Simulation and prediction of hydrometeorological processes based solely on dynamical equations has proven to be unrealistic, a fact mainly reflected by the inability of dynamical models to reproduce the huge variability of these processes, particularly at a point and at a very fine time scale. Coupling stochastic approaches with deterministic hydrometeorological predictions, in order to better represent predictive uncertainty, remains nevertheless a non-trivial matter. As expected, when the processes are aggregated at coarser spatial scales, up to the global scale, or at coarser time scales up to climatic scales of decades to centuries, the variability is reduced. Yet such decrease of variability is not as drastic as predicted by classical statistical approaches. While to date sufficient evidence has accumulated on the departure of statistical properties of hydrometeorological processes from the classical statistical prototype, in most cases the latter is still used, due to scientists' familiarity, in such notions as detection and diagnosis of hydroclimatic trends and estimation of future hydroclimatic variability. This session aims to highlight the use of stochastic approaches in the representation of hydrometeorological processes and to seek the approaches that are most consistent with the hydrometeorological reality.
The session is organized by the EGU/HS Subdivision on Precipitation and Climate and is cosponsored by the International Association of Hydrological Sciences/Working Group on Statistical Hydrology (IAHS/STAHY).