An agricultural flash flood loss estimation methodology: the case study of the Koiliaris basin (Greece), February 2003 flood

Abstract

River flooding causes significant losses to crops and negatively affects local agriculture economies, particularly in rural riverine areas. In this work, a techno-economic methodology for the monetary estimation of crop losses due to flash flooding is presented. The methodology takes into account flood depth and flow velocity, as provided by MIKE FLOOD, as well as the season of flood occurrence, and provides monetary estimates of crop damage based on synthetic logistic flow velocity–flood depth–crop damage surfaces. The development of the flood damage surfaces involved a questionnaire survey targeting practicing and research agronomists. Subsequently, a weighted Monte Carlo simulation was performed in order to enhance the questionnaire-based loss estimate information. Finally, synthetic flow velocity–flood depth–crop damage surfaces were developed for every crop under study and for every month using logistic regression analysis. The damage surfaces are an essential component of the developed model which was implemented in Python, enabling the GIS visualization of the estimated agricultural damage. The aforementioned methodology was applied for estimating the damage caused by a flash flood that took place in the Koiliaris River Basin in Crete for which no historical data exist. The novelty of the proposed methodology is the development of local synthetic flow velocity–flood depth–crop damage surfaces. Furthermore, the velocity parameter, which is taken into account, makes the methodology suitable for flash flood events, where significant discharges and high velocities dominate, or for flood event cases which are characterized by high flow velocities. The methodology identifies rural areas and agricultural land uses that are most prone to flooding and serious crop damages and thus require greater attention. Furthermore, the methodology aptitude for developing local damage surfaces could be modulated in order to confront different flood scenarios on various crops distributions and be used to address agricultural planning activities.