SPICOSA´s central objective of developing the Systems Approach Framework (SAF) centres on methodologies for simulating the function of coastal zone systems. The simulation software EXTEND applied at SPICOSA Study Sites makes it possible to show how complex Coastal Zone systems react to a wide range of changes in the use and management of these systems and, in return, how changes in the natural systems influence economic and social sectors. Currently, the SPICOSA Study Sites are kicking off the formulation step using EXTEND for making models.

Demonstration video on how to use EXTEND

The Model Support Team of SPICOSA regularly produces a Special EXTEND Newsletter for the EXTEND community of SPICOSA and beyond with information about developments of EXTEND applications, examples, and progress.

EXTEND Software
We have pre-selected EXTEND™ for the simulation software because it best matches our needs. This will not be a constraint for future users of the SAF, which will not be dependent on any single software nor on any individual researcher. A part from being inexpensive the Extend software has a number of characteristics critical to our needs, a few are listed here (see description):
User interface. Extend models are very user friendly, portable and can be cloned, which allows researchers with relatively little or no modelling experience to read, write, and operate Extend models. This is essential to our objectives that researchers, not software experts, construct the basic model components.
Hierarchical. Very simple representations can be easily expanded or coupled. This allows researchers to email components to each other for use or critic and facilitates the use of a shared model library. In addition, the user support web site provides access to additional model blocks.
Programmable. The Extend modeller (familiar with ”C” language) can access the built-in, compiled language, “ModL”, to modify or change any operating block. All models and modifications belong to the user.
Tools. The basic software comes with a set of tools for graphics, database system for input and output, data analysis, built-in optimization schemes, animation routines, an internal notebook text and cloned model blocks, hierarchical blocks (specialized clusters of smaller blocks), and connectivity such that inter application communication is a drag and drop operation.
Information. The software home page is http://imaginethatinc.com and further articles are available. An example of an environmental application is given in the following figure. Please also visit the website section Publications for access to Extend training materials and model examples. For more information please contact Jean-Luc De Kok, (Model Support, VITO)

Buy. 1Point2 is exclusive distributor of ExtendSim v7 in France, Belgium, Italy, Spain, Portugal, and Greece. www.1point2.com


SPICOSA definition of system simulation
The words ‘system’ and ‘simulation’ have broad definitions. For our purposes, we define a system as some meaningful collection of processes that demonstrate a distinct function, and a simulation as representation of a system that can quantify a system’s function to some reliable degree of accuracy. Of particular importance is our definition of a coastal zone system involves all interactive components, i.e. both the natural ecosystems and the human social-economic systems. Recognizing that an acceptable level of sustainability requires that these interactions are mutually constructive and stabilizing; consequently, our simulations must include the cultural, institutional, and economic process of these interactions.

In other words, the goal of the SAF is to develop the capacity to simulate, in an operational manner, changes in status and in value of coastal-zone systems induced by human influences and vice versa. In our first attempt, we will focus on providing decision-makers with the best-possible information on ‘what-if scenarios” for the use in implementing the Sustainable Development in coastal zones. For example, these scenarios will simulate the costs and benefits mutually occurring between the costs of sustaining the total value of natural systems and monetary benefit of their increasing use. Both natural and anthropogenic systems are strongly time dependent, inherently so, because their inputs (e.g. sunlight, meteorology, socio-economic needs) are all strongly time-dependent and often importantly spatially dependent. For decision-making processes, the first-order simulations will have to do with changes in cost or public disapproval as a function of time; and among the important spatial questions will be the distributions of these costs over the policy domain. These simulations and their interpretations will be converted to a language and format convenient for end-users and packaged in a way that will facilitate future users through its self-evolving and accumulative mechanisms.


Systems Approach Framework
Representing a system’s function in a simple, prognostic manner requires the best possible understanding of the system. The SAF is based on the Systems Theory, e.g. that we cannot understand a system by researching its disassembled parts or by ignoring its external interactions, and on the emerging concepts governing complex systems, e.g. that systems self-organize under the influence of their external constraints and by the quality of their internal interactions. The simplifying tricks of the Systems Approach concern:

  • Designing simulations with the best possible knowledge of the system, its inputs, its internal interactions and its constraints to change,
  • Initiating first-order processes such that their results can be calibrated,
  • Iterating to higher-order processes, specifying the degree of resolution required,
  • Specifying how the spatial dimension will be represented, e.g. as in virtual space or more detailed as with a spatial numerical grid,
  • Approximating the unknown inputs and functions through statistical, empirical, etc. means, which can be later replaced with more accurate data

These tricks will serve us to the degree that we want an integrated response and that we can accept some degree of error. These concessions are offset by the possibility that we can continue refine the exercise to the limit of our understanding and information base. With modern advances in commercially available software, we are much less limited by computation than were the researchers of the sixties when system simulations were first attempted. Likewise, the availability today of large supportive data sets and greater knowledge allows us to tackle problems that would have been impossible a generation ago.


Lagoon Salinity. A simulation of the salinity in the Pamlico Sound (North Carolina, USA) was run for the years 1998 through 2000 with the only input being the observed Atlantic salinities, and the local meteorology. The series was run without adjustments after its initial calibration for mixing and friction in the inlet. Blue line is the simulated salinity values, the red dots are observed salinities, and he green line the observed runoff. The large runoff peak was due to the flooding caused by Hurricane Floyd. Full agreement would not be expected because of approximations in the input data. Close agreement over the 4-yr interval indicates that the dynamics are correctly formulated (otherwise the solution would not be stable). This model is constructed using the non-linear thermohaline method for estuarine exchange, see Hopkins, T.S., 2001. Thermohaline feedback loops and natural capital. Scientia Marina, 65 (Suppl.2): 233-258