions does not satisfy theinformation users. Why is this the case? Reasons in-clude: (1) Specification of information need is insuffi-cient. (2) The information need as specified is not the‘‘real’’ information need (too little effort went into theprocess of defining the information need). (3) Thestrategy to collect information has not produced theright information. (4) The obtained information gener-ates new questions, making the originally agreed uponinformation appear to be inadequate. (5) The situationhas changed (e.g., new policies), causing other informa-tion to be needed. (6) New methods have been devel-oped.

Generally, the conclusion is that new information isneeded. The new (i.e., changed) information needmust be determined and the process starts again, butfrom a new position. Because we have learned from theprevious process, this next cycle should lead to bettertailored information. By continuous evaluation andfeedback, the cycle becomes a spiral, leading to aconstantly improving quality of the information. Thecycle thus represents a continuously adjusting informa-tion system that can be presented as a spiral towards abetter information product (Figure 6).

Each step in the information cycle puts requirementson the previous step of the cycle. If, for example, a trendwith a defined reliability has to be calculated, samplinghas to take place with a certain frequency, depending

on the variability of the data. Thus, data analysis putsrequirements on data collection. Furthermore, eachstep limits the following steps. For example, if thesampling frequency is too low, no reliable trend can becalculated. By theoretically going through the informa-tion cycle both clockwise and counterclockwise, formu-lating the prerequisites and restrictions of every step,these requirements and limitations may be made ex-plicit.

Application of the Information Cycle

The information cycle is a framework stating theessential steps in the continuous process of informationproduction. Each of these steps has to be elaborated,depending on the required information and the strat-egy to collect this information. Monitoring, for in-stance, may require methods of data analysis differentfrom modeling. Thus, for different purposes, the infor-mation cycle may be adjusted. In the Guidelines onWater-Quality Monitoring and Assessment of Trans-boundary Rivers (ECE TFMA 1996, ECE 1996), thenotion of an information network being a cycle ofactivities was added to the steps that Ward and others(1990) described. This resulted in the monitoring cycleas shown in Figure 7.

The monitoring cycle shows the different steps thathave to be defined to be able to specify all requirementsfor a (chemical) monitoring system. Information forone step is used to develop and define information forthe next step. The resulting information as produced bythe monitoring system is the sum of a chain of activities.When the quality of the information in one of the stepsis poor, it will, in turn, affect the quality of the resultinginformation. The strength of the cycle is thus deter-mined by its weakest link (Stroomberg and others1995).

Figure 5. The information cycle.Figure 6. A spiral towards better tailored information (Co-fino 1995).

Defining Information Goals for Monitoring 233

Monitoring is defined as the long-term, standardizedmeasurement and observation of the aquatic environ-ment in order to define status and trends. A survey isdefined as an intensive program of finite duration tomeasure and observe the quality of the aquatic environ-ment for a specific purpose. (Chapman 1996).

The monitoring cycle can be adapted for other typesof monitoring. For biological monitoring, for instance,the steps ‘‘sample collection’’ and ‘‘laboratory analysis’’may be supplemented by the step ‘‘field observations.’’In the draft Guidelines on Monitoring and Assessmentof Transboundary Groundwater (ECE TFMA 1998),‘‘quantity measurements’’ is added, replacing ‘‘samplecollection’’ and ‘‘laboratory analysis.’’

Issues to Consider in SpecifyingInformation Needs

In specifying the information needs for integratedwater management, it is suggested (ECE TFMA 1996) toapproach the task from a triangle of core elements:functions and use of the water, problems and threats forthis use or function, and measures that may be taken(Figure 8). For example, a large part of the Dutchshoreline consists of dunes that protect the land behindit against flooding (function). During storms, parts ofthese dunes are washed away (problem). The damage isrepaired by supplying sand (measure). Another prob-

lem for this function is changing of erosion andsedimentation patterns due to, e.g., land acclamation inthe sea near Rotterdam. As a measure, small dams areconstructed along the shoreline to prevent washingaway of sand. By elaborating problems and measures ofall function, a comprehensive analysis of the situation athand is made.

