Post-Doc position in my group at the Department of Civil, Environmental and Architectural Engineering, University of Padova (Italy), focusing on stochastic hydrology. Opening late October. Likely start in December 2023/January 2024. 18 Months duration with possible extension (pending funding). 25 kEuro/year. Contact me if interested: email@example.com .
This project aims to generate estimates of local hourly and daily extremes for the whole of Italy to be made available as maps published online for high quantiles of typical practical interest (e.g. 10, 100, and 200 years). The recently-developed Metastatistical Extreme Value Distribution (Marani and Ignaccolo, 2016) and the associated downscaling methodology (Zorzetto and Marani, 2019) will be used and applied to datasets including: Climate Hazards Center InfraRed Precipitation withStation data (CHIRPS version 2, daily, 0.05°, 1981-2021; Funk & al 2015), Integrated Multi-satellitE Retrievals for GPM (IMERG Final precipitation version 6, 30 min, 0.1°, 2000-2021; Huffman & al 2019), Climate PredictionCenter Morphing Technique (CMORPH Climate Data Record version1.0, 30 min, ~0.08°,1998-2020; Xie & al 2019), SM2RAIN-CCI (daily, 0.25°, 1998-2015;Ciabatta & al 2018), and the I2-RED database (doi:10.5281/zenodo.4893122).
The novel MEVD, to be used in this project activities, assumes a Weibull distribution of “ordinary” rainfall values, from which yearly maxima are “drawn”. The Weibull distributions fitted to satellite rainfall retrievals will be downscaled from the daily scale to the hourly scale in time, and from the satellite spatial scale (1-25 km) to the point scale in space. The theory of stochastic fields will be used to link statistical moments of a rainfall field averaged over areas and time intervals of different sizes once its space-time correlation is defined. Matching downscaled Weibull parameters from satellite rainfall estimates to those from hourly rain gauge observations will define correction factors for all satellite pixels for which rain gauges are available. This will allow the adjustment of the satellite estimates of extremes. These approaches have been successfully used to downscale rainfall in time (Marani & Zanetti 2007) and space (Zorzetto & Marani 2019). The extension of this approach to a simultaneous space-time context is a specific objective of the proposed work. This will lead to estimates of local hourly and daily extremes for the whole of Italy through regionalization and interpolation of the correction factors derived. Maps of extremes will be published for high quantiles, together with maps at the original satellite scale (i.e. no downscaling) produced through traditional GEV-based methods for reference.