ENSO impact on simulated South American hydro-climatology

摘要

The variability of the simulated hydro-climatology of the WaterGAP Global HydrologyModel (WGHM) is analysed. Main object of this study is the ENSO-driven variabilityof the water storageof South America. The horizontal model resolution amounts to 0.5 degreeand it is forced with monthly climate variables for 1961-1995 of the Tyndall CentreClimate Research Unