Environmental and Economic Impacts of an Irrigated Agriculture System
Bill Bland, Brad Barham, George Kraft - University of Wisconsin
Introduction
North of Madison, Wisconsin, the tallgrass prairie ecosystem that provided the edaphic base of the cornbelt gives way to the Central Plain of Wisconsin . The Central Plain is 7,000 km2 of predominantly sandy outwash soils formed as the last glaciers retreated 10,000 years ago. The soils of this region have little inherent productivity, and early in the century barely supported agriculture. During the 1940s, on an abandoned farm on the southern edge of the region, Aldo Leopold thought and wrote about, and acted on, an ethic of stewardship for the land. Leopold's book, "A Sand County Almanac" is now a classic work on the philosophy of stewardship for the environment .
At about the time Leopold was formulating his prescription for healing of his "sand farm," the technology of aluminum fabrication, motivated by aviation, was making feasible machines to irrigate the sands. Three decades after the appearance of "A Sand County Almanac" the University of Wisconsin published a brochure, "Irrigation of the Central Sands of Wisconsin," which described a bright future for irrigated agriculture in the region . The agricultural system described is now nearly 50,000 ha of vegetables grown under sprinkler irrigation.
Thus two distinct views of the human relationship to the land took root in the Central Plain at about the same time. The question of their mutual compatibility was to come into focus in the 1980s and 90s with the emergence of environmental quality concerns linked to irrigated crop production. Prominent among these concerns is the effect of this agricultural system on regional groundwater quality. Here groundwater quality is more than a theoretical or future concern: municipal water suppliers and private well owners must employ expensive treatment techniques to meet drinking water purity criteria.
Yet the regional economic benefits of irrigated crop production are significant. High value vegetable crops create income from the region's water, relatively short growing season, and sandy soil. Processing of these crops within the region adds value and income. Thus, economic, public health, and environmental quality issues are in conflict.
The present and future states of this conflict are not defined simply by economics and unambiguous biological outcomes. Rather, the debate will involve philosophical perspectives (e.g., an entrepreneurial farmer and a commonweal-oriented municipal official), attitudes toward risk (e.g., a new parent and a water utility manager), and concepts of environmental justice (e.g., local field workers and absentee landowners). There appear to be irreconcilable, fundamental conflicts of interest, but our society's democratic principles demand that all voices be heard and private property rights respected. In addition to our legal tradition of due process is an evolving appreciation that public participation and education in environmental decision making is a powerful and perhaps necessary step to obtain the broadest possible acceptance of outcomes. Our long-term goal is to discover how disparate disciplinary knowledge can be integrated so as to be accessible enough to inform public debate and decision-making related to regional agricultural systems and their societal and environmental costs and benefits. The project described here furthers our pursuit of this goal, through the following objective:
Creation of a comprehensive analysis model of the Central Wisconsin irrigated vegetable agricultural system, simultaneously considering environmental, economic, and social aspects of the relationship of this agricultural system to the region.
This project is based on two premises, which we will test:
Systems View
The project takes a systems view of the Central Sands irrigated vegetable production agricultural system (hereafter SIVS), in which we seek to elucidate properties which emerge from interactions among the components. The term "systems analysis" has many shades of meaning, but we are using the term in sense of Doyle : "The nub of the systems approach is a belief that the whole is more than the sum of its parts…separate parts are linked in an interacting manner and it is the interactions and inter-relationships between the various components which give the system its identity…." We believe that interactions between agricultural productivity, environmental quality, and economic opportunity will shape future study of agricultural systems.
One of the challenges in systems analysis is determining the bounds of the system . In the domain of agricultural systems the bounds are being expanded beyond individual crops to farms, farm households, and the ecosystem in which the biophysical farm resides . Expansion beyond the crop field brings the problem of boundary definition to the fore. A reasonable approach is to view any system as a human construct, an idea related to the concepts of "soft systems," "problem-defined systems," and "system of interest" ; . Certainly within a system under study there are defined biophysical domains, like a crop growing in a field, and simulation techniques are available for these , but the outer bounds of the system remains a human construct. We will describe the bounds we propose for our study of the SIVS, and then comment on some issues that were excluded.
The components of the SIVS are taken as:
Some of the relationships among these components are depicted in Figure 1. In particular, management practices on individual fields are nested within the array of cultivated land in the region, which in turn sits atop the groundwater flow system. The groundwater system receives recharge and contaminants from the production fields, and other land uses. Both practices in individual fields and the aggregated area cultivated influence economic yield. The economic yield is likely distributed in ways related to the number and nature of the jobs needed to operate the SIVS, and land tenure. Distributed economic benefits then impact economic opportunity in the region. Groundwater chemistry (presence and fate of contaminants) and flow impact the regional water resources, which are linked to health and infrastructure implications of groundwater quality, and ecosystem health issues.
