IMPACT Project field tests data analysis
Hassan, M.A.A.M.; Morris, M.W. (2008). IMPACT Project field tests data analysis. FLOODsite Report, T04-08-04. [S.n.]: [s.l.]. 30 pp.
Part of: FLOODsite Report, more
|  |
| Authors | | Top |
- Hassan, M.A.A.M.
- Morris, M.W.
|
|
|
| Abstract |
Part of the FLOODsite Task 4 research programme included a review and more detailed analysis of the IMPACT Project field and laboratory test data relating to breach growth and breach modelling (see www.impact-project.net work package 2). As part of FLOODsite Task 6, HR Wallingford established links with the CEATI facilitated Dam Safety Interest Group project on breach modelling. Through this project, copies of data arising from the IMPACT project, but provided via the original IMPACT Project Norwegian partner, and a number of intermediary organisations, was received by HR Wallingford and noted to significantly differ from the original master data sets held by HR Wallingford (as coordinator of the IMPACT Project, and work package leader for the breach work). This report presents the conclusions of a long term investigation into the source of the differences in the IMPACT data sets, and concludes recommended values for future use of the IMPACT data. The investigation concluded that:• There are significant differences between the IMPACT data now being used, as compared to the data used during breach model testing under the IMPACT project. The extent to which this affects research conclusions from the IMPACT project is unclear. Researchers using data originating from the IMPACT project itself should refer to this report and associated data sets to ensure they are using correct data• It appears that much of the data provided from Norway during the IMPACT project (i.e. the field test data) related to proposed rather than as built test conditions. Significant differences between proposed and as built conditions which were not reported at the time have now been identified. However, whilst these explain a significant number of the data differences, they do not explain all differences, including some of the major differences.• Some flow data has been recalculated. This includes reservoir releases providing inflow to the test site as well as measured breach flow• Some revised calculations provided from Norway still appear to be incorrect, hence some questions still remain as to actual data valuesThese findings are disappointing, both for the IMPACT partners who were assured of data quality at the time of the original research and more recently for a growing number of researchers around the world who have been using data from the original project reports. Whilst the original data was not formally released pending final analyses at the end of the project, it is clear that data has been shared and a number of researchers have published work using the original data. It is recommended that these researchers review their analyses in light of the findings reported here.This report endeavours to identify errors and offers a ‘best estimate’ of correct data sets that may subsequently be used in breach model verification and validation. Copies of the IMPACT project breach data are also now being made publicly available alongside this report. |
|
