Application of GIS to Bay-Delta Hydrodynamic and Salinity Modeling

Chuching Wang and Kevin Reagan

Abstract

An ArcView application has been developed that imports output files from a post-processing program of a Bay-Delta hydrodynamic and salinity model and creates production-ready layouts that graphically communicate the complicated model results.

Custom buttons are provided in an ArcView project that automatically import four model output files, create 52 different symbolized views based on key hydraulic/salinity variables posted to the link schematic, and generate 14 summary layouts that completely characterize the results and impact analyses associated with one case alternative model run. The process is easily repeated for numerous model runs, performing in minutes steps that conducted manually in ArcView had taken many hours or days of tedious repetition. The application utilizes “data-driven” programming concepts, where the program flow, view legend symbolization, and layout titling are controlled by the contents of ArcView tables that can be copied and edited for each specific model run, rather than having to modify “hard-coded”, case-specific Avenue scripts.

From each symbolized view, a custom button is provided to trigger 24 different types of tidally averaged daily time series displays associated with any chosen model link. The ability to rapidly generate graphics which visually portray conditions in every one of the 600+ model links, instead of the traditional practice of examining model results from only a limited set of (assumed) critical, “representative” model points, has greatly improved analysis efficiency and has led to important new insights into the behavior of the model and its applications.

Introduction/Defining Issues

California’s Bay-Delta estuary is an important habitat for many fish and wildlife species. It is also critical to California’s economy, supplying drinking water for 20 million Californians and irrigation water for 200 crops, including 45% of the nation’s production. The Bay-Delta ecosystem is in a declining state due to decades of competing or conflicting interests on water among various stakeholders. CALFED, a consortium of state and federal agencies, in collaboration with other stakeholders is currently in an effort to identify the long-term solutions to ensure reliable water supplies and a healthy ecosystem in the Bay-Delta.

Metropolitan Water District of Southern California (MWD), as one the primary stakeholders in the Bay-Delta water resources, is actively participating in the Agricultural/Urban Coalition to facilitate consensus among stakeholders and to develop preferred management options for the Bay-Delta long-term solution. A Delta Simulation Model (MWDDSM) has been developed and applied by MWD as part of the decision-making support tools in order to derive preferred alternatives and to evaluate the potential short- term and long-term hydrodynamic and salinity impact due to potential CALFED solution alternatives. The model is a modification of the California Department of Water Resources’ Delta Simulation Model, Suisun Marsh Version (DWR, 1995) which in turn is a modified version of the Fischer Delta Model (Fischer, 1984).

The simulation model is a one-dimensional numerical model, solving tidally influenced unsteady-state open-channel flow and solute-transport equations with a link and node configuration representing the complex network of waterways in the Bay-Delta. Output of the model consists of time-series information on stage, flow, velocity, and salinity (total dissolved solids) at every model link and node with time interval in terms of minutes.

Operation of the model is resource intensive. It requires long CPU hours to run. It requires large quantities of input data. It also produces more than 800 MB worth of information (time series of various parameters for every link and node of the model) for every single year real tides are run. How to digest the output information in a timely and efficient manner, and how to present the information in a compact and effective way so they can be incorporated in the decision-making process becomes a challenging task.

Hydrodynamic/Salinity Impact Analyses

For a hydrodynamic/salinity impact analysis among alternatives, we first establish a baseline condition. MWDDSM is then used to simulate the stage, flow, velocity, and salinity behaviors over a given time period for both baseline and case alternative conditions. The impact due an alternative can be defined as the differences of the above mentioned properties between case alternative and the baseline.

For the tidally influenced Bay-Delta estuary, the properties such as flow, stage and salinity are varied contiunously all the time due to the tidal influence. For a typical short term impact analysis (over one year time frame), we use one-minute time step to solve the flow equation, and use 3-minute time step to solve the solute transport equation. If we sample the numerical solution of a property (for example, flow) at 15-minute interval, then we would have a time series with 31,000 data points for any given model node. It is not easy to graphically present the results, especially the comparison of two alternatives.

In our study, we propose to present and evaluate results based on tidally averaged daily properties, which are averages over a whole daily tidal cycle. A whole daily tidal cycle or a tidal day is approximately 25 hours. Coupling the tidally averaged daily mean with daily maiximum, daily minimum and daily range, we can characterize the whole time series of a property and tide amplitude in a much more comprehensible way.

ArcView Application Design

In this paper, a GIS approach is presented to deal with the massive information in conducting impact analyses. An ArcView application has been developed that imports output files from a post processing program of the hydrodynamic and salinity model and creates production- ready layouts that graphically communicate the complicated model results. Graphical visualization of summarized time-series results obtained from the model, by symbolizing a schematic link representation of the model grid, has not been possible until now.

Custom buttons are provided in an ArcView project by applying Avenue that automatically import four model output files, create 52 different symbolized views (Table 1) based on key hydraulic/salinity variables posted to the link schematic, and generate 14 summary layouts (Table 2) that completely characterize the results and impact analyses associated with one case alternative model run. The application utilizes “data-driven” programming concepts, where the program flow, view legend symbolization, and layout titling are controlled by the contents of ArcView tables that can be copied and edited for each specific model run, rather than having to modify “hard- coded”, case-specific Avenue scripts. The process is easily repeated for numerous model runs, performing in minutes steps that conducted manually in ArcView had taken many hours or days of tedious repetition.

Additionally, from each symbolized view, if more details are desired, a custom button is provided so that one can click on a model link and trigger a custom designed program which can display 24 different types of tidally averaged daily time series associated with the chosen model link.

For our typical application of the developed tool, once the simulation model completed, we ported the results into the tool and quickly generated hard copies of the 14 layouts. From the hard copies, we studied the impact, checked for any counter-intuitive solution bahaviors, and noted the significant impacted areas. We then went to the tool and opened views of interest, and applied the time series display tool to further investigate the details of those links of interest. From this we can quickly check the integrity of input assumptions and interpretate and characterize the model results in a very effective and efficient manner.

A sample set of layouts of a case scenario is included in the Appendix for reference. Also included in the Appendix are images to illustrate the time series display tool.

Conclusion

Acknowledgments

References

Appendix: Sample Layouts