Because this system is a human construct, every human who thinks about it for long will likely develop an alternative perspective of where the boundaries should be. For example, the human health standard for nitrate concentration is controversial and so a model of the health implications of alternative values for the standard is relevant. We have chosen to leave this issue external to the system for now, because of the predominant role that human value judgements take in setting the standard . We anticipate that the certainty to be ascribed to the current standard might become an issue of public debate if our model reveals the incompatibility of an economically vital SIVS and groundwater concentrations of nitrate at the current standard.
Another sub-system not present in our SIVS model is that of the sociology of farmer adoption of management practices that might reduce nitrate escape to groundwater. The sociology of farm household decisions is complicated, but appears computationally tractable . However, we do not currently have enough data to inform even the simplest model. Surveying of farmers within the SIVS is now underway and will in coming years prove useful to our analysis (P. Nowak, Dept. Rural Sociology, UW-Madison, personal communication, Jan. 1998).
A final example of a relevant but excluded phenomenon is the effect of fertilizers from upper Midwestern agriculture on a hypoxic "dead zone" in the Gulf of Mexico . The Wisconsin SIVS contribution to the fertilizer load of the Mississippi will someday be tallied and compared to other regional agriculture, but the issue is small relative to concern over groundwater nitrate concentrations in the Central Plain of Wisconsin.
A Similar Agricultural System Analysis
The Murray-Darling Basin of Australia is the major irrigated region in the country. The industry faces an uncertain future, however, due to rising water tables and salinization. To provide farmers and policy makers with insight into potential alternative scenarios a system simulation is being developed by the Bureau of Resource Sciences, Department of Primary Industries and Energy. Named the "Murray-Darling Basin Irrigation Futures Framework," the model links biophysical, production, and socio-economic models of the region. Users can explore the impacts on the economy and environment of changes to such factors as water markets, land use, and drainage schemes ; . A webpage description is: http://www.nric.gov.au/nric/projects/mdbirrig/mdbirrig.html.
The ambitious Australian effort serves as a template for much of what we will accomplish, although our SIVS involves a much smaller area, different crops, and a different agroenvironmental problem. Most significantly, the Australian model demonstrates that the software technology is in place to link disparate models operating at different resolutions. According to , "This framework has integrated models operating at four spatial scales and two temporal resolutions. The components include numerically intensive hydrological models which use a finite difference grid at daily time steps, linear algebra formulations of economic Input/Output models and more simple formulations of crop growth and response models."
Rationale and Significance
The fundamental motivation for this work is the belief that agricultural scientists in universities must develop expertise and tools to "interpret" farming systems to the general public. As agricultural land use and practices increasingly come into conflict with other sectors of society, decisions will be required about their mutual compatibility. University agricultural scientists can serve an important role in articulating the implications of societal decisions about agricultural systems. We are intrigued by the proposition put forth by that society has yet to develop an agricultural ethic to reconcile the need for food with the environmental impacts of farming. An informed debate toward this ethic will require a great deal of integrated and broad assessment of the impacts and contributions of farming systems.
The Wisconsin SIVS is an excellent model system on which to learn how to develop and use an integrated model of a modern agricultural system. The array of crops is relatively simple and well understood, and weather effects are relatively limited, given the irrigation and high level of pest control. Growers are well organized and aware of public concerns, and there is a sophisticated and aware citizenry in the urban areas. Leaching of agricultural chemicals has been studied for decades at the State Agricultural Experiment Station at Hancock in Waushara County (just to south), and results from numerous regional hydrogeological studies are available.
In addition to being well understood, the SIVS appears to have an additional characteristic that makes it interesting for exploring the relationship of an agricultural system to a region. That is, the SIVS does not appear to cause any degradation or depletion of its own resource base (in fact, some nitrate in the irrigation water may be a useful input). Thus threats to the sustainability to the SIVS are external and political. This aspect removes from the situation any traditional conservation motivation from farmer behavior, but might also make them more sensitive to perceptions of the broader public.
The major innovation of the work is the combination of agronomic, environmental, and economic aspects in a spatial and temporal framework. A comprehensive analysis of the SIVS without consideration of the spatial arrangement of the crop fields would be of little value to decision-makers and farmers concerned with groundwater impacts. The spatial context is essential to revealing significant time lags inherent in the system. For example, regulations on N fertilization or cropping patterns aimed at reducing groundwater nitrate might cause immediate reductions in employment, but much delayed impacts on water quality.
As the final item in this section, we speculate about questions that might be asked of the system by diverse user communities. These questions are one expression of our vision of how users (regulators, officials, farmers, public) will in the future approach decisions regarding the costs and benefits of the SIVS, and how our tool might contribute to the debate. Some example questions are:
Research Methods
The study area is described first, followed by how the research will be partitioned into tasks. Methods for each task will then be described.
Study Area
Initial model development will be focused on Portage Co., Wisconsin, because data availability is substantially better than in the other counties which contain portions of the SIVS. Figure 2 shows the county boundaries, major roads and the area in which most of the irrigated agriculture is conducted.
