The Palouse Basin Participatory Model Pilot Project

Typically water models are designed with software systems such as Modflow or Riverware. These hydrologic models help scientists understand the storage and movement of water in aquifers or river systems. These models can capture the effects of pumping from such systems but do not include social issues that affect demand. System dynamics helps us to define a more holistic "system" that includes human elements. System dynamics software allows modelers to use insights learned from hydrologic studies and combine them in a model that can include economics, growth, conservation and other non-physical parameters of the system as a whole. Furthermore system dynamics simulations can be run in a matter of seconds so that users may explore multiple "what if" scenarios in a short period of time. [More on system dynamics]

The Palouse Basin Participatory Model currently includes (or will soon include) parameters such as aquifer storativity, recharge, pumping history, projected demand and potential conservation by Pullman, WSU, Moscow and UI, current and potential use of reclaimed water, a hypothetical reservoir, aquifer storage and recovery (ASR) and the potential contributions of individual conservation measures such as low flow toilets or reduced lawn watering. To compare potential costs of such efforts the group is using dollars per gallon of water; i.e. cost of conservation measures or cost of "new water" development.

The management of natural resources is becoming increasingly complex. Agencies which manage resources have come to realize the value of public input into the decision making processes and encourage participation in a variety of forums which attempt to both educate and gather public sentiments. Due to the complexity of explaining often highly scientific information and juggling a plethora of social values, agencies and communities are exploring the use of participatory modeling processes using system dynamics. [More on participatory modeling].

This methodology uses a common language to integrate various types of information into simulation models. These models assist stakeholders with problem definition and the evaluation of potential management or policy alternatives. The process of building a model helps stakeholders clarify their own mental models and gain a better understanding of important scientific relationships. (Beall 2007)

Building the Palouse Basin model has followed an iterative process that includes workshops and model building between workshops. Modelers build simulations of their understanding of stakeholder concerns expressed during the group meetings. They then return to the group with a simulation model to be vetted by the group. The model continues to be developed with the input and commentary described at each meeting. The process encourages members of the groups to explore individual mental models of water issues on the Palouse. It has promoted discussion of scientific uncertainty, and individual perceptions of availability and conservation. The iterative nature of the process promotes a shared vision of "the problem and potential solutions" through discussion and model development.

Participatory modeling can add value to a group process in a variety of ways. It can improve comprehension of the problem, and assist with group dynamics and group learning. It can be used to integrate professional science and different forms of local knowledge including street science and experiential knowledge. (Beall 2007)

Participatory modeling using system dynamics creates a transparent nexus of science, policy options, social concerns and local knowledge that enhances discussion of issues surrounding the use of natural resources. This process uses the tenets of scientific theory, hypothesis testing and clear statements of assumptions.

The pilot project was designed to help participants "explore the use of participatory modeling to assist with water resource management decisions". As part of the project the group is identifying what the process should produce. One product will be a presentation to the 2008 Palouse Basin Water Summit that includes an example of "what if" scenarios and insights to the use of this kind of process. If the participants decide that this is something that could assist with water management on the Palouse they will collaboratively design a process, which would potentially include more participants, to create a decision support tool. A decision support tool can be used to store information, as a venue to discuss scientifically uncertain parameters and clarify assumptions, as a venue to explore different social values (i.e. growth and conservation), and to assist with public education.

Palouse Basin Participatory Model Workshop 1
March 14, 2008: Jack's Creek Conference Room, University of Idaho

The morning began with a brief introduction which included a synopsis of PBAC's Mission Statement which is quoted below from the PBAC website:

2006 PBAC Mission and Goals

Mission: To ensure a long-term, quality water supply for the Palouse Basin region.

Goals:

• Develop and Implement a balanced basin wide Water Supply and Use Program by 2020.
• Create an action plan for aquifer system enhancement and alternate water supply development by 2010.
• Direct research and implement pilot projects necessary to understand the basin hydrogeology in a manner sufficient to support the Water Supply and Use Program and the affiliated supply projects.
• Update the Palouse Basin Groundwater Management Plan to reflect the Water Supply and Use Program.
• Encourage and facilitate entities in meeting their specific pumping, conservation, efficient use, water recycling and other goals.
• Educate entities and the public on the state of the basin water supply and options for carrying out PBAC's mission and goals.
• Maintain harmonious and effective working relationships across the state line to fairly meet the needs of all entities.
• Revisit the goals and mission statement annually to ensure that they are current and reflect the goals of the group.

