Author(s): Marisa Rosana Silva Davila

Linked Author(s): Marisa Rosana Silva Dávila

Keywords: Uncertainty on PMF; PMP; Rainfall duration

Abstract: In the analysis of maximum floods in basins without and/or with few discharge measurements, many sources of uncertainty make it necessary to question not only the data but also the results obtained. The objective of this research is to reduce the uncertainty in the determination of the probable maximum flood through an exhaustive analysis of the basic information and the understanding of the hydrological and meteorological processes of the basin to determine, among others, the validity of the existing records, the representativeness of the regional information, besides of the rainfall depth and duration, temporal pattern, and spatial distribution of the PMP. The study was applied to the Mantaro River basin (Atlantic Ocean basin) located in the central region of Peru. For this, the information quality study of all available hydrometric and rainfall information was carried out through a rigorous review of the data and some values had to be discarded to conclude that the series obtained are valid for the study. There is one stream flow gauging station near the outlet of the basin, with an approximate area of 27 000 km2, with only measurements of daily flows, whose record shows very high values, and which also had to be corrected due to a diversion upstream of the streamflow gauging station. Flood hydrographs were determined with the statistical method for different return periods. To calculate the PMF, the precipitation-discharge model was performed with HEC-HMS using the Probable Maximum Precipitation (PMP). The information of pluviographic stations is very scarce, so there is no information about the temporal pattern of the storms including duration and the onset of these in the entire extension of the basin. Regional information had to be used to determine the intensity-duration-frequency (IDF) curves. The determination of the PMP in the basin with Hershfield's Method resulted in values that were too large. For the calibration process of the precipitation-discharge model, it was necessary to conceptualize the characteristics of the heterogeneous behavior observed in the basin considering its great extension, its vegetation cover characteristic, geological conditions, flood routing, and the temporal pattern, occurrence, and synchronization of storms throughout the basin. This made it necessary to include in the modeling a lag in the onset of the storm between the northern portion and the southern portion of the basin.

DOI: https://doi.org/10.3850/978-90-833476-1-5_iahr40wc-p0662-cd

Year: 2023