SC5.16

You have observed timeseries or observed fields from hydroclimatic variables (e.g., rainfall, wind, etc.) or from other environmental variables. You wish to generate synthetic ones that reproduce precisely the observed statistical properties, but you do not how to do it. No worries! Join us and you will find out!

The short course will introduce you to a unified method of stochastic modelling and the CoSMoS R-package that makes generation of random fields and of univariate or multivariate time series piece of cake. The generated random fields or time series preserve any desired probability distribution and correlation structure including features like spatial and temporal intermittency. We will talk about the stochastic properties of hydroclimatic processes such as precipitation, streamflow, wind, temperature, etc., and highlight features such as stationarity, cyclostationarity, marginal distributions, spatiotemporal correlations structures, and intermittency. We will explain how AR and multivariate AR models work and describe step-by-step the parent-Gaussian framework that allows precise and easy simulation of random fields and time series. Real-world examples include rainfall simulation at different spatiotemporal scales as well as simulating variables such as temperature, relative humidity, etc.

Early Career Scientists (ECS) and student are more than welcome! As always, we organize this short course in cooperation with the Young Hydrologic Society (YHS; younghs.com)!