The meeting agenda included a discussion of ground rules, an introduction to collaborative modeling and system dynamics, creating a Palouse Basin timeline, two model mapping exercises, and a discussion of where the process should go from here.

Due to the fact that the participants had never been exposed to a participatory modeling process it was suggested that the topic of ground rules be left until the end of the day. It was noted that until the group of attendees understood what they were getting in to the concerns of ground rules may be irrelevant. At days end the topic was revisited with the following considerations: 1) The meetings are technically open meetings due to the nature of the participants however the meeting room can only hold so many people. 2) The group had already experienced the first set of system dynamic exercises together. In consideration of this, all those on the original invitee list would be considered part of "the participatory group" (even those who were unable to attend and those originally considered observers). Any others that may show up in the future would be welcomed as observers. 3) The group would also like the pilot project to be presented at the 2008 Water Summit and at the 2008 American Planning Association Regional Meeting in Spokane. 4) Washington Dept. of Ecology and Idaho Dept. of Water Resources representatives will be involved in any and all capacities that they can offer.

An introduction to collaborative modeling and system dynamics followed. (see more on system dynamics and participatory modeling)

The first group exercise was a collaborative effort to create a Palouse Basin timeline (jpg, pdf). Recognizing our collective knowledge and the historical context of our current situation helped the participants view the aquifers in the context of the greater system. Participants were asked the following question: How did the system get to the place that it is today?

The first model mapping exercise had the participants working in three groups who were asked to define the stocks, flows and parameters that were important to the Palouse water system. During a plenary session the groups identified stocks, flows and parameters that were common to all three groups. These considerations were split into issues that affected demand and issues that affected supply. Stocks included water in the Grande Ronde, water in the Wanapum, and populations of people in Moscow, UI, Pullman and WSU. There was discussion of the issue of "double counting" people if the model was based on per capita use. Flows included pumping by the four entities, Palouse and Colfax. Rates and other parameters that affected the flows included: precipitation, evapo-transpiration, recharge, growth, conservation, education, social capital, and numerous other factors. The participants recognized that many of the issues which affect the use of water were extremely difficult to quantify or predict but none the less were very important to the behavior of the system.

The Second model mapping exercise split the participants into two groups, one working to create a picture of a supply model and one group working on demand issues. Though somewhat ambiguous this exercise helped to clarify relationships and to further distill the participant's mental models of the system.

During the final plenary the group revisited the ground rules as noted above and discussed who had data for the modelers. The group agreed that it was appropriate for the modelers to talk with individuals on a one to one basis as needed for data and clarification. The group decided that they wanted to go forward with the project and agreed to meet again on April 11, 2008 from 9-12. They were also asked to complete a survey to assess the day and the potential usefulness of this type of approach. Most of the surveys indicated that the participants were optimistic about the use of participatory modeling for explaining scientific information and social concerns.

Palouse Basin Participatory Model Workshop 2
April 11, 2008: Jack's Creek Conference Room, University of Idaho

After a welcome, introductions and quick review of last workshop there was a model review session. A simple generic model of supply and demand was presented to familiarize participants with model interfaces, sliders (which allow the user to change a parameter instantaneously) and the impacts of a percent increase in population. The participants were shown the impact of a 1% population growth over a sustained period of time and reminded that the impact of a percentage growth rate is exponential.

Participants then reviewed the first version of the pilot project model and shown the disclaimer on the first page which indicates that this is an experimental pilot project. This was added because participants had voiced concern that "someone" may get a hold of the model and decide that the parameters in that version of the model reflected a consensus of the group. It was noted that it would be difficult for people, who were not in the room when a model version was explained, to understand the current assumptions. For example, this first version set the potential storage as the area above pump level with an option to lower pumps (pump level in model is an average and the actual varies). This created an impression to some that we were running out of water very quickly. (In response to this concern the version presented at the next meeting reflects an average basement number for well depth.) Concern was expressed with the current infrastructure approach in general because it masks that water exists below pump level. It was suggested that the model use a slider with a different name than "pump level" to illustrate depth of aquifer. The participants were reminded that system models should be used for trend analysis rather that point prediction tools but it was acknowledged that "people in general" do not understand this type of model analysis and that care should be taken as to how the model was used.

