National Academies Press: OpenBook
« Previous: Model Calibration and Testing
Suggested Citation:"Model Coupling." National Academies of Sciences, Engineering, and Medicine. 2016. Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/23557.
×

17. Calibration and validation should consider efforts to compare model output of rates as well as state variables.

It is a common practice to use state variables such as biomass to validate numerical models of primary producers. However, especially in this SAV model where three critical rate processes are simulated (photosynthesis, respiration, growth), it is important to look at output from the model of these rate processes and to compare, in some fashion, these simulated rates to measured rates. If measured rates are not available for all species, their acquisition can be identified as a critical research activity to be done while the literature is scanned to provide some confidence regarding rates for some of the modeled species.

18. Developing a SAV model that can accurately simulate the observed maps of SAV coverage is unlikely and not advisable. Rather, validation exercises should be considered that take SAV coverage into account at larger spatial scales and compare patterns of SAV coverage between predicted and observed maps.

A “pattern-oriented approach” similar to that described by Grimm et al. (2005) could be considered for guiding model evaluation and validation. This approach also influences model development, but in a way that is complementary to the currently described efforts for the SAV model. This might include validation exercises comparing important patterns generated from the model that were not simply an outcome of the model inputs.

In addition to considering a pattern-oriented strategy, with a spatial scale of 1 m2 and the decision to permit just one species per grid cell, it is highly unlikely that a simulated map of SAV coverage will directly mirror actual species distribution maps. A more realistic validation exercise might consider comparisons at the reach scale or some intermediate spatial scale above 1 m2. Model evaluation should focus on aggregate measures (e.g., total biomass by type) and their seasonal and spatial patterns, rather than trying to match predicted and observed biomasses on a cell-by-cell basis within a survey. Validation should also consider the use of the SAV model as both a standalone model and in its role as generating habitat input for the FD model, to ensure that the appropriate aggregate measures are evaluated.

Model Coupling

Model coupling is a special topic because of the plans by the model development team to use the results of the SAV modeling to provide habitat inputs to the FD model. Running models in a coupled mode involves additional issues beyond those identified above, which were based on running the two models independently (stand-alone).

There are four submodels within the overall ecological model: hydraulics (steady state 2D model), water quality (QUAL2E), SAV, and FD. The hydraulics model is Dr. Thom Hardy’s existing MD_SWMS model for both the Comal and San Marcos systems. The grid size is 0.25 m2. Hydrology (flow and depth) is generated by having 7-day averages over the time period 2000 to 2013. The water quality model for both systems is QUAL2E. The model outputs from the hydraulics and water quality submodels that have been used to date in the FD model are maximum daily water temperature and minimum daily DO. The SAV model will require hydraulic and water quality model outputs of depth, temperature, and (eventually) some measure

Suggested Citation:"Model Coupling." National Academies of Sciences, Engineering, and Medicine. 2016. Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/23557.
×

(e.g., average daily) of flow. There is currently no direct role for velocity or depth as inputs to the FD submodel.

The SAV submodel is a stand-alone model that can be used to examine questions related to HCP activities, and it is also planned to provide the habitat information for the FD model grid. While these are highly related uses of the SAV modeling, it is likely that compromises are needed in order for the same SAV model to be able to perform both uses. For example, trying to use the SAV modeling results as input to the FD model may push the SAV modeling to a finer spatial scale to match the FD model than if the only goal of the SAV modeling was to assess flow effects on SAV dynamics.

The actual coupling between the SAV and FD models is planned on being one-way, which is reasonable. That is, SAV affects FD, but FD does not affect SAV. This is biologically realistic and also allows for the SAV and FD models to be run separately if needed for computational reasons. The SAV modeling should generate outputs on spatial and temporal scales realistic for how FD uses these habitats within the model; that is, how do these habitats affect FD growth, mortality, reproduction, and movement on hourly to daily time steps for roughly 1-m2 spatial resolution and within the FD model domain of a reach. For example, conversion of a grid cell to a different SAV species occurs just one time per month at this stage of model coupling. Careful consideration of whether this, as well as how other variables are transferred, is sufficiently accurate for use in the FD model is warranted. The SAV modeling is still unsettled but it seems that a reasonable compromise can be found such that the SAV modeling can be used for both simulating SAV responses to flow and for providing habitat inputs to the FD modeling.

