GC23D-0662:
Stochastic Hourly Weather Generator HOWGH: Validation and its Use in Pest Modelling under Present and Future Climates

Tuesday, 16 December 2014
Martin Dubrovsky1, Martin Hirschi2,3 and Christoph Spirig3, (1)Institute of Atmospheric Physics ACSR, Praha 4, Czech Republic, (2)ETH, Institute for Atmospheric and Climate Science, Zurich, Switzerland, (3)MeteoSwiss Federal Office of Meteorology and Climatology, Zurich, Switzerland
Abstract:
To quantify impact of the climate change on a specific pest (or any weather-dependent process) in a specific site, we may use a site-calibrated pest (or other) model and compare its outputs obtained with site‑specific weather data representing present vs. perturbed climates. The input weather data may be produced by the stochastic weather generator. Apart from the quality of the pest model, the reliability of the results obtained in such experiment depend on an ability of the generator to represent the statistical structure of the real world weather series, and on the sensitivity of the pest model to possible imperfections of the generator.

This contribution deals with the multivariate HOWGH weather generator, which is based on a combination of parametric and non‑parametric statistical methods. Here, HOWGH is used to generate synthetic hourly series of three weather variables (solar radiation, temperature and precipitation) required by a dynamic pest model SOPRA to simulate the development of codling moth.

The contribution presents results of the direct and indirect validation of HOWGH. In the direct validation, the synthetic series generated by HOWGH (various settings of its underlying model are assumed) are validated in terms of multiple climatic characteristics, focusing on the subdaily wet/dry and hot/cold spells. In the indirect validation, we assess the generator in terms of characteristics derived from the outputs of SOPRA model fed by the observed vs. synthetic series.

The weather generator may be used to produce weather series representing present and future climates. In the latter case, the parameters of the generator may be modified by the climate change scenarios based on Global or Regional Climate Models. To demonstrate this feature, the results of codling moth simulations for future climate will be shown.

Acknowledgements: The weather generator is developed and validated within the frame of projects WG4VALUE (project LD12029 sponsored by the Ministry of Education, Youth and Sports of CR), and VALUE (COST ES 1102 action).