A quality assessment framework for natural hazard event documentation: application to trans-basin flood reports in Germany

Abstract

Written sources that aim at documenting and analysing a particular natural hazard event in the recent past are published at vast majority as grey literature (e.g. as technical reports) and therefore outside of the scholarly publication routes. In consequence, the application of event-specific documentation in natural hazard research has been constrained by barriers in accessibility, concerns of credibility towards these sources and by limited awareness of their content and its usefulness for research questions. In this study we address the concerns of credibility for the first time and present a quality assessment framework for written sources from a user's perspective, i.e. we assess the documents' fitness for use to enhance the understanding of trans-basin floods in Germany in the period 1952–2002. The framework is designed to be generally applicable for any natural hazard event documentation and assesses the quality of a document, addressing accessibility as well as representational, contextual, and intrinsic dimensions of quality. We introduce an ordinal scaling scheme to grade the quality in the individual quality dimensions and the Pedigree score which serves as a measure for the overall document quality. We present results of an application of the framework to a set of 133 cases of event-specific documentation relevant for understanding trans-basin floods in Germany. Our results show that the majority of flood event-specific reports are of good quality, i.e. they are well enough drafted, largely accurate and objective, and contain a substantial amount of information on the sources, pathways and receptors/consequences of the floods. The validation of our results against assessments of two independent peers confirms the objectivity and transparency of the quality assessment framework. Using an example flood event that occurred in October/November 1998 we demonstrate how the information from multiple reports can be synthesised.