The next element in specifying the informationneeds may be a division into the necessary informationcategories (CIW/CUWVO, 1996):

1. Information for policy evaluation. This informa-tion should be linked to the policy as defined. Thestarting point is the existing or desired function. Theresulting information should show to what extent prob-lems are still relevant and measures have had theintended effect. Indicative variables may be useful inthis respect to keep track of problems. If there isevidence of deterioration in water quality, additionalinformation is needed to be able to (re)define mea-sures.

2. Information for policy preparation. Policy prepara-tion needs information on present status and futuredevelopments. Major sources of information includeexperiences from other parts of the world and extrapo-lation of progress, e.g., in pollution reduction. Regularinventorying of the status of the water also is needed.This information should be related to its direct environ-ment, to previous inventories, and to information fromother regions or countries.

3. Information for operational water management.Operational water management needs direct usableinformation. For intake of drinking water, for example,the exact concentration of substances may be lessimportant than whether they are above or below certainlevels. For an overview of water quality, biologicalalarms may be sufficient. Flood forecasting at firstinstance requires rough measurement of rising orfalling of the water level but when critical levels arereached, each centimeter becomes important. Essentialinformation for operational water management is real-time availability to allow for instant reactions.

A next important factor in specification of informa-tion needs is quantification. Information needs shouldbest be stated in terms such as: ‘‘Twenty percentreduction of pollution in the next five years,’’ or ‘‘Nomore interruptions in the intake of drinking waterwithin two years.’’ An element of relativity (percentagereduction of . . .) or quantity (no more . . . or lessthan . . .) can also be useful in the specification. Next tothat, an element of time (. . . within two years) shouldbe included. ‘‘By the year 2000 the ecosystem of theriver Rhine should be improved to such an extent thathigher species, such as the salmon, may become indig-

Figure 7. The monitoring cycle (ECE TFMA 1996).

Figure 8. Core elements in water management (ECE TFMA1996).

J. G. Timmerman and others234

enous’’ (Hogervorst 1993) is a statement that can bemeasured. There is an element of quantity (a more orless stable population of salmon) and an element oftime (the year 2000). Naturally, the quantity elementmust have a link to water management measures. In thiscase, the element of quantity should be expressed interms of a specific water-quality variable(s) and tophysical factors such as suitable spawning grounds andthe possibility of migration. When these criteria arequantified, the information need can be converted intoan information strategy (Timmerman and others 1996).

When specifying information needs, it is also impor-tant to look at information utilization, in other words,what will be done with the information once it iscollected. This is a determining factor for the informa-tion need (De Jong and Timmerman 1997).

Indicators and Information Needs

Indicators can be helpful to further define theinformation needs and may facilitate going through thesteps of the information cycle. Indicators are useful incommunicating between the different worlds of informa-tion users and information producers, because they canpresent information in a condensed/aggregated formatand are often linked to specific problems or issues,which are in turn based on specific management needs.Simplification and quantification of information on thewater system under consideration makes this informa-tion accessible and comparable to other water systems.

Indicators are defined here as an observable ormeasurable quantity, variable, or parameter, represent-ing a process in the environment and having signifi-cance beyond its face value (Bakkes 1994, OECD 1993).The construction of an indicator is a means of achievingreduction in data volume while retaining significancefor particular questions. Suitable indicators provide forthe three core elements (Figure 8). As an example, ininternational river basin programs such as Salmon 2000and the return of the beaver in the Elbe, the species arethe indicator of a greater goal. They are measurablevariables, representing processes in the environment byputting requirements on physical and chemical condi-tions of the water, thus also having a significance for thetotal ecological functioning of a river. These species alsoembody the core-elements function (ecological func-tion), problems (water quality and land use), andmeasures (reduction of emissions, ecological restora-tion). Moreover, they are very suitable in communicat-ing to policy-makers or the public. Because of this,return of popular species is often stated as an objectiveof ecological restoration measures, facilitating in-creased social acceptance of these measures (Timmer-

man and Ottens 1997). When indicators are furtheraggregated, the term ‘‘index’’ is used. This can be a setof aggregated or weighted parameters or indicatorsdescribing a situation (OECD 1993).

