Personal tools
You are here: Home Catalogue of methods Reframe
Related terms
Navigation
What's up ?
Be notified when a document is published in this folder or below.
 
Views

Hamish Harvey, Jim Hall

The Reframe software, being developed through the FRMRC and FLOODsite projects, is is designed to provide support for end-to-end flood risk analysis. A major part of this is to provide support for the explicit handling of uncertainty in the flood risk management decision-making process.

Reframe is built around a new metamodel, a formal account of the various artefacts involved in computational decision support for flood risk management. This metamodel is designed to enable the definition of flood risk calculations in a way which is sufficiently complete to be executable by computer, and sufficiently succinct to be used as the basis for human communication about the content of an analysis.

The current Reframe prototype has two main parts, each based on this metamodel.

  1. A compiler for a programming language which allows computations to be defined in terms of the Reframe metamodel.

  2. A web-browser based viewer for computations defined in this language, offering textual and diagrammatic representations.

The focus of development (as of April 2007) is on handling non-native data formats. This in turn will allow the encapsulation and reuse of external codes, as using these unmodified depends on reliable data format translation. A case study is under way based around an Environment Agency/ HR Wallingford Thames risk model.

Compiler

The long term goal is for users to be able to import data sets and to define computations over these through a web--based user interface, and then to enact these computations without leaving this interface. The Reframe system will then enact the overall analysis defined, coordinating the potentially many underlying computational codes required.

The compiler has so far been developed to the stage that it can compile definitions of idealised risk analysis calculations. This stage provides an important proof of concept. Further development will take place on the basis of need in the case study mentioned above.

It is possible that the Reframe tools may be used by some in a purely descriptive mode, to aid communication about computional analysis within the decision-making process. In this case it is still important to know that the Reframe metamodel is sufficient to describe these analyses. The most effective way of building this confidence is proof by example.

User interface

The web-browser based viewer will be developed into a fully interactive interface to the system. The rich metamodel behind Reframe opens up a range of possibilities for user-supporting features. It supports the formal encoding of much more of the inherent structure of data sets and computations than current techniques. This will enable

  1. A hypertext-based user interface.

  2. Extensive consistency checking of user specified computations (eventually with real-time feedback as changes are made).

  3. The automatic generation of diagrammatic representations of computations.

The generation of diagrams already exists in prototype. Diagrams are a powerful tool for communication, a critical aspect of flood risk management as a collaborative process, but most diagrammatic representations of computations (such as flow diagrams) are informal in nature and must be developed manually quite separately from the code which they purport to represent. In contrast, diagrams generated by the Reframe system can be presented with perfect confidence that they correspond exactly to the computation.




subject:
 

Powered by Plone CMS, the Open Source Content Management System

This site conforms to the following standards: