Integrated Modeling Planning: Five Year Science Plan for the Tampa Bay Study

Home | Tampa Bay Study | Reports | Five Year Science Plan for the Tampa Bay Study

Appendix 3: Integrated Model Planning

Five Year Science Plan for the Tampa Bay Study
April 15, 2003 - Kimberly Yates

Appendix 3: Integrated Model Planning:

There have, historically, been numerous modeling efforts in Tampa Bay that will be helpful in laying the foundation for an integrated model. For example, in 1992, the Tampa Bay Estuary Program held a Modeling Workshop to examine how best to go forward toward water quality, nutrient and seagrass goals. 50 scientists and managers, both local and national participated. They recommended a three-pronged approach:

Develop an empirical model to relate nitrogen load to water quality
Develop a mechanistic model for the same
then, and only if the first two approaches did not provide the framework and technical basis needed for development of loading targets, develop to an integrated water quality, circulation, loading and response model approach.

The first two modeling exercises are complete. The Tampa Bay Estuary Program and its partners believe that the first two models have and will continue to provide the level of certainty we need for maintenance of the Tampa Bay Nitrogen Management Strategy. Thus, they are not looking toward further model development for that management issue. However, the TBEP is interested in application of a more integrated approach for issues such as sediment transport (which may be important for seagrass restoration efforts), water quality, and potential salinity changes with changes in freshwater inflow that may result from bay-wide dredging activities and other natural and anthropogenic alterations in the Bay.

The Coastal/Navigation Section, Planning Division, Jacksonville District, U.S. Army Corps of Engineers is conducting Tampa Harbor Project General Re-evaluation Study to examine the entire 70 miles of Federal channels in Tampa Bay. This study will determine if modifications are warranted, with a particular focus on anchorage areas. The non-Federal sponsor for the Tampa Harbor Project is the Tampa Port Authority. The St. Petersburg Harbor Project study is also intended to determine if modifications of St. Petersburg Federal channels are warranted. The non-Federal sponsor for the St. Petersburg Harbor Project is the City of St. Petersburg.

The Corps has a need for modeling of Tampa Bay, in conjunction with the general Re-evaluation studies. In particular, the Corps is in need of two types of modeling for these studies. The first type is hydrodynamic modeling, to provide input for the ship simulation study. The second type is impact modeling, to address impacts to the environment from changes in channel geometry from dredging and port expansion activites. Impacts in which the U.S.A.C.E. is interested in addressing include impacts to seagrasses, circulation changes, alteration of the salinity regime, and sedimentation.

Both the Tampa Bay Estuary Program and its partners, and the U.S. Army Corps of Engineers have agreed to partner with the U.S. Geological Survey to plan and implement the development of an integrated model that will address circulation/hydrodynamics, sediment transport, water quality, submerged aquatic vegetation, and groundwater to facilitate improved dredging and dredged material management – a key TBEP CCMP goal.

On Septermber 18, 2002, we held our first planning workshop, co-hosted by U.S.G.S., the T.B.E.P., and the U.S.A.C.E.-Jacksonville District (see attached meeting summary) for integrated modeling. Sixty-one participants from 24 different federal, state and local entities participated. The primary purpose of this meeting was to become more familiar with the modeling efforts already existing in Tampa Bay that may help provide the foundation to move forward with an integrated model in Tampa Bay, and to generate discussion on the following topics:

Applications for an integrated model – Tampa Harbor Study?
What components should be addressed?
Do we have some pieces already underway that can be incorporated?
What are some of the challenges we may face?

The Tampa Harbor Study provides a timely and significant opportunity for a focused model application. The changes the U.S.A.C.E. will be making over the next several years will impact the entire Tampa Bay System including circulation, water quality, sediment transport, seagrass, etc. One possibility for an integrated model application is to address the Tampa Harbor Study with a feedback monitoring mode model. For example, we can model the impact of specific engineering/constructional phases of the Tampa Harbor Study, and monitor, test, and validate the model as the phases of the Tampa Harbor Study are carried out, making modifications as we proceed. The timing of the Harbor Study is consistent with the U.S.G.S. project time frame and this would provide us with a good example of an applied integrated model. It seems that many past modeling efforts may have “gone awry” because they were not focused enough. By focusing on a particular application or issue, this will provide the impetus to create a foundational integrated model that can later be modified to address the numerous questions that various agencies would like to have answered. We anticipate that a foundational integrated model will later be modified to address the many needs of other agencies, and will ultimately be coupled with a watershed model in production by the Southwest Florida Water Management District for complete ecosystem modeling.

As a result of the predictive modeling workshop, we have recognized that all of the agencies participating have may individual needs from an integrated model, and are willing to collaborate on its development and implementation. We have also recognized that successful and efficient model development will be best achieved by choosing a well-defined focus for the model application with a very specific set of questions to be answered by the modeling effort. After the foundational model is developed and tested, the model may, then, be altered to address the many questions of partnering agencies. There are many modeling efforts already in existence for various components of estuarine function (circulation, seagrass, etc.). We will encourage the continued development of these component models. Ultimately, we hope to compare results from integrated model activities to other model results.

During FiscalYear 2001 and 2002, several meetings were held with federal agencies and state universities that are currently engaging in integrated model development activities to determine the currently existing “state of the art” with respect to integrated modeling. It is clear that most models that integrate circulation, water quality, sediment transport, groundwater, and submersed aquatic vegetation components are very costly (requiring a minimum of $1M for development), require many years to develop (5 or more), and have not, yet, been successfully developed for direct end-use by non-modeller users through easily accessible web-based systems. There have been many examples of large modeling efforts in estuaries across the nation that have been perceived as failed attempts. These “failed attempts” have had in common the following: 1) they have tried to answer too many questions (lack of focus), 2) the application was not well-defined prior to development, 3) were not accepted by resource managers, 4) no long-term ownership and maintenance, 4) not user-friendly (inaccessible and long-turnaround time for answers).

We envision the development of an integrated model by a consortium led by the U.S.G.S., and consisting of other federal, state, and local partners including TBEP, USACE, FMRI, USF, SWFWMD, and Tampa Bay Water, and the City of Tampa. Through careful planning, we will engage the resource management community as partners in the development of modeling capabilities to foster long-term investment and interest in the development, use, and maintenance of a model. We will engage the expertise of the numerous scientists who have participated in previous modeling efforts in Tampa Bay. Throughout the planning process we have consulted with our (U.S.G.S.) nationally recognized modeling experts (including Chris Sherwood, Harry Jenter, Jared Bales, John Warner) to ensure that Tampa Bay modeling efforts complement the modeling goals of the U.S.G.S., nationally, and to ensure technology transfer of new modeling approaches that may be developed for use in Tampa Bay and other U.S. coastal environments.

A primary measure of success for development of an integrated model will be the degree to which it supports and is used by the scientific and resource management community, and the interest and support of regional stakeholders in long-term maintenance and use of the model. This model will target three audience components through a web-accessible user interface at http://gulfsci.gov.

First, the model will be used to provide information and results for the very complex questions of scientists, engineers, and modelers, realizing that many model “manipulations” are very complex, require modification of algorithms, and the degree of accessibility for model manipulation is much greater than for many resource managers and the general public. Use of the model in this capacity will have restricted accessibility to prevent misuse of the model. This level of accessibility will likely be limited to the modeling consortium and scientists directly contributing to data input. GIS products resulting from model scenarios will be accessible through the Interactive Map Server or Digital Library.

Second, a web-accessible user interface will be created that enables resource managers to manipulate prescribed scenarios by choosing from a variety of parameter options to predict future outcomes with respect to the various model components. This will enable resource managers to easily navigate through the model, test scenarios on the own, develop confidence in the results, and use these test cases for resource management decisions. Accessibility at this level will be limited to prescribed scenarios that are updated with “real-time” data and information (e.g. tides, water temperature, water quality, winds, etc.).

Third, a web-accessible user interface will be developed for general public use with the same level of accessibility based on prescribed scenarios and “real-time” data. The focus of this component will target public interest issues, for example, water quality information pertinent to the fishing community, beach goers, etc. Allowing the general public to access and use the same model that is used for very complex resource management decisions will develop confidence and public support in the mechanisms used to make resource management decisions.