The dominant irrigated soils in the region were formed in outwash sand and gravel. They are well- to excessively drained and are nearly level to gently sloping. The most important series include Rosholt sandy loam, Plainfield loamy sand, and Billet sandy loam .
The behavior of nitrate and water in the irrigated potato production system are extensively studied. Part of this body of work illuminated the complex nature of water infiltration due to both the micro-relief created by tillage, and effects of natural layering in the soil (Saffigna et al., 1976; Donohue, 1990; Kung, 1990; Ju and Kung, 1997a; Ju and Kung, 1997b, and Ju et al., 1997). Fates of nitrogen fertilizer and implications for plant growth were also studied (Endelman et al.1974; Jury et al. 1976; Kirkham et al. 1974; Lesczynski and Tanner 1976; Liegel and Walsh 1976; Olsen et al. 1970; Saffigna and Keeney 1977).
The hydrogeology of the Central Sands region consists of an unconfined aquifer formed from glacial meltwater deposits, primarily glacial sands and gravel outwash, overlying Precambrian granitic bedrock. There is a considerable body of literature available on the hydrogeology of the Central Sands region (e.g., Holt 1965; Weeks et al. 1965; Weeks and Stangland 1971; Karnauskas 1977; Lippelt and Hennings 1981; Rothschild 1982; Bradbury and Rothschild 1985; Allen 1985; Stoertz 1985; Brownell 1986; Stoertz et al. 1991; Bradbury et al., 1992; Kraft 1997). Our own work in this region includes lake studies (Karnauskas and Anderson 1978), recharge studies (Riemersma et al. 1997; Stoertz and Bradbury 1989; Stoertz et al. 1991), investigations into the occurrence and transport of aldicarb (Rothschild et al. 1982; Chesters et al. 1982), and investigations of the role of drainage ditches in intercepting contaminants (Zheng et al. 1988a, b; Chambers and Bahr 1992; Meigs and Bahr 1995).
The hydraulic conductivity of the aquifer averages 5.1 x 10-4 m/s and recharge rate averages 9-10 inches per year (Kraft 1997). Field studies (Bradbury et al. 1992; Stoertz et al. 1991) in the Buena Vista Basin in Portage County south of Stevens Point indicate, however, that there is considerable spatial and temporal variability in recharge.
Portage County contains nearly one-half of the irrigated vegetable area planted in the SIVS (Table 1), with Waushara and Adams Counties accounting for most of the rest. Small areas of two other counties adjacent to these three might be considered within the SIVS, but their irrigated areas are small.
Table 1. Areas planted to the major processing crops in 1996 in the three counties which constitute the majority of the SIVS .
|
County |
Potatoes |
Snap Beans |
Sweet Corn |
|
- ha - |
|||
|
Portage |
9,900 |
5,800 |
3,400 |
|
Waushara |
6,200 |
2,900 |
2,500 |
|
Adams |
3,600 |
1,600 |
1,300 |
Agriculture is prosperous in the irrigated portion of the county, but nitrate and pesticide pollution from the industry is a major public issue. About 18% of county wells are estimated to exceed nitrate standards, but this number rises to 40% in some irrigated townships. Long term data from stream baseflow and municipal wells indicate nitrate levels are continuing to rise. Pesticide pollution of drinking water wells is also common, with 40-50% of county wells containing atrazine. The county is currently engaged in a groundwater management planning exercise to deal with the effects of agricultural pollution, but finds its efforts stymied by information gaps on the causes, effects, and economics of agriculture and groundwater pollution.
The municipalities of Stevens Point, Whiting, and Plover (populations 24000, 1850, and 8337 respectively) are clustered at the northwestern corner of the SIVS, and all rely on the glacial outwash aquifer for drinking water. While productive, the aquifer is highly susceptible to degradation due to current land uses and low pollution attenuation capacity of the overlying soils. Groundwater nitrate in excess of the drinking water standard forced closure of the Whiting municipal well from 1978 to 1991, during which time the municipality procured water from the City of Stevens Point. Completion of a nitrate removal facility constructed at a cost of $670,000 brought the well back into service. Plover wells began exceeding the nitrate standard in 1993, forcing the Village to install a nitrate removal facility costing over $2 million. A detailed study of the recharge zones of these municipal wells created for the USEPA was just completed by one of the Investigators .
System Development
Our approach for meeting the objective of this project is to create needed submodels, link these submodels with the whatIf? integration framework, and evaluate the usefulness of the tool. The following tasks were identified:
Task 1-Soil and Vegetation Model
The soil and vegetation model must simulate crop yield and nitrate leaching below the rootzone as a function of weather and management decisions. Crops that must be simulated are potato, sweet corn, and snap bean. Additionally, soybean, field corn, and small grain models will eventually be required to evaluate their roles as relatively low input alternative crops to include in rotations.
