H13H-1661
Space-time Precipitation Variability of Global Scale Precipitation: Empirical Analysis, Stochastic Modeling, Anthropogenic Effects

Monday, 14 December 2015
Poster Hall (Moscone South)
Isabel deLima, University of Coimbra, Coimbra, Portugal; Institute of Marine Research (IMAR) and Marine and Environmental Sciences Centre (MARE), Coimbra, Portugal and Shaun Lovejoy, McGill University, Montreal, QC, Canada
Abstract:
The characterization of different precipitation scaling regimes represents a key contribution to the improved understanding of space-time precipitation variability. We analyze three global scale precipitation products (one instrument based, one reanalysis based, one satellite and gauge based) from monthly to centennial scales and planetary down to several hundred kilometers in spatial scale. Results from space-time scaling analyses of spectra and Haar fluctuations in precipitation show the presence – similarly to other atmospheric fields - of an intermediate “macroweather” regime between the familiar weather and climate regimes. These regimes qualitatively and quantitatively alternate in the way fluctuations vary with scale. In the macroweather regime, precipitation is characterized by negative temporal fluctuation exponents, which implies – contrary to the weather regime – that fluctuations tend to cancel each other out. This regime is important for seasonal, annual and decadal forecasts, and it is also important for assessing lower frequency anthropogenic effects that can be detected because (in the industrial epoch) they break the scaling.

We characterize systematically the macroweather precipitation temporal and spatial, and joint space-time statistics and variability, and the outer scale limit of temporal scaling (the critical scale is about 20 - 40 yrs, depending on the product, on the spatial scale). Our results clarify the different scaling regimes and further allow us to quantify the agreement (and lack of agreement) of the precipitation products as a function of space and time scales. We show how to detect the anthropogenic signal in precipitation above the natural variability noise. We also show how to make new space-time precipitation models that can be used for monthly to interannual forecasting. These clarifications are relevant for a full understanding of the changes affecting the hydrological cycle and the interactions between different hydrologic processes.