Date of Event: 05/11/2021
Start Time: 9:00 am
THE UNIVERSITY OF CONNECTICUT
Civil & Environmental Engineering
DEPARTMENT OF CIVIL & ENVIRONMENTAL ENGINEERING
UNIVERSITY OF CONNECTICUT
9:00 AM – TUESDAY, MAY 11TH, 2021
Emmanouil N. Anagnostou (Major Advisor)
Efthymios I. Nikolopoulos (Associate Advisor)
Marina Astitha (Associate Advisor)
Xinyi Shen (Associate Advisor)
Francesco Marra (Associate Advisor)
Using Remote Sensing Observations and Model Simulations for the Analysis of Hydrological Extremes.
Hydrological extremes can harm society and ecosystems. However, many parts of the world lack in situ observations for quantifying hydrological extremes. Physically-based distributed hydrological model simulations driven by atmospheric simulations and remote sensing precipitation observations can be used to alleviate the issue of data scarcity in estimating return periods of hydrological extremes, but the short data record length associated with these datasets limits the application of traditional statistical methods (GEV/LP3/GPD) that rely on extreme value theory. Also, the errors in these indirect measurements or model simulations may lead to large biases in the quantification of extremes. The novel Metastatistical Extreme Value Distribution (MEVD) framework is proposed in this research as a mean of overcoming the limitations imposed by the short record length and obtaining more reliable assessment of high quantiles. The error estimates of MEVD applied on the data generated from satellite-based precipitation products and hydrological model simulations are thoroughly evaluated across different regions and hydroclimatic conditions. It is shown that MEVD is able to address the fundamental issue of data record limitations in deriving robust estimation of hydrological extremes, and alleviate the biases in hydrological model simulations of flood peaks. The application of the MEVD framework in conjunction with simulated streamflows and high-resolution precipitation products from remote sensing observations bring new opportunities for estimating hydrological extremes at global scale, including areas with limited or no in situ records.
Published: May 6, 2021