We are progressing with the soil N and the potato model, the most important of the crops. Potato is the most important in the SIVS economically, and appears to lead to the highest nitrate loading of the usual crops. The soil N and potato models are being developed by Mr. Seth Wilner and Dr. Bland, building on Wilner's M.Sc. thesis research . The model will be created to reproduce the results of years of field trials conducted at the Hancock Agricultural Experiment Station by Dr. Keith Kelling, and trials conducted on irrigated sands in Michigan and Minnesota by Drs. Vitosh, Ritchie (Michigan) and Rosen (Minnesota). Wilner's research is supported by grants to Drs. Bland and Kraft, including: "Dynamic Simulation of Soil Nitrate for N Fertilizer Management," and "Recovery of Nitrate Leached to Groundwater and Re-Use in Irrigation Water: A Feasibility Study."
Crops other than potato will be simulated using recognized models including CERES-maize and -wheat and CROPGRO-soybean, component models of the DSSAT family. Crops grown under the relatively stable conditions of the irrigated sands can be reasonably simulated using existing models.
Finally, for land areas not irrigated, e.g., forest or grasslands, only water use need be simulated because economic yield need not be carefully estimated and fertilizer is not applied. Evapotranspiration will be taken as the potential rate when the plant community-specific rootzone is at field capacity and decline linearly to nothing as plant available water is depleted. Recharge of groundwater from these areas will be the flux of water below the rootzone, at the very low nitrate concentration typical for such land uses.
Weather data for the model will be from the National Weather Service's Cooperative Observation Network, and the solar radiation database created and maintained at the Midwest Climate Center (see for an example application of these same databases).
The various vegetation models will be spatially distributed using GIS coverages of landuse. These are available from Portage County and will be maintained and manipulated by Dr. Kraft and his data specialist, at the Central Wisconsin Groundwater Center of the Univ. of Wisconsin-Stevens Point. This team will also maintain and increase their database of hydrogeologic information for the area, in support of Task 2.
Task 2-Water Resources Model
To capture the temporal and spatial aspects of groundwater movement beneath the SIVS, we will develop a three-dimensional, finite-difference numerical flow and transport model. The groundwater model will consist of two components: a flow model and a transport model. The composite model will simulate groundwater flow and nitrate transport in response to changes in landuse and agronomic management.
A multilayer (three-dimensional), finite-difference grid will be superimposed over the study region. Model boundaries will be defined using physical and hydraulic boundaries. Input parameters (hydraulic conductivity, porosity, depth to bedrock, specific yield, dispersivity, and recharge) and head and concentration values needed for model calibration will be obtained from the literature and from relevant agencies, e.g., the Wisconsin Geological and Natural History Survey (WGNHS), the Wisconsin Department of Natural Resources and the Central Wisconsin Groundwater Center).
MODFLOW, a commonly used and rigorously verified, three-dimensional, finite-difference groundwater flow code (McDonald and Harbaugh, 1988), will be used to develop the flow model using standard modeling protocol (Anderson and Woessner, 1992). Briefly this protocol is as follows: develop a conceptual and mathematical model, use MODFLOW to build the numerical model, calibrate the numerical model using field data, and finally run predictive simulations to assess proposed management schemes. Hydrogeological parameters will be obtained from published reports, maps, and existing well logs of the area. The model output will consist of hydraulic head values at nodal points. Field-measured head values will be used to calibrate the model.
In previous work (Riemersma et al. 1997; Stoertz and Bradbury 1989) we constructed groundwater flow models for the Buena Vista Groundwater Basin in Portage County south of Stevens Point. Information used to construct these models can be transferred to use in the proposed project.
MT3D96, a widely used, three-dimensional, transport code (Zheng, 1996), will be used to simulate the transport of nitrates. The head solution from MODFLOW will be input to MT3D96 and used to generate a groundwater velocity distribution. This distribution, together with groundwater recharge rates and nitrate loading data provided by the agronomic model, and estimates of dispersivity, will be used to simulate nitrate transport subject to advection and dispersion. Nitrate will be modeled as a chemically conservative contaminant. Model output will consist of a value of nitrate concentration for each grid cell.
These two models will be accessed through the Windows-based Groundwater Vistas platform, a pre-and post-processing system (Environmental Simulations Inc., 1996) that facilitates input of parameter values and visualization of results.
Task 3-Economic Model
The economic analysis developed in this model centers on estimating the on-farm and off-farm net benefits associated with crop production and the distribution of those net benefits among sectors and groups within the region. Because there are no major negative feedbacks between current agronomic practices and future production possibilities (e.g., salinization or depleting ground water supplies), dynamic concerns with the evolution of soil and water quality are not explicitly incorporated into estimates of net-benefits. Similarly, because the Central Sands region is already highly commercialized, the potential dynamic effects of changes in agrarian structure and resulting land use patterns are also ignored. In other words, an essentially static approach is taken to estimating the level and distribution of net-benefits associated with alternative crop production practices. The main sustainability concern explored in this effort is the policy one associated with the environmental and health concerns surrounding groundwater contamination and the costs of remediation or of health problems.
The innovative aspect of the economic modelling emerges in the effort to use a conjoined input-output/econometric model to devise estimates of both direct and indirect economic impacts of changes in the levels of vegetable crop production and processing in the Central Sands region. This effort involves three steps, the first two being classic input-output type analysis, and the third which integrates the outcomes of steps 1 and 2 into an econometric model of impacts on the rest of the regional economy (as measured by income and employment, government fiscal changes, housing market effects, and changes in commuting patterns). In this way, changes in the regulatory structure or in farm production practices can be traced through to their effect on the regional economy. However, the changes have to be fairly substantive (more than a few percentage points in the volume of production) in order to provide meaningful comparative scenarios.
In the first step, available data on crop yields, prices and input costs will be combined to estimate the direct economic yields of the SIVS. Because the input costs to that production include some locally provided goods and services (e.g. fertilizers, manual labor, tax accountants), it will be necessary to construct representative farm budgets with an eye toward how the value of locally provided inputs might vary depending on the crop. Some additional data on crop production outcomes may also need to be gathered from cooperating farmers. The key measurable output of the first step in the economic model will be the level of income or value-added produced locally by the observed pattern of vegetable crop production and the distribution of that income among different participants in the industry (farm-owners, farm workers, input providers, and so on).
In the second step, an attempt will be made to assess the economic impacts associated directly with the processing and distribution of the vegetables. For this stage of the model, the crops will be viewed as an essential input to a larger processing operation, in the sense that without the local production of these vegetables the plants would be unlikely to operate in the region. Data will need to be gathered on the cost structure of the vegetable processing and distribution units operating in the area, but will build on existing research done by Deller and other researchers cited below. The local value-added (labor, goods, and other service inputs provided by local sources) will be estimated and added to the farm benefit estimates developed in the first step. The accounting process here should also provide a view on how those benefits are distributed within the local area among the various participants in the industry.
Finally, to simulate the economic changes in on-farm production and agricultural processing on the local economy, the Wisconsin Economic Impact Modeling System will be employed (Shields 1998; Deller and Shields 1996; Deller, Shields and Stallmann 1998; Shields and Deller 1997). The Wisconsin System (WS) is a conjoined input-output/econometric model of Wisconsin county economies. Using expenditure pattern data (employee wage expenditure patterns and non-employee related firm expenditures) to describe alternative production regimes, the input-output component of the WS computes changes in industrial output through the direct and indirect rounds of impacts. Changes in industrial output then are used to estimate changes in employment, wages (income and poverty), commuting patterns, unemployment and population. These estimated changes are then used to simulate fiscal impacts on local governments through expenditures and revenues, the local housing market through new construction and housing prices, and local retail markets through consumer expenditures. Effects on the local economy and community associated with changing agronomic practices and farm regulations can thus be traced by comparing and contrasting the outcomes across scenarios.
Task 4-Integration to a System Analysis Tool
Integration of the agronomic, water resource, and economic models into a single framework allows us to see how these system components interact. How does turning off 20% of the irrigation systems each year affect employment and water quality? An interactive, coupled system will allow such questions to be explored. The infrastructure that makes it possible to couple these disparate models is a commercial suite of software called whatIf? from ROBBERT Associates of Ottawa, Canada. This software is an "integration framework system" in the classification of Malafant and Fordham (1997); he further described it as, "…an object-oriented scenario modelling package providing a structured set of tools for groups of experts to interact, express their ideas and apply concepts to achieve resolutions to debate(s) of economically and ecologically sustainable resource issues." Information on whatIf? is available at the ROBBERT Associates website, http://www.magma.ca/~robbert/.
We will take advantage of the several years of experience accumulated within the Australian study through continuing informal (at present) collaboration with Mr. Kim Malafant, who is Manager of Advanced Modelling Systems, National Resource Information Centre, Bureau of Resource Sciences, of the Department of Primary Industries and Resource Sciences. The Project Director of this proposal met Mr. Malafant at the 1997 American Society of Agronomy Meeting, and is planning to visit Mr. Malafant's laboratory in the first half of 1998.
Mr. Malafant collaborates with ROBBERT Associates on extending the capabilities of whatIf?, particularly in the area of simplifying linking to it process models still in the form and language in which they were developed. It is this capability that will make it possible for us to join our agronomy models, MODFLOW, and our economic input-output models into a linked system analysis of SIVS. Thus whatIf? provides the environment, control capabilities, and output management and display to answers to diverse questions posed of the SIVS. A question leads to formulation of alternative scenarios, which are then simulated by the various component sub-models, under the direction of the whatIf? model of the SIVS.
In operation, whatIf? consists of a UNIX-based server and PC clients, linked by the Internet. Client software may be freely distributed, although the license limits the number of simultaneously connected users. The simulation language and code, scenario modeler, and documentation functions reside on the server. Clients linked to the server can formulate scenarios, cause them to be executed by the server, and then review any output variable in several formats. For the diverse set of users that we envision for this model, we anticipate that previously developed and executed scenarios will provide answers to most of their questions. Particularly motivated users, however, will be able to formulate additional scenarios.