To date, we have identified only one modeling group that has the technology to create a fully integrated model with the level of accessibility for different user groups as described above. The Danish Hydrologic Institute based in Denmark with a regional office in St. Petersburg, Florida has developed a generation of modeling software that provides the architecture within which to develop highly complex integrated models with a web-accessible user interface for various levels of difficulty described above. This modeling software has proven successful for integrated modeling related to large-scale dredging activities at Oresund in the Baltic Sea (Jensen and Lyngby 1999) and other locations. The Baltic see modeling effort targeted circulation, sediment transport, water quality, and seagrass components, and is similar to the integrated modeling effort planned for Tampa Bay. While the software is proprietary and is available with all of the algorithms required to develop a fully integrated model, it is designed for ease of replacing pre-programmed algorithms so that open-source code algorithms can be used. We liken the DHI software to spreadsheet software such as MS Excel. While Excel is proprietary software, it can be used to program mathematical models, the equations and results of which are not proprietary, are owned by the software user, and publishable as such.

DHI software has been tested and widely used by the international modeling community (Universities, Federal and state entities) for a wide variety of coastal issues ranging from coastal erosion, submerged habitat impact, water quality, and hydrodynamics (Bach et al. 1998, Rasmussen et al. 2000). DHI has the flexibility to provide software licensing to a consortium of agencies understanding that they are working toward a common goal. We feel that use of this software will bring U.S.G.S. modeling to the forefront of integrated modeling technology. Integrated model development using DHI software will require a fraction of the cost (approximately $300,000 as opposed to $1M) and a fraction of the time (approximately 1 to 1.5 years as opposed to 5) that U.S. modeling groups have required. A $300,000.00 budget is required for purchase of software ($96,000), hiring of a term or FTE, Ph.D. level modeler to lead Tampa Bay modeling efforts, 6 months to 1 year of intensive training for the modeler by DHI representatives. As a best-case scenario, using the DHI software may enable us to generate a fully integrated model within 1 to 1.5 years that is user-friendly and web-accessible. The remaining years of the project can then be used to calibrate, validate, and modify the model for answering a variety of client questions. As a worst-case scenario, if our attempts to use DHI software are not satisfactory, we have only spent 1.5 years and $300,000 to find that it doesn’t work rather than 5 years and $1M. At that point we still have the option and time to pursue other integrated modeling efforts. Thus, we feel that pursuing use of the DHI software is the most cost effective and efficient mechanism to pursue integrated modeling for Tampa Bay.

The U.S.G.S. will take the lead in organizing a modeling consortium, development, calibration, validation, and implementation of an integrated model. The objectives for model development and questions that it will address will be defined in conjunction with our primary partners (the Tampa Bay Estuary Program and U.S. Army Corps of Engineers) to ensure a very focused goal for efficiency of model development. The modeling consortium consisting of representatives from regional agencies participating in Tampa Bay modeling efforts and resource managers will provide technical expertise, input data, and other advice for customising the model product for Tampa Bay and its end-users.

Development, validation, and testing of an integrated model is one of the major products that we anticipate will result from a five-year project in Tampa Bay. Thus, we will focus many (but not all) of the data collection and research activities on providing the information necessary to build, validate and calibrate an integrated model, and to link the estuary model with a watershed model under development by our partners. Many of the monitoring activities required for validation and calibration of an integrated model are very costly. Thus, as research objectives are completed, resources from those objectives will be redirected to the purchase, maintenance, and utilization of data for monitoring instrumentation in the bay.

Scientists working on the Tampa Bay project are knowledgeable on the general types of information required for integrated model development. However, we lack the expertise required to develop detailed science plans with timelines and milestones that accurately reflect our predictive modeling goals. The timeline in Table 1 reflects our best attempt to define a schedule without having the predictive modeling component in place. This schedule will be modified as predictive modeling plans are further developed. In order to effectively develop future science plans, we need, immediately, expert advice from a predictive modeller familiar with the goals of Tampa Bay integrated model plans, and with the information required to develop all components of an integrated model. We would like to immediately hire in either a term or permanent position a Ph.D. modeller to provide this advice, participate in further development of science plans, to lead integrated model development for Tampa Bay, and to participate in technology transfer activities for U.S.G.S. modeling in other coastal environments.