Joint probability analysis predicts the probability of occurrence of events in which two or more partially dependent variables simultaneously take high or extreme values. The objectives of this project are to:

• involve and consult the wider industry including relevant TAG leaders and framework consultants on their joint probability requirements,

• bring together recent joint probability work at HR Wallingford, CEH Wallingford and the Proudman Laboratory,

• extend it where necessary to the whole of England, Scotland and Wales,

• map dependence around and within the UK for several variable pairs relevant to flood and coastal defence,

• develop best practice guidelines for use of joint probability methods and results,

• draw up proposals for an open workshop and/or training seminars to explain methods and their appropriate use,

• assess research needs for development and take up of joint probability methods and results.

Several different environmental variables are potentially important in design and assessment of flood and coastal defences, including waves, tides, surges, river flows, rainfall, swell and wind. Joint probability analysis methods have been developed, tested and applied in consultancy studies by the researchers involved, and benefits demonstrated, but take up within the industry has been patchy. The aim is to increase appropriate take up of joint probability methods in flood and coastal defence design and assessment.

• Dependence maps and tables will provide guidance on realistic levels of dependence between different variables, or whether independence would be a reasonable assumption, addressing the problem of lack of appropriate data for use in joint probability work.

• Best practice guidance will provide clear and relevant notes on when, where and how to apply joint probability methods and results, addressing the issue of reluctance to use methods at present limited to specialists.

• The assessment of industry needs will increase the chance of appropriate take up of methods and results. The assessment of research needs will identify gaps needing to be filled or updated by future research.

• DEFRA/EA sponsored open workshops will not only seek to publicise and disseminate methods, but will also clarify and demonstrate support for the use of such methods.

During the first few months of the project, an outline of the best practice guide was prepared and circulated for consultation within the industry. This consultation included an open meeting at Wallingford on 30 May 2002.

Dependence between key pairs of variables will be summarised around England, Wales and Scotland in a form suitable for use in simplified joint exceedence analysis methods such as that given in the CIRIA beach manual. Confidence in the dependence estimates will be indicated in a way that can be used in design calculations. The variable pairs involved are:

• Wave heights & water levels; wave heights and surges

• Tides & surges

• Rainfall & river flow; surge & river flow; rainfall and surge

• Wind sea & swell

The guidelines will summarise best practice based on the experience of the project team and steering group. They will pass on clear and relevant advice about how, when and why joint probability analyses and results should be used in project work. They will cover use of simplified methods, the dependence maps, JOIN SEA, extreme water level predictions and rainfall/surge/flow results. The guidelines will cover data requirements, types of variable and applications amenable to joint probability analysis, methods of checking results, how to incorporate assumptions about climate change, benefits and potential pitfalls.