Task 5-Assessment of Impact
The usefulness of our model of the SIVS ultimately depends on whether the system provides timely and reliable information that is relevant to decision making and is understandable to the intended audience(s). Evaluation of a system requires examination of each of these characteristics.
The quality of information can be judged as part of the system development process. Input data will be characterized in terms of its specificity, accuracy, and so forth. Most of the data will reside in a geographic information system, so the Federal Geographic Data Committee's Spatial Data Transfer Standard and the Content Standards for Geospatial Metadata will provide criteria and a framework for documenting data quality. The reliability of processed information (e.g., model output) will be evaluated as part of the model development process through sensitivity testing, error propagation estimates, and comparison to "ground truth" and other sources of greater accuracy.
The relevance of information from the system must be judged in the context of use. Although a broader potential range of users exist, we will confine our evaluation to growers and those making policy decisions at a local level (town, village, county) who could influence agricultural practices or land use options. The initial sub-task will be to characterize these two audiences' information requirements through a needs assessment. Interviews will be conducted with growers and decision-makers during early phases of implementation of the system to determine requirements such as information content, form, delivery mechanism, timeliness and accuracy. These interviews will both guide system development and provide the basis for evaluation: does the system fulfill the design criteria?
Two approaches will be used to evaluate the relevance of the completed system: "internal" evaluation of components such as information content and interface design based on criteria established from the original needs assessment, and "external" judgment from the intended audiences in pre- and post- use observations. We will adapt a structured approach for internal evaluation (e.g., Johnston 1984). In addition to looking at whether the
SIVS model provides reliable information in an understandable form (as specified in the original needs assessment), the internal evaluation will include accounting (how much does it cost to gather data, run the system, and generate output). The external evaluation will put the information provided into an organizational context. It is quite likely that users' expectations will change, both as a function of intervening time during the development period and as a function of actually using the system. Pre-use interviews will capture user expectations at the time of implementation, with questions similar to the needs assessment on content, form, delivery and reliability. Post-use interviews will be used to observe users reactions to the system, particularly the relevance of its information to the kinds of decisions they make, and to record changes in expectations.
Meeting Our Expectations
Relative to our early statements of objective and premises, how will we judge our success? Our objective is to create an integrated model of the SIVS. The final product can only be judged to have met this goal subjectively, because in such an endeavor there is a large gray area between success and failure. The more significant test of the results of the project will be whether the model has usefulness to user communities like the growers' association, local government officials, state regulators, and concerned citizens. We will gain insight into this from the assessment activities in Task 5. Our subjective evaluation of the model and the reactions of users will allow us to judge if either or both of the premises stated earlier are in fact true.
References
Addiscott, T. M., A. P. Whitmore, 1991. Farming, fertilizers and the nitrate problem. Oxford University Press, New York.
Allen, R. M., 1985, Slug test determination of hydraulic conductivity in the Central Sand Plain of Wisconsin, WGNHS, unpublished report.
Anderson, M.P. and W.W. Woessner, 1992. Applied Groundwater Modeling, Academic Press, 381 p.
Armour, D. L., L. R. Massie, W. L. Bland, G. D. Bubenzer 1997. Specification of center-pivot irrigation based on load-control constraints. Trans. ASAE 40: 89-95.
Beardsley, T. 1997. Death in the deep. Scientific American (Nov.): 17,20.
Bradbury, K.R., J.M. Faustini, and M.W. Stoertz, 1992. Groundwater Flow Systems and Recharge in the Buena Vista Basin, Portage and Wood counties, Wisconsin, Wisconsin Geological and Natural History Survey Information Circular 72, 31 p.
Bradbury, K.R. and E.R. Rothschild, 1985. A Computerized Technique for Estimating the Hydraulic Conductivity of Aquifers from Specific Capacity Data. Ground Water 23(2), p.240 - 246.
Brownell, 1986, Stratigraphy of unlithified deposits in the Central Sand Plain of Wisconsin, Univ. of Wisc.-Madison, M.S. thesis, 172 p.
Butler, K. S. 1978. Irrigation of the Central Sands of Wisconsin. Res. Bull. R2960, University of Wisconsin-Madison, Madison, WI
Chambers and Bahr, 1992, Tracer test evaluation of a drainage ditch capture zone, Ground Water 30(5),667-675.
Chesters, G., M.P. Anderson, B. Shaw, J.M. Harkin, M. Meyer, E. Rothschild, and R. Manser, 1982, Aldicarb in groundwater, Univ. of Wisconsin Water Resources Center, Special Report, 38 p.
Deller, Steven C. and Shields, Martin. 1996. "The Wisconsin Fiscal Impact Model: Preliminary Results." Paper presented at the Mid-Continent Regional Science Association Meeting, Madison, WI (June 6-8).