Indicators may be developed on the basis of differentperspectives. Van Harten and others (1995) distinguishindicators on the pressure on, state of, and impact on awater system and the response of the socioeconomicsystem (Figure 9). This classification is based on anecological stress–response chain (Friend and Rapport1979). A pressure indicator describes the intensity ofhuman activities that cause changes in quality and/orquantity of the water system. The state indicator de-scribes the status of the quality and/or quantity of thewater system. A pressure may result in a new state. Theimpact indicator describes the influence of the state ofthe water system on functions and uses of the watersystem. When functions or uses are affected, a societalresponse may be expected. This is described by theresponse indicator. The response aims at a new balanceof the water system and the protection of functions anduses.

The PSIR (Pressure state impact response) chain canbe related to the core elements in water management(Figure 8). Impact indicators reflect the way functionand use of a water system is influenced, while statusindicators illustrate the level of functioning. Problemsand threats may be quantified by pressure indicatorswhere response indicators describe the measures taken.By providing surveillance for the different functionsand uses of water systems, with often conflicting de-mands, indicators can help in the management of thiscomplex situation.

When problems are recognized, information is firstneeded on the source of the problem and the level ofeffect on the use of the water system. This informationmay be provided by pressure and impact indicators.When the problem is fully recognized and measures are

Figure 9. PSIR concept (after Van Harten and others 1995).

Defining Information Goals for Monitoring 235

formulated, the focus will be on the measures taken andthe present situation in the water system. Status andresponse indicators provide for this information. Gener-ally starting from issues, for policy evaluation or prepa-ration, the information need will concentrate mostly onpressure and response indicators. For operational watermanagement, focusing primarily on functions of waters,status and impact indicators will provide the greaterpart of the information.

As an example, suppose the issue is the influence ofliquid manure on a groundwater system in relation tothe use of groundwater as a source of drinking water. Anindicator for the pressure on the groundwater systemcould be the nitrogen load of the soil. The NO3

concentration in shallow groundwater can be selectedas an indicator for the state of the system. The impactindicator may be the suitability for use of the groundwa-ter as a source of drinking water. A response indicatorcould be the percentage of the total amount of liquidmanure produced that is processed in a way that is notharmful to the environment. If the issue is fully recog-nized, policy makers will focus mostly on status andresponse indicators.

Defining an Information Strategy

After defining the information need, it is importantto recognize that the information cycle involves defin-ing a strategy for collecting this information. The informa-tion strategy chosen is a mixture of the informationneed and the best practical way to collect information.

Integrated water management requires integratedinformation. To provide such information, many differ-ent sources are needed. In water management, usuallymonitoring is used to collect data. Often models areused to make predictions on the basis of the monitoringdata. Surveys are used to provide insight into specificissues. Other sources of information may be data from

other disciplines such as agriculture, recreation, sociol-ogy, ecology, and economics. The information cyclesuggests that only after specification of informationneeds, can it be determined what part of the informa-tion needed will be collected from what source. Incorpo-rated into the information cycle, different sources ofinformation as well as different ways of collecting andanalyzing data are shown in Figure 10.

When information is collected from monitoring,some general strategies in water quality monitoring are(e.g., ECE TFMA 1996, Van Luin and Ottens 1997,Witmer 1995): (1) Variable oriented: collecting data ofindividual variables. This strategy is useful when thenumber of potentially harmful substances is limited orfor examination of specific processes. (2) Effect ori-ented: data collection is focused on indicative variablesthat point out if something is going wrong. If there is anindication of some kind, specialized investigation maypoint out exactly what is going on (CIW/CUWVO1996). Ecotoxicological testing for instance, may revealeffects that highly exceed the toxicity accounted for bychemical analysis (Kristensen and Krogsgaard Jensen1997). This strategy is useful when there are no dominat-ing problems that may be related to a limited set ofvariables. (3) Source-oriented: collecting data of sourcesthat cause adverse effects. This strategy may be used,e.g., for licensing (effluent monitoring) or policy prepa-ration but is also indicative for pressure on a watersystem. (4) Achievement-oriented: collecting data on apolicy goal that is set to be achieved within a given timeperiod. The strategy aims at pointing out the distance tothe goal or the effectiveness of measures. (5) Predictive:data collection on events that are expected to happen.This strategy is used in flood forecasting and for policyanalysis. The use of models is imperative in this strategy.(6) Knowledge-building: data collection to understandspecific processes in the water. Usually this strategy isperformed on a short-term basis. (7) Tiered testing: aStepwise approach in which additional data are collectedonly when available data provide insufficient information(Figure 11). (8) Investigative: collection of data of whichthe relevance is not yet clear. This may be done for instanceby storing samples that have not been analyzed. If someevent (e.g., a fish-kill) occurs, these samples may be ana-lyzed to trace back the source of the event.

