Pscheidt, Ieda: Generating high resolution precipitation conditional on rainfall observations and satellite data. - Bonn, 2017. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5n-48088
@phdthesis{handle:20.500.11811/7240,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5n-48088,
author = {{Ieda Pscheidt}},
title = {Generating high resolution precipitation conditional on rainfall observations and satellite data},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2017,
month = sep,

volume = 79,
note = {This study is part of the high resolution reanalysis project proposed for Germany and Europe (Bollmeyer et. al., 2014) within the framework of the Hans Ertel Centre for Weather Research (HErZ). The reanalysis for Germany assimilates among other variables high resolution rainfall rates. For the most recent years, radar data is assimilated, however, for periods before 2007 this data is not available and another radar-like dataset is required. This study proposes the method HIRAIN to generate an ensemble of probable space-time precipitation fields given a set of observational data. HIRAIN works in two steps. First, a Bayesian statistical model conditional on observations from synoptic stations and on satellite information simulates the latent spatial Gaussian process that drives the occurrence of precipitation exceeding a selected threshold. In a second step, realisations of occurrence/non-occurrence of precipitation exceeding the same thresholds are obtained given the simulated latent process. The occurrence/non-occurrence of precipitation is generated through two different methodologies. HIRAIN is extended to several thresholds of precipitation amount and the final precipitation product is generated from the fields occurrence/non-occurrence of the individual thresholds. A Bayesian approach is used in HIRAIN to provide more realistic fields than those produced by interpolation methods. In the Bayesian approach the data at the observation locations are honored and the spatial covariance structure of the spatial process is reproduced in each realisation. Moreover, the ability to generate ensemble of possible precipitation patterns provides valuable information of precipitation uncertainties that plays also an important role in ensemble reanalysis. HIRAIN produces precipitation dataset with hourly and 4 km resolution. This product presents a more appropriate resolution for the purposes of the reanalysis than the rainfall datasets available by the time the Germany reanalysis project started.},
url = {https://hdl.handle.net/20.500.11811/7240}
}

Die folgenden Nutzungsbestimmungen sind mit dieser Ressource verbunden:

InCopyright