Non-Stationary Rainfall Intensity-Duration-Frequency Curves

  • Latifa A. Yousef

Student thesis: Master's Thesis


Efficient management of water resources is of crucial importance, especially in arid regions such as the United Arab Emirates (UAE). Climate change has induced variability in rainfall around the world. Current research advances recognize the existence of non-stationarity in hydro-meteorological variables and the importance of incorporating it in the modeling of these variables. Despite the wide-spread use of Intensity-Duration-Frequency (IDF) curves, developments in their methodology have not happened as extensively as other hydro-meteorological models. The methodology developed in this thesis incorporates non-stationarity existing in rainfall records into the general IDF relationship, by making the parameters involved dependent on covariates of time and climate oscillations. The best suited probability distribution functions are case-specifically chosen and applied in the general IDF models. The parameters of the models are found by optimizing the least square error of the fitting of the models to the data. Non-stationarity in rainfall intensity records is tested for using the revised Mann-Kendall test. Correlations between the rainfall intensities and climate indices are used to determine which climate oscillations to test as covariates. The methodology is tested on stations located in Abu Dhabi, Quebec and Ontario in Canada and the southwestern United States. IDF relationships for all stations are developed using both the regular stationary methodology and the proposed non-stationary methodology, in order to compare between the results of both methods. The non-stationary framework for IDF modeling exhibits better results for almost all tested stations when fitting the models to the data. Improvements in terms of goodness-of-fit measured by the least square error of the models reach approximately 60% for some stations. The results prove that the proposed methodology has the potential to be significant in the formulation and revision of IDF relationships where non-stationarity is found in rainfall records.
Date of Award2014
Original languageAmerican English
SupervisorTaha B. M. J. Ouarda (Supervisor)


  • Water Resources
  • UAE
  • Climate Change
  • Intensity-Duration-Frequency
  • hydro-meteorological models
  • Abu Dhabi.

Cite this