These different strategies are mostly interrelated.They are also closely related to the information needand choosing a strategy may lead to rethinking of theinformation need.

Other Steps in the Cycle

The information need and the strategy chosen deter-mine the way data collection, data analysis, and informa-

Figure 10. Sources of information (after ECE TFMA 1996,ECE 1996).

J. G. Timmerman and others236

tion utilization are conducted. The resulting informa-tion should fit the information need, but as statedbefore, there are reasons why there could be a mis-match. Therefore it is important that in designing thesesteps, choices made are traceable. In case there isdeviation of the information obtained from the informa-tion need as specified, the process by which the informa-tion was produced is known (ECE TFMA 1996, Timmer-man and others 1996).

Some aspects of the steps data collection, dataanalysis, and information utilization are usually recog-nized but in practice not always implemented in moni-toring operations. In this section, these subjects will behighlighted; more detailed information can be found inWard and others (1990), Bartram and Ballance (1996),and Chapman (1996).

In designing a monitoring network, choice of sam-pling frequencies or locations is important. When, forinstance, water-quality data are available, use of statisticscan minimize the number of locations through correla-tion between stations (Sanders and Loftis 1995). Knowl-edge of chemical, biological, and physical processes, ofspecific local characteristics or properties of the objectof investigation, and of the analytical methodology andstatistics also should be used in devising a program ofmeasurements (Cofino 1993).

The collected data have to be stored in a way thatthey can be easily retrieved, but the data also have to bedocumented in a comprehensible manner. Data shouldbe validated or approved before they are made acces-sible to any user or entered in any data archive (ECETFMA 1996).

One of the most important challenges is the aggrega-tion of the huge amount of available data into clearinformation. To illustrate the volume of data, for eachlocation on the Rhine River alone, the Dutch collectsome 5000 data points on water quality each year.Various tools have been developed for analyzing water-quality data. The Netherlands have developed theSTOWA method for ecological judgment (Klapwijk andothers 1995), the water dialogue (Latour and others1996), and the AMOEBE (Ten Brink and others 1991).In New Zealand, a method has been developed for analyz-ing water-quality trends (McBride and Smith 1997).

Discussion

Because of a growing demand for information onprocesses in the water system and the need for inte-grated assessment of this information, informationproduction for water management has become moreand more complex. Since water-quality monitoring

usually has a long-term character, decisions for develop-ing a monitoring network have a long-term conse-quence. This increasing complexity, together with thepossible long-term consequences, necessitates reassess-ment of the mechanisms of supplying information. Theinformation cycle, as presented in this paper, provides aconceptual framework that incorporates the specifica-tion of information needs into the process of informa-tion gathering and provides possibilities to manage thequality of the process. By elaborating the steps in theinformation cycle, a better information product may beobtained. Each step puts requirements on the previousone and limits the following step. These requirementsand limitations can be clarified by theoretically goingthrough the cycle both clockwise and counterclockwise.The cyclic character comprises regular evaluation of thegathered information, thus quantitatively connectingthe monitoring system design and operations with theinformation expectations and/or products required bymanagement. In this way, up-to-date tailor-made infor-mation may be produced.

The information cycle offers a framework as a basisfor the dialogue between information users on onehand and information producers on the other hand. Itprovides a method to direct the dialogue to thosequestions that must be answered if we are to formulatean information strategy and that is accountable for theinformation produced. The information cycle and itsmonitoring version were been introduced in the Dutchwater management bodies in 1997 and have become apart of the common language since.

Figure 11. Tiered testing strategy (ECE TFMA 1996).

Defining Information Goals for Monitoring 237

The information cycle may be adapted for its use, asis shown in the monitoring cycle. Through adaptation,the information cycle can also be used for specific typesof monitoring as, for instance, groundwater quantitymonitoring or for short term information production,as in surveys.