Deller, Steven C., Shields, Martin, and Stallmann, Judith. 1998. "The Aging of a Small Rural Region: An Application of the Wisconsin Economic Impact Modeling System." Paper presented at the Western Regional Science Annual Meetings. Monterey, CA (February 19-21).
Dent, J. B., G. Edwards-Jones, M. J. McGregor 1995. Simulation of ecological, social and economic factors in agricultural systems. Agricultural Systems 49:337-351.
Donohue, S.V. 1990. Microrelief and preferential flow: factors affecting solute leaching and vadose zone monitoring. M.S. thesis. Univ. of Wisconsin-Madison.
Doyle, C. J. 1990. Application of systems theory to farm planning and control: Modelling resource allocation. p. 89-112 in: Systems theory applied to agriculture and the food chain. J. G. W. Jones and P. R. Street., Elsevier, New York.
Endelman, F.J., D.R. Keeney, J.T. Gilmour, and P.G. Saffigna. 1974. Nitrate and chloride movement in the Plainfield loamy sand under intensive irrigation. J. Environ. Qual. 3:295-298.
Environmental Simulations, Inc., 1996. Guide to using Groundwater Vistas.
FGDC 1997. Federal Geographic Data Committee: Framework Introduction and Guide. Washington DC, FGDC.
Faustini, J.M., 1985. Delineation of groundwater flow patterns in a portion of the Central Sand Plain of Wisconsin, M.S. thesis, 117 p.
Fordham, D. P. and K. W. J. Malafant (1997). The Murray-Darling Basin irrigation futures framework (IFF?). MODSIM 97 International Congress on Modelling and Simulation., Hobart, Tasmania, Australia.
Grant, W. E., E. K. Pedersen, S. L. Marin 1997. Ecology and natural resource management systems analysis and simulation. John Wiley, New York.
Hole, F. D. and C. E. Germain 1994. Natural divisions of Wisconsin (map). Wisconsin Department of Natural Resources, Madison, WI.
Holt, C.L.R. Jr., 1965. Geology and Water Resources of Portage County, Wisconsin. USGS Water-Supply Paper 1796.
Ison, R. L., P. T. Maiteny, S. Carr 1997. "Systems methodologies for sustainable natural resources research and development." Agricultural Systems 55: 257-272.
Johnston, J. 1984. Evaluating the New Information Technologies. Jossey-Bass, San Francisco.
Ju, S.H., and K.J.S. Kung. 1997a. Impact of funnel flow on contaminant transport in sandy soils: Numerical simulation. Soil Sci. Soc. Am. J. 61:409-415.
Ju, S.H., and K.J.S. Kung. 1997b. Steady-state funnel flow: Its characteristics and impact on modeling. Soil Sci. Soc. Am. J. 61:416-427.
Ju, S.H., K.J.S. Kung, and C.S. Helling. 1997. Simulating impact of funnel flow on contaminant sampling. Soil Sci. Soc. Am. J. 61:427-435.
Jury, W.A., W.R. Gardner, P.G. Saffigna, and C.B. Tanner. 1976. Model for predicting simultaneous movement of nitrate and water through a loamy sand. Soil Sci. 122:36-43.
Karnauskas, R.J., 1977, The hydrogeology of the Nepco Lake watershed in central Wisconsin with a discussion of management implications. Univ. of Wisc., M.S. thesis, 249 p.
Karnauskas, R.J., and M.P. Anderson, 1978, Ground-water lake relationships and ground- water quality in the Sand Plain Province of Wisconsin - Nepco Lake, Ground Water 16(4), 273-281.
Kirkham, M.B., D.R. Keeney, and W.R. Gardner. 1974. Uptake of water and labeled nitrate at different depths in the root zone of potato plants grown on a sandy soil. Agro. Ecosystems 1:31-44.
Kraft, G.J., 1997. Contaminant Source Assessment and Management Using Groundwater Flow and Contaminant Models in the Stevens Point - Whiting - Plover Wellhead Protection Area. Central Wisconsin Groundwater Center.
Kung, K.J.S. 1990. Preferential flow in a sandy vadose zone: I. Field observation. Geoderma 46:59-71.
Leopold, A. 1949 A Sand County Almanac. Oxford Univ. Press., New York.
Lesczynski, D.B., and C.B. Tanner. 1976. Seasonal variation of root distribution of irrigated, field-grown Russet Burbank potato. Am. Potato J. 53:69-78.
Liegel, E.A., and L.M. Walsh. 1976. Evaluation of sulfur-coated urea (SCU) applied to irrigated potatoes and corn. Agron. J. 68:457-463.
Lippelt, I.D. and R.G. Hennings, 1981. Irrigable Lands Inventory - Phase I, Groundwater and Related Information. Miscellaneous Report 87-1. Wisconsin Geological and Natural History Survey, Soil Conservation Service, Madison, Wisconsin. 13 pp.
