Rainfall intensity-duration-frequency curves at ungauged locations with uncertainties due to climate change

• Main Author: Muhammad Noor, Muhammad Noor
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.utm.my/id/eprint/97944/1/MuhammadNoorPSKA2020.pdf; http://eprints.utm.my/id/eprint/97944/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:144895
• Institution: Universiti Teknologi Malaysia

Intensity duration frequency (IDF) curves are important in designing and managing urban hydraulic structures for mitigation of floods. The objective of the study was to develop IDF curves at ungauged locations with associated uncertainties due to climate change. Peninsular Malaysia was considered as the case study area. The novelty of the study was to propose a new methodology for reliable estimation of IDF curves at any location with consideration of non-stationary behaviour of rainfall due to climate change which can be used for robust designing of climate change resilient urban hydraulic structures. Hourly observed rainfall data at 80 locations distributed over Peninsular Malaysia and four remote-sensing rainfall datasets namely, GSMaP_NRT, GSMaP_GC, PERSIANN and TRMM_3B42V7 were used for this purpose. Four widely used probability distribution functions (PDFs) and four methods for estimation of PDF parameters were compared to determine the most suitable PDF and its parameter estimation method in the study area. Subsequently, the estimated parameters of the selected PDF were used to generate IDF curves at all the observed locations. The performance of four remote sensing rainfall datasets in construction of IDF curves at observed locations was compared to find the best product. The bias in the IDF curve of the best rainfall product was corrected to generate the IDF curves at ungauged locations. To update the IDF curves for future climate change scenarios, high-resolution rainfall projections data were generated through selection of suitable global climate models (GCMs) of Coupled Model Intercomparison Project Phase 5 (CMIP5) and their downscaling at remote sensing rainfall grid locations. Climate change factor at each grid location was estimated through comparison of PDF of historical and future simulations of GCMs for different radiative concentration pathways (RCP) scenarios. The factors were used to perturb the historical IDF curves to generate IDF curves with associated uncertainties for future climate change scenarios. Results revealed general extreme value (GEV) as the best-fitted PDF and maximum likelihood as the best parameter estimation method at 62% of the stations. Performance assessment of remote sensing rainfall datasets revealed all datasets underestimated rainfall intensities for different durations and return periods. Comparative performance of the products revealed GSMaP_GC as the most suitable product for developing IDF curves at ungauged locations with least biases (8% to 27%). BCC-CSM1.1 (M), CCSM4, CSIRO-Mk3.6.0 and HadGEM2-ES were found as the most suitable GCMs models for the projection of daily rainfall in Peninsular Malaysia. The ensemble mean of projected rainfall showed a maximum increase in annual rainfall by 15.72% and an increase in variability by 26.15% during 2070-2099 compared to the base period (1971-2000) under RCP 8.5. The assessment of IDF curves with uncertainty revealed a maximum change in rainfall intensity for different durations under RCP 8.5 and the minimum for RCP 2.6. The rainfall intensity for different durations was found to increase with time. The highest increase was observed up to 96.8% for the period 2070-2099. The assessment of uncertainty in rainfall IDF for different RCP scenarios revealed higher uncertainty for higher return periods and vice versa. The IDF curves generated in this study can suitably be used for designing hydraulic structures at locations where observed rainfall data is not available. It can also be used for designing hydraulic structure for adaptation to climate change induced rainfall extremes and mitigation of urban flood.