One of the major challenges is to define the elementof information need. As discussed in this paper, ques-tions may seem simple but on second thought appear togenerate new questions—questions that must be an-swered in the process of designing a water-qualitymonitoring system. The actual information need has tobe defined in a dialogue between information users andinformation producers. A method for specification ofinformation needs is currently under development andis being tested in several pilot projects in the Nether-lands and in some pilots under the ECE Task Force onMonitoring and Assessment in European transbound-ary rivers. First results from these pilot projects indicatethat a better link of information to policy in a moreintegrated manner becomes possible.

Acknowledgments

Many thanks to Dr. Jim C. Loftis, Dr. James A. Perry,and an anonymous reviewer who critically reviewed themanuscript. Their useful suggestions greatly improvedthis manuscript.

Literature Cited

Adriaanse, M. 1993. Optimalisatie routinematig onderzoekkwaliteit rijksbinnenwateren. Hoofdrapport (Optimizationof routine research into water quality of national inlandwaters). RIZA notanr. 92.055 (in Dutch).

Adriaanse, M. 1995. Information requirements as designcriteria for surface water monitoring. Pages 126–133 in M.Adriaanse, J. Van de Kraats, P. G. Stoks, and R. C. Ward(eds.), Proceedings of the international workshop: Monitor-ing tailor-made. Beekbergen, The Netherlands.

Adriaanse, M. 1997. Tailor-made guidelines: A contradictionin terms? Pages 391–399 in J. J. Ottens, F. A. M. Claessen,P. G. Stoks, J. G. Timmerman, and R. C. Ward (eds.),Proceedings of the international workshop: Monitoringtailor-made—II. Nunspeet, The Netherlands.

Adriaanse, M., and G. H. Broseliske. 1998. Information strate-gies—the challenge of common sense. Pages 299–309 in M.Landsberg-Uczciwek, M. Adriaanse, R. Enderlein (eds.),Proceedings of the international conference: Managementof transboundary waters in Europe, Mrzezyno, Poland.

Adriaanse, M., J. Van de Kraats, P. G. Stoks, and R. C. Ward(eds.). 1995. Conclusions. Pages 345–346 in Proceedings ofthe international workshop: Monitoring tailor-made. Beek-bergen, The Netherlands.

Bakkes, J. A., G. J. Van der Born, J. C. Helder, and R. J. Swart.1994. An overview of environmental indicators: state of the

art and perspectives UNEP/EATR. 94-01; RIVM/402001001.Bilthoven, The Netherlands.

Bardwell, L. V. 1991. Problem-framing: A perspective onenvironmental problem-solving. Environmental Management15(5):603–612.

Bartram, J., and R. Ballance (eds.). 1996. Water qualitymonitoring. A practical guide to the design and implemen-tation of freshwater quality studies and monitoring pro-grammes. UNEP/WHO, London.

Blom, G. 1995. Dutch water management in practice, theREGIWA projects: Experiences, results, future develop-ments. Pages 3–12 in E. Romijn, and J. C. Schutte de Roon(eds.), Netherlands experiences with integrated water man-agement: considerations for international cooperation.RIZA, Lelystad, The Netherlands.

Chapman, D. (ed.). 1996. Water quality assessment—a guideto the use of biota, sediments and water in environmentalmonitoring. Chapman, and Hall, London.

CIW/CUWVO. 1996. Progress report on integrated watermanagement and North Sea matters 1996. CommissionIntegrated Water management/CUWVO (in Dutch).

Cofino, W. P. 1993. Quality assurance in environmental analy-sis. Pages 359–381 in D. Barcelo (ed.), Environmentalanalysis: Techniques, applications and quality assurance.Elsevier Science Publishers, Amsterdam.

Cofino, W. P. 1995. Quality management of monitoringprogrammes. Pages 178–187 in M. Adriaanse, J. van deKraats, P. G. Stoks, and R. C. Ward (eds.), Proceedings of theinternational workshop: Monitoring tailor-made. Beekber-gen, The Netherlands.

De Jong, J., and J. G. Timmerman. 1997. Opening andintroduction. Pages 1–4 in J. J. Ottens, F. A. M. Claessen,P. G. Stoks, J. G. Timmerman, and R. C. Ward (eds.),Proceedings of the international workshop: Monitoringtailor-made—II. Nunspeet, The Netherlands.

