Employment of Compound Events in Developing and Validating the Stochastic Spatial Weather Generator

0518229 - ÚFA 2020 RIV US eng A - Abstrakt
Dubrovský, Martin - Lhotka, Ondřej - Mikšovský, J. - Salis, M. - Štěpánek, P. - Skalák, P. - Meitner, J.
Employment of Compound Events in Developing and Validating the Stochastic Spatial Weather Generator.
AGU Fall Meeting. Washington: American Geophysical Union, 2019. GC43G-1321.
[AGU Fall Meeting 2019. 09.12.2019-13.12.2019, San Francisco]
Grant CEP: GA ČR(CZ) GA18-15958S
Institucionální podpora: RVO:68378289
Klíčová slova: circulation weather generator * stochastic weather generator (WG) * GRIMASA project
Obor OECD: Climatic research
https://agu.confex.com/agu/fm19/meetingapp.cgi/Paper/580946

GRIMASA project aims to develop a spatial (not only, but especially a gridded version) stochastic weather generator (WG) applicable at various spatial and temporal scales, for both present and future climates. The first version of the multi-purpose SPAGETTA generator, which is based on a parametric approach suggested by Wilks (1998, 2009), was presented already at AGU-2017 (contribution NG41A-0115). The present version is run at daily time step and allows to produce multivariate weather series for up to 100 (approximately) grid-points. In developing and validating the generator, we employ also various compound weather indices defined by multiple weather variables, which allows to include inter-variable correlations into the validation process. In our previous experiments, the WG was run at 100 km resolution (50 km version of EOBS data was used for calibrating the WG) for eight European regions, and its performance was compared with RCMs (CORDEX simulations for EUR-44 domain). At AGU-2018 (GC33J-1492), our WG was validated in terms of characteristics of spatial temperature-precipitation compound spells (including dry-hot spells). Most recently, after implementing wind speed and humidity into the generator, the WG was run at much finer resolution (using data from irregularly distributed weather stations in Czechia and Sardinia) and validated in terms of spatial spells of wildfire-prone weather (using Fire Weather Index).
Present project activities aim mainly at (A) going into finer spatial and temporal scales, and (B) conditioning the surface weather generator on larger scale circulation simulated by circulation weather generator run at coarser resolution: the development of the circulation generator has started this year. In our contribution, we will show results related to these two activities, focusing on the WG’s ability to reproduce spatial temperature-precipitation spells and fire-weather spells at various spatial scales (down to EUR-11 resolution) for selected European regions. Effect of using the circulation generator to drive the surface weather generator will be also assessed.
Trvalý link: http://hdl.handle.net/11104/0303389