Dienstag, 04. Dezember 2018, 12:15 - 13:30 iCal

Vortrag von Dr. Christoph Frei

Uncertainty in spatial precipitation analyses, and why climatology should care about it.

Hörsaal 1 (Raum 2A120)
Althanstr. 14, UZA 2, 1090 Wien


Vortrag im Rahmen des Meteorologisch-Geophysikalischen Kolloquiums:

Today, most national weather services offer analyses of the spatial distribution of precipitation as an operational climate service. Commonly the analyses are provided on a regular grid, at km-scale spacing, and cover the past 30-plus years at daily resolution. Such data products are widely used for evaluating climate models, for driving runoff models, for estimating extremes, and for deriving regional climate-change scenarios. The underlying rain-gauge networks, however, have limited spatial coverage and the measurements are subject to error. This translates into uncertainties of spatial analyses that may be relevant in applications. This presentation introduces a probabilistic methodology that spawns an ensemble of possible distributions of precipitation rather than a single “best” estimate. Results of the ensemble will be presented for the Alpine region. They provide insight into the magnitude and dependencies of analysis uncertainty. Moreover, they pinpoint to a dangerous pitfall when using traditional grid datasets for the analysis of extremes, for the inference of long-term trends and, likely, for several other applications.


Institut für Meteorologie und Geophysik


Mag. Herta Gassner
Institut für Meteorologie und Geophysik