This inspired discussion as to whether model is to be used for public education or decision making. The potential of multiple versions was considered with a technical version for internal use and an educational version for use with the public that a used a "simple water balance" approach.

Further discussion inspired questions on the impact of well casing and discussion of leakage between aquifers due to poor finishing of wells. It was noted that the large (newer) municipal wells are cased almost all the way to the bottom.

The supply side of the model included values for potential storage of the aquifers, potential storativity, and potential recharge. The demand side was based on per capita demand with some generic per capita or bulk gallons per year sliders for conservation. Initial model structures to include re-treatment and a recharge surface reservoir were added to illustrate how these components could be added to the model. There was concern expressed with double counting. For example WSU per capita use was based on an initial population 17,000, and Pullman on an initial population of 25,000. This was addressed in later versions of the model with the Universities segregated as "pumping entities" rather that "residential entities".

There were a variety of issues discussed at this point. The facilitators decided to develop a "parking lot" of requests and information that the participants felt were important but that 1) the modelers may not be able to include in the pilot project model due to time constraints and complexity or 2) the time needed to gather data or 3) the lack of data altogether which would not preclude the inclusion of a particular parameter in the model but rather require discussion and consensus on a value or range of values.

The “Parking lot” discussion and comments included the following issues:

• Separate pump levels by well to reflect the depth of some of the newer wells
• Use depth to basement rather than depth to well pump to avoid an artificial view of less water available
• Split recharge efficiency to reflect relation to surface water/instream flow
• Leakage rate between aquifers
• Spring loss from Wanapum
• Elevation level-cap on decline of water (elevation of streams in the system) Possibly set cap when elevation drops below the surface elevation.
• Additional exempt wells will reduce recharge
• Shallow rural use could be affected by this
• Washington has minimum flows on Paradise Creek and Palouse River and closure to new development in the summer [Surface Water Source Limitation – SWSL]
• Per capita use – tease out details of split between Universities and cities
• Estimates of rural groundwater use
• Scope of the basin that the model should cover
• Rainwater harvesting (legalities)
• Supply Augmentation (watering landscape)
• Detention ponds
• Pipelines from NF and Snake
• In Stream flow, seasonal availability, use of.
• Run the model on a seasonal basis?
• Add Colfax and Palouse? Are they in the basin?
• Change the model units and convert centimeters to inches and kilometers to miles
• Growth rates of universities and cities cannot operate independently
• Start developing FAQs to guide graph formation
• Long term impacts of climate change
• Costs of various conservation and supply measures
• External source ranking of alternatives
• ASR, where does the water come from?
• Representation of the costs of supply, and ease of implementation, impacts on and from land purchases
• Would like to be able to use the model for economic decisions (new industries, etc.)

The length of this list and the list of topics and details that are important to the participants indicates the complexity of water management in the basin. Surveys taken at the end of the session again indicated that the participants felt optimistic about this type of approach. Some indicated that they were getting a better understanding of the scientific issues. It was also noted that people were starting to understand that there were diverse perspectives of the system.

Palouse Basin Participatory Model Workshop 3
May 2, 2008: Jack's Creek Conference Room, University of Idaho

In response to comments from Workshop 2 the following changes were made to the model between workshops:

• Units were changes to inches/gallons/miles
• Pump levels were dropped to an average well basement depth of both aquifers
• 2005 pump rates = initial conditions
• Universities as "corporation" with cities reflecting population growth
• Bulk rate conservation – shows up in overall pump rate reduction rather than entered as per capita reductions
• "Imagination reservoir" with water coming from "Barelya Creek" was added to illustrate the impact of a potential reservoir. It was available to offset pumping from the Wanapum
• Uncertainty issues with high sensitivity – dealt with by picking middle, but can model several scenarios to bracket risk [worst case/best case]
• Storativity
• Wanapum recharge efficiency
• Grande Ronde recharge
• Possible next additions
• Scenarios for addressing uncertainty with high sensitivity (see below)
• Add specific conservation and supply measures
• Add cost estimates
• Sensitivity analysis

Several issues were illustrated though model simulations and then discussed.

A demonstration of the impact of storativity values upon the system was shown to the participants to illustrate the sensitive nature of the highly uncertain storativity values on the model. It was noted that storativity in Grande Ronde is a range of values reflecting studies at places like Hanford. Numbers from pump tests are both short term and localized, so are difficult to translate to a larger spatial-temporal time. Participants discussed that one way to address uncertainty in storativity and recharge (both of which were not likely to be resolved in near future) would be first to eliminate unrealistic scenarios [note that back calculation from water balance gives numbers that are unrealistic according to the geologists]. Then, run the model at 2 different scenarios [worst case/best case] and look at management options for each.