De Jong, J., P. T. J. C. Van Rooy, and S. H. Hosper. 1994. Livingwith water: At the cross-roads of change. Pages 477–493 inProceedings of the international conference Living withwater. Amsterdam, The Netherlands.

ECE. 1996. Protection of transboundary waters, guidance forpolicy- and decision-making. Water Series No. 3, UN, Geneva.

ECE TFMA (Task Force on Monitoring and Assessment).1996. Guidelines on water-quality monitoring and assess-ment of transboundary rivers. RIZA report no. 96.034.

ECE TFMA (Task Force on Monitoring and Assessment).1998. Guidelines on monitoring and assessment of trans-boundary groundwaters (in preparation).

Friend, A., and D. Rapport. 1979. Towards a comprehensiveframework for environment statistics: A stress–responseapproach. Statistics Canada, Catalog 11-510, Ottawa.

Helsel, D. R. 1997. The value of water quality information fornational environmental policy. Pages 47–52 in J. J. Ottens,F. A. M. Claessen, P. G. Stoks, J. G. Timmerman, and R. C.Ward (eds.), Proceedings of the international workshop:Monitoring tailor-made—II. Nunspeet, The Netherlands.

Hofstra, M. A. 1995. Information is vital for the nationaldecision maker. Pages 43–54 in M. Adriaanse, J. Van deKraats, P. G. Stoks, and R. C. Ward (eds.), Proceedings of the

J. G. Timmerman and others238

international workshop: monitoring tailor-made. Beekber-gen, The Netherlands.

Hogervorst, D. 1993. The Rhine Action Programme. Pages35–42 in W. J. Walley, and S. Judd (eds.), Proceedings of thefreshwater Europe symposium on river water quality moni-toring and control. Birmingham, UK, 22–23 February 1993.

Klapwijk, S. P., J. J. P. Gardeniers, E. T. H. M. Peeters, and C.Roos, 1995. Ecological assessment of water systems. Pages105–117 in M. Adriaanse, J. Van de Kraats, P. G. Stoks, andR. C. Ward (eds.), Proceedings of the international work-shop: Monitoring tailor-made. Beekbergen, The Netherlands.

Kristensen, P., and J. Krogsgaard Jensen. 1997. Integratedapproach for chemical, biological and ecotoxicologicalmonitoring—a tool for environmental management. Pages111–120 in J. J. Ottens, F. A. M. Claessen, P. G. Stoks, J. G.Timmerman, and R. C. Ward (eds.), Proceedings of theinternational workshop: Monitoring tailor-made—II. Nun-speet, The Netherlands.

Latour, P. J. M., E. Stutterheim, and A. J. Schafer. 1996. Fromdata to information: The Waterdialogue. H2O 29(23):693–696 (in Dutch with English abstract).

MacDonald, L. H. 1994. Developing a monitoring project.Journal of Soil and Water Conservation May–June: 221–227.

McBride, G. B., and D. G. Smith. 1997. Results of a trendassessment of New Zealand’s national river water qualitynetwork. Pages 135–142 in J. J. Ottens, F. A. M. Claessen,P. G. Stoks, J. G. Timmerman, and R. C. Ward (eds.),Proceedings of the international workshop: Monitoringtailor-made—II. Nunspeet, The Netherlands.

Ministry of Transport, Public Works and Water Management.1989. Water in the Netherlands: A time for action. Summaryof the national policy document on water management.Ministry of Transport, Public Works and Water Manage-ment, The Hague.

Ministry of Transport, Public Works and Water Management.1997. Fourth national policy document on water manage-ment, government proposals. Sdu uitgevers, The Hague,The Netherlands, 43 pp.

OECD. 1993. Group on the state of the environment workshopon indicators for use in environmental performance re-views. Draft synthesis report, Paris.

Ongley, E. D. 1998. Modernisation of water quality pro-grammes in developing countries: Issues of relevancy andcost efficiency. Water Quality International September/October:37–42.

