M. Isabel Pedroso De Lima, IMAR - Institute of Marine Research; Polytechnic Institute of Coimbra, Department of Forest Resources / ESAC, iplima@esac.pt
Probable impacts of extreme precipitation on society and the environment are strongly determined by regional specificities and are expected to be affected by changes in the climate. Thus, the study of local precipitation regimes is essential in this context. However, many models have difficulties in adequately describing the large variability in precipitation, which is still a task that is in need of further insight. This work reports analyses of rainfall extremes across time scales, which is conducted in a multifractal framework that handles large rainfall dynamic ranges, brings together the behavior of rainfall observed at different scales and overcomes the problem of using different models to describe data of different resolutions. Moreover, this framework implicitly assumes the presence of heavy tails in the probability distribution of rainfall intensity. Results suggest that some ‘conventional’ approaches might underestimate often high return levels of rainfall. The data analyzed are from Portugal.
Keywords: rainfall; extremes; multifractals.