Regionalisation of the “intensity-duration-frequency” curves in Northern Algeria

The flood events observed during last years in the urban areas are subject of main interest for quantification of the hydro-climatic risks and climatic change to the regional scales. The establishment of a statistical relationship between the intensities of intense rains and the recurrence of these events allows us to determine the dimensions of the works according to a previously defined level of risk. They constitute today a leading tool for various users. This work concerns the study of the maximum annual rains, recorded at 49 stations in the northern Algeria. The objectives of this work are to determine the estimators who are the “intensity-duration-frequency” curves and to extract from these whole of information the b Montana climatic parameter to be regionalized for the calculating the river flow and for the dimensioning of the networks of cleansing in the event of insufficiency of data. Different durations going from 15 min to 24 h are studied. We utilised the collocated co-kriging as multivariate estimation method for interpolation in order to yield the space distribution maps of b Montana climatic parameter, with the benefit of using spatially correlated secondary variables, such as the digital elevation model and the distance from the coastline that are known at any localisation. All features led to choose the digital elevation model as covariate for interpolating b Montana values, yielding a better regionalisation of the studied climatic parameter. The geostatistical handling of b Montana values strictly related to auxiliary variables that constitute physical factors overcomes the data shortage in planning, managing and preventing the rain flood risk.