Raskin, P., P. Gleick, P. Kirshen, G. Pontius, and K. Strzepek.1997. Water futures: Assessment of long-range patterns andproblems. Stockholm Environment Institute, Stockholm,Sweden, 77 pp.

Sanders, T. G., and J. C. Loftis. 1995. Factors to consider inoptimization of a monitoring network. Pages 146–152 in M.Adriaanse, Van de Kraats, P. G. Stoks and R. C. Ward (eds.),Proceedings of the international workshop: Monitoringtailor-made. Beekbergen, The Netherlands.

Somlyody, L. 1994. Water quality management: can we im-prove integration to face future problems? Pages 343–359 inProceedings of the international conference: Living withwater. Amsterdam, The Netherlands.

Stroomberg, G. J., I. L. Freriks, F. Smedes, and W. P. Cofino.

1995. Quality assurance and quality control of surface watersampling. In P. Quevauvilier (ed.), Quality assurance in environ-mental monitoring. VCH, Weinheim, Germany (in press).

Ten Brink, B. J. E., and J. H. Woudstra. 1991. Towards aneffective and rational water management: The AquaticOutlook Project—integrating water management, monitor-ing and research. European Water Pollution Control 1(5):20–27.

Ten Brink, B. J. E., S. H. Hosper, and F. Colijn. 1991. Aquantitative method for description and assessment ofecosystems: The AMOEBE approach. Marine Pollution Bulle-tin 23:265–270.

Timmerman, J. G., and J. J. Ottens. 1997. The informationcycle—a framework for the management of our waterresources. Pages 407–415 in With rivers to the sea, proceed-ings of the 7th Stockholm water symposium, 10–15 August,Stockholm, Sweden.

Timmerman, J. G., M. J. Gardner, and J. E. Ravenscroft. 1996.Quality assurance. Volume 4, 1994/1995 working pro-gramme of the ECE task force on monitoring and assess-ment. Institute for Inland Water Management and WasteWater Treatment (RIZA), Lelystad, The Netherlands. RIZAreport no. 95.067.

Timmerman, J. G., M. Adriaanse, R. M. A. Breukel, M. C. M.van Oirschot, and J. J. Ottens. 1997. Guidelines for waterquality monitoring and assessment of transboundary rivers,European Water Pollution Control 7(5):21–30.

Van Harten, H. A. J., G. M. Van Dijk, and H. A. M. De Kruijf.1995. Waterkwaliteitsindicatoren overzicht, methode-ontwik-keling en toepassing (Water quality indices, overview, meth-odologies and application). RIVM report no. 733004001 (inDutch).

Van Luin, A. B., and J. J. Ottens. 1997. Conclusions andrecommendations. Pages 401–403 in J. J. Ottens, F. A. M.Claessen, P. G. Stoks, J. G. Timmerman, and R. C. Ward(eds.), Proceedings of the international workshop: Monitor-ing tailor-made—II. Nunspeet, The Netherlands.

Van Rooy, P. T. J. C., and J. De Jong. 1995. Towards comprehen-sive water management in The Netherlands: (1) develop-ments. European Water Pollution Control 5(4):59–66.

Ward, R. C. 1995. Monitoring tailor-made: What do you wantto know? Pages 16–24 in M. Adriaanse, J. van de Kraats, P. G.Stoks, and R. C. Ward (eds.), Proceedings of the interna-tional workshop: Monitoring tailor-made. Beekbergen, TheNetherlands.

Ward, R. C., J. C. Loftis, and G. B. McBride. 1986. The‘‘data-rich but information-poor’’ syndrome in water qualitymonitoring. Environmental Management 10(3):291–297.

Ward, R. C., J. C. Loftis, and G. B. McBride. 1990. Design ofwater quality monitoring systems. Van Nostrand Reinhold,New York, 231 pp.

Wisserhof, J. 1994. Enhancing research utilization for inte-grated water management. Proceedings of the internationalconference: Living with water, Amsterdam, The Nether-lands.

Witmer, M. C. H. 1995. Information needs for policy evalua-tion. Pages 55–61 in M. Adriaanse, J. van de Kraats, P. G.Stoks, and R. C. Ward (eds.), Proceedings of the interna-tional workshop: Monitoring tailor-made. Beekbergen, TheNetherlands.

Defining Information Goals for Monitoring 239