Olsen, R.J., R.F. Hensler, O.J. Attoe, S.A. Witzel, and L.A. Peterson. 1970. Fertilizer nitrogen and crop rotation in relation to movement of nitrate nitrogen through soil profiles. Soil Sci. Soc. Am. Proc. 34:448-452.
Malafant, K. W. J. and D. P. Fordham 1997 An irrigation futures decision framework: Integrating biophysical, production and socioeconomic models. Agron Abs. p.23 ASA, CSSA, SSSA.
McDonald, M.G. and A.G. Harbaugh, 1988. A modular three-dimensional finite difference groundwater flow model. U.S. Geological Survey Open-File Report 83-875. 528 pp.
Meigs, L.C., and J.M. Bahr, 1995, Three-dimensional groundwater flow near narrow surface water bodies, Water Resour. Res. 31(12), 3299-3307.
Otter, A. J. and W. D. Fiala (1978). Soil survey of Portage County, Wisconsin
Riemersma, P.E., M.P. Anderson, and K.R. Bradbury, 1997, Sensitivity of Recharge Mapping Method to Scale of Discretization, AGU Spring meeting, Eos 78(17), p. S125.
Rothschild, E.R., 1982. Hydrogeology and Contaminant Transport Modeling of the Central Sand Plain. M.S. Thesis, University of Wisconsin, Madison, Wisconsin. 135 pp.
Rothschild, E.R., R.J. Manser, and M.P. Anderson, 1982, Investigation of aldicarb in groundwater in selected areas of the Central Sand Plain of Wisconsin, Ground Water 20(5), 437-445.
Saffigna, P.G., and D.R. Keeney.1977. Nitrogen and chloride uptake by irrigated Russet Burbank potatoes. Agron. J. 69:258-263.
Saffigna, P.G., C.B. Tanner, and D.R. Keeney. 1976. Non-uniform infiltration underpotato canopies caused by interception, stemflow, and hilling. Agron. J. 68:337-342.
Shields, Martin. 1998. An Integrated Economic Impact and Simulation Model for Wisconsin Counties. Unpublished PhD dissertation, Department of Agricultural and Applied Economics, University of Wisconsin-Madison.
Shields, Martin, and Deller, Steven C., 1997. "A Conjoined Input-Output / Econometric Model for Wisconsin Counties." Paper presented Annual Meetings of the Southern Regional Science Association. Memphis, TN (April 17-19).
Stoertz, M.W., 1985. Evaluation of Groundwater Recharge in the Central Sand Plain of Wisconsin. M.S. Thesis, University of Wisconsin, Madison, Wisconsin. 159 pp.
Stoertz, M.W., 1989. A new method for mapping groundwater recharge areas and zoning recharge for an inverse model, PhD thesis, University of Wisconsin-Madison, 178 pp.
Stoertz, M.W. and K.R. Bradbury, 1989. Mapping recharge areas using a groundwater flow model- a case study, Ground Water 27(2), p. 220-228.
Stoertz, M.W., M.P. Anderson, and K.R. Bradbury, 1991, Field investigations and numerical studies of groundwater recharge through unsaturated sand: A methodology applied to Central Wisconsin, WGNHS Info. Circ. 71, 52 p.
Thompson, P. B. 1995. The spirit of the soil-Agriculture and environmental ethics. Routledge New York,.
Veitch, S. M., D. P. Fordham, K. W. J. Malafant 1993 WhatIf? scenarios to explore land management options. Applications of advanced information technologies: Effective management of natural resources., Spokane, WA, American Society of Agricultural Engineers.
Weeks, E.P., D.W. Ericson, and C.L.R. Holt, Jr., 1965. Hydrology of the Little Plover River Basin, Portage County, Wisconsin, and the Effects of Water Resource Development. USGS Water-Supply Paper 1811.
Weeks, E.P. and H.G. Stangland, 1971. Effects of Irrigation on Streamflow in the Central Sand Plain of Wisconsin. USGS Water Resources Division, Open File Report.
Wilner, S. A. (1998). Potato nitrogen management under two irrigation systems. Soil Science. Madison, University of Wisconsin-Madison Wis. Agric. Statistics Service 1997 Wisconsin 1997 agricultural statistics., Wisconsin Dep. of Agriculture, Trade, and Consumer Protection, Madison, WI
Wisconsin Department of Natural Resources, 1987. Groundwater contamination susceptibility in Wisconsin, Wisconsin Department of Natural Resources Report 5, Wisconsin State Groundwater Management Plan, 27 p.
Zheng, C., K.R. Bradbury, and M.P. Anderson, 1988a, Role of interceptor ditches in limiting the spread of contaminants in ground water, Ground Water 26(6), 734-742.
Zheng, C., H.F. Wang, M.P. Anderson, and K.R. Bradbury, 1988b, Analysis of interceptor ditches for control of groundwater pollution, Jour. of Hydrol. 98, p. 67-81.
Home || Vision
|| Background || Model
|| Tutorial
Literature || Calendar
|| Links || Contact
Us