This inspired discussion about the Wanapum recharge numbers. The efficiency of capture is based on the understanding that the formation is less interconnected. It was noted that that we are beginning to have sufficient information to separate this out. This level of understanding does not apply to the Grande Ronde.

A scenario that operated with the "Imagination reservoir" addition with 1cfs year around basically solves the problem; however concerns such as cost, engineering, and water rights were not considered in the scenario.

For purposes of the pilot project participants recommended that "Supply/Demand Scenarios" be modeled as reductions in pumping and include cost estimates of those reductions. This would allow for simulations that could offer a comparison between approaches for simplicity. These simulations could be run with the best case/worst case options of storativity.

The following "supply/demand" parameters were requested for inclusion in the model:

• ASR: Preliminary estimates exist on cost and volume although uncertainty remains on water rights and on the efficiency of capture.
• Waste Water Reclamation
• Reservoir with realistic numbers
• Kamiak Butte recharge
• Conservations measures from Pullman and WSU water plans
• Lawn watering reductions
• Tiered pricing [data for this location are not available]
• Specific growth scenario which could illustrate impact of one new industry – e.g brewing company and reflect industry use and growth in number of households

Participants then discussed where to go with pilot and whether to extend it to a larger project. It was noted that the pilot process model could be used as an educational tool for public discussion. It could be used to illustrate comparison of costs and benefits of measures only, or to illustrate uncertainty, or to illustrate need for research on storativity. It could be used as a platform for discussion of the tradeoffs between different types of expenditures that could potentially fix the problem.

Discussion continued to a presentation at the 2008 Water Summit. It appeared that there was general consensus that for purposes of the summit, a comparison of the costs and benefits of various measures should be the focus. Uncertainty issues should be noted but not brought forward as a topic for discussion, but rather use illustration of two starting scenarios that bracket the realistic range of estimates. There was discussion of potentially having a focus group prior to the summit to see where issues/concerns/questions arise. Also potential presentation to entities that feed into PBAC and CAG (include greater complexities of uncertainty) in advance of the summit were discussed.

It was noted that the model could assist with identification of critical data gaps and a potential presentation to the legislature which illustrated the need for research funding.

With respect to a "bigger/larger model and process" there was concern expressed with the project having a life of its own separate from PBAC. Discussion continued surrounding the utility of developing a larger process. It was noted that this could help tie in the public aspect to PBAC and help educate and develop consensus in community. The issue of funding such a project was also a concern and that help from the states would be important. Trying to visualize "what" the process would look like was difficult but participants were encouraged to remember that this would be something that they would collaboratively develop over time.

It was decided that a list of FAQs would be developed to assist with graphing and interfaces that depict information and that the list would be available on this website.

Palouse Basin Participatory Model Workshop 4
May 30, 2008: 108 Law, University of Idaho

The workshop began with a review of changes and additions to the previous version of the model. Conservation measures described in both the WSU and Pullman water plans were added to the model as were projections of future demand. Moscow and UI are not required by the State of Idaho to create this type of water plan but will attempt to provide similar sorts of information for the next iteration of the model.

Multiple scenarios of recharge and storativity were provided. The hydrologists that were present agreed to meet again this summer to review current scientific knowledge and to discuss the potential for reducing the ranges of uncertainty.

Supply scenarios included simulations with the following mix of parameters:

• Recharge
• No Grande Ronde Recharge: does not calibrate with what we see
• .545 recharge – Grande Ronde [calibrates at .0005 storativity]
• Wanapum recharge based on water balance


• Storativity
• .001
• .0005
• .0001
• .00001
• The potential for using data from 50's – 60's on Wanapum to calibrate was discussed as were the limitations of this information and inherent uncertainties.


• Demand scenarios:
• Conservation Plans and Reclaimed water: WSU/Pullman
• No conservation/no reclaimed water
• Conservation/no reclaimed water
• Conservation and reclaimed water
• Cost of conservation measures and reclamation/savings added - first 6 years of Pullman/WSU plans
• Allows comparison of value – e.g. Pullman is currently getting more bang for the buck, however the costs will go up as the easy measures are completed
• $16 million is the projected cost to upgrade treatment at Pullman and pipe to WSU for plan for reclaimed water use

There were suggestions to have sliders reflect maximum realistic savings. This might impact behavior by showing the relation between reductions in demand and costs. Discussion ensued of how population is divided up – possible solution: just use total water use and total population. Current data indicates that this comes up with about 156 gpd per person.