The use of steady state hydraulics and dynamic QUAL2E as potential inputs to the SAV and FD model is reasonable provided the limitations of this coupling (hydraulics-QUAL2E) for use in the FD model are clearly detailed. The hydraulics model is used in two ways: a series of constant flows is simulated (steady-state for each flow) for direct use of depths and velocities in SAV and FD models, and using 7-day average values of flow (also to steady-state) as input to the QUAL2E model to generate hourly temperature and DO. The steady-state velocities and depths are re-gridded from the 0.25 m2 of the hydraulics to the 1-m2 grid of the FD model. The hourly temperature and DO are processed to obtain daily maximum temperature and daily minimum DO values. All of the FD model cells fall within a single QUAL2E segment, and thus the temperature and DO values in the appropriate QUAL2E segment are applied to all of the cells in the FD model. While the idea of model coupling is sound and seems simple and intuitive, the details are very important for conveying the limitations (and strengths like higher confidence) when the fully coupled set of models are used to simulate SAV and FD responses to HCP actions.

At every step of passing output from one model to be input to the next model in the chain, some information is lost (often variance) and the receiving model inherits the assumptions used to run the donor model. In the situation here, these steps include aggregation of 0.25 m2 scale in the hydraulics to 1-m2 resolution in the SAV and FD models, steady state hydraulics used dynamically in the SAV and FD models, steady state hydraulics used differently to generate velocities and depths versus as input to QUAL2E to generate temperature and DO, and all of the FD and SAV model cells being within a QUAL2E segment (i.e., no spatial variability). Careful evaluation and bookkeeping of the assumptions, of how information is generated (e.g., steady-state versus dynamic) and then passed to the next model, is needed to ensure the information from different sources is consistent and to know what types of scenarios can be realistically

Suggested Citation:"Model Coupling." National Academies of Sciences, Engineering, and Medicine. 2016. Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/23557.
×
Page 18
Suggested Citation:"Model Coupling." National Academies of Sciences, Engineering, and Medicine. 2016. Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/23557.
×
Page 19
Next: Concluding Remarks »
Evaluation of the Predictive Ecological Model for the Edwards Aquifer Habitat Conservation Plan: An Interim Report as Part of Phase 2 Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

An ad hoc committee of the National Academies of Sciences, Engineering, and Medicine is in the process of reviewing the many different scientific initiatives underway to support the Edwards Aquifer Habitat Conservation Plan (HCP). The Committee to Review the Edwards Aquifer Habitat Conservation Plan is focusing on the adequacy of information to reliably inform assessments of the HCP's scientific initiatives, ensuring that these initiatives are based on the best-available science. Relationships among proposed conservation measures (including flow protection measures and habitat protection and restoration), biological objectives (defined by the HCP as specified flow rates), and biological goals (such as maintaining populations of the endangered species) are central to the HCP, and are being evaluated during the Academies review. The study spans from 2014 to 2018 and will result in three reports. At the conclusion of Phase 1, the Committee issued its first report (NRC, 2015), which focused on hydrologic modeling, ecological modeling, water quality and biological monitoring, and the Applied Research Program. The Committee will issue its second report in late 2016 and its third and final report in 2018.

This interim report is part of Phase 2 activities and will be incorporated, as an appendix, into the second report. This interim report focuses on the ecological modeling only and is being provided prior to the issuance of the second report in order for the Committee's comments (which take the form of conclusions and recommendations) to be considered while the ecological modeling team is still in place.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!