Conference on Stochastic Weather Generators

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

The Conference on Stochastic Weather Generators (SWGEN 2018) will be held in Boulder, Colorado on October 2-4, 2018. The conference will be held at the National Center for Atmospheric Research. Stochastic weather generators (SWGs) are mathematical or statistical algorithms whose simulated values capture the statistical distribution of weather or climate variables of interest. Weather and climate-related phenomena can have substantial impacts on human safety, for example flooding, heat waves and severe storms. They also have significant societal and economic impacts: seasonal water resource planning, crop yield studies and weather volatility risk assessments all require high resolution simulations of relevant weather quantities such as temperature, precipitation, wind speed and solar insolation. Historically, SWGs were developed by domain scientists for applied projects such as statistical downscaling of climate models. Recently the applied mathematics and statistical science communities have taken increasing interest in their formal development, which feed back into novel modeling, estimation and simulation techniques that then warrant theoretical study. SWGEN 2018 serves as the first conference devoted to SWGs in the United States, and is intended to introduce a wide audience of mathematical, statistical and domain scientists to the current state of the field, disseminate novel techniques for precipitation, temperature, wind speed and solar radiation modeling, to forge new collaborative opportunities and to identify areas of future research. The conference will consist of keynote lectures, a poster session and research talks featuring topics of major interest in mathematical and statistical communities for weather simulation including 1) modeling techniques for renewable energy applications such as wind and solar irradiance simulation, 2) modeling and simulation of complex nonstationary and inhomogeneous processes that evolve over space and time, and 3) large-scale spatiotemporally-coherent simulation methods. Generating realistic spatiotemporal weather processes requires novel techniques from areas such as spatial and computational statistics. These methods dovetail with current interests in the statistical community involving big data modeling, estimation and simulation techniques. Other topics will be considered, including nonlinear processes, high frequency data, multi-scale models, extremes, analog or resampling techniques, and non-Gaussian processes with applications in hydrology, agriculture, air quality, insurance and environmental engineering. Links between topics will be identified and discussed to better understand the role of mathematics and statistics in this field. The conference website is at: https://www2.cisl.ucar.edu/calendar/swgen-2018-stochastic-weather-generators-conference This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.