The next modifications to the model should address:

• Further refining on use of conservation plan
• Lawn watering
• Brewing company [i.e. major water use industry]
• Add growth back in through back-calculation to Pullman demand projection
• Add more information on costs – try to tie to water rates

The meeting closed with another discussion on "Where to go with pilot and whether to extend to a larger project". Potential products of the pilot project could include an education tool to be used for public discussion and to illustrate the cost/benefit trade-offs of taking more or less conservative approaches now versus later. Discussion of longer term use of such a tool included taking an adaptive approach and whether, as a policy matter, it would be acceptable to start conservatively and a built in review process.

Participants were asked to start thinking about how to display information to the public both in the short term (water summit) and long term. The use of dollars per gallon was suggested as a unit. The concept of bringing all costs/benefits back to present value and comparing these to the cost of doing nothing was also suggested.

The presentation at the Water Summit will be called: Developing a Collaboratively Built Palouse Basin Decision Support Tool. By the next meeting the facilitator will develop preliminary outline of the summit presentation. The group will review the presentation and then help schedule time in the regular PBAC/CAG meetings for a pre-summit vetting.

Palouse Basin Participatory Model Workshop 5
August 22, 2008: Jack's Creek Conference Room, University of Idaho

The workshop opened with a welcome, introductions and quick review of last workshop. We then moved to a review of additions and changes to model that included improvement of the physical side, and added conservation for Pullman and Moscow. Some of Pullman’s short term costs for conservation were added but only went as far as their six year plan.

Participants were informed of meetings that were held with hydrologist and geologists during the summer. During and between these meetings the modelers were able to update the physical side of the model to the point that the hydro/geo experts were fairly comfortable with the model outputs.

Demonstrations of the current version of the model opened discussions that emphasized that the purpose of the model was to highlight trends rather than absolute numbers. One addition to the model puts a lag between pumping and aquifer response. Current model show about 45 years. But data may suggest about 15 years. Allyson can add any real data on this that exists. It was also noted that historic demand curves are currently “best guess”—these can also be adjusted with data.

Example model scenarios indicated that the addition of current growth rates seems to be addressable with conservation. However, there is a need to add costs to be able to compare this to supply measures. It is particularly important to have real estimates on costs since people tend to react to absolute numbers rather than trends and we do not want to alienate people with hypothetical numbers that could cause sticker shock.

The group then moved to a discussion about the Palouse Basin Water Summit presentation and what kinds of information should be highlighted. Participants made the following requests with respect to the draft presentation:

• Use elevation above sea level versus time in years to avoid the implication that we know the volume of water.
• Use currently known production zone from the WSU test well as a proxy for total volume of water when projecting how long we have to implement action.
• Concern was expressed that the lag masks what is going on since it is an arbitrary choice – on the other hand, make sure it is communicated that lag time in taking action and seeing a response is real even if uncertain.
• Overlay actual data from 1935 (could use knowledge of artesian wells to project back to early 1900's).
• Instead of using term “growth” use term such as no change – however, if using sliders, the model needed to breakdown increased use into population change, use change, etc.

It was agreed that modified slides would be emailed for review before the summit. In addition the revised presentation would be presented for a pre-summit vetting by both PBAC and the CAG.

The Palouse Basin Water Summit October 7, 2008

The Palouse Basin Participatory Model Pilot Project was presented during the later part of the morning to a full house (view presentation as pdf). Feedback from both participants and summit attendees was very positive. Summit attendees appreciated the emphasis on process and the fact that we had a wide cross section of participants that added to both credibility and transparency. They also seemed to understand that the process and the model had equal importance. One concern of the participants throughout the modeling exercise was that those not part of the process would take something from the model scenario outputs and use the information for further an agenda. This did not seem to be a problem; however the presenter did go to lengths to explain that all models had basic assumptions that should be openly explained and understood before using a model for decision support. In addition it was emphasized that questioning model outputs was an important part of not only building models, but also using them. Or in other words models are educational tools to help us make better decisions, not tools to make the decisions for us.