- Identification: GA205/05/2265 - Grant Agency of the Czech Republic
- Duration: 2005 - 2007
- Principal investigator: RNDr. Martin Dubrovský Dr. (Institute of Atmospheric Physics
ASCR) - Co-investigator: Ing. Martin Růžička CSc.
- Co-investigator: RNDr. Ladislav Metelka Dr. (Czech Hydrometeorological Institute)
- Co-investigator: Ing. Mgr. Miroslav Trnka PhD. (Institute of Landscape Ecology ASCR)
Subject of research
The project focuses on calibrating stochastic single-site daily weather generator (WG) for sites with non-existent or incomplete historical daily weather series (which are normally used to determine WG parameters). The present WG is designed to provide input data for crop models and hydrological models. Aims of the project: 1) lf only some daily variables are observed, the missing WG parameters (e.g., parameters for the daily temperature range) will be estimated from those available using empirical relationships established from the learning data. 2) If no weather observations are available in the given site, the WG parameters will be interpolated from surrounding stations. Various interpotation techniques (with a stress on the neural networks) will be tested. In interpolating the parameters, elevation of the site will be taken into account. 3) Accurracy of estimating WG parameters from a globál 0.5°x0.5° climatologicat data available from Climate Research Unit (U.K.) will be assessed. 4) Two
Research Goal
(aim 0) For use in the present project, we collected a database of 6?variate station daily weather data for Czechia (125 stations x 40 years) and Nebraska [28 stations x (20-25 years)]. The data were used to calibrate the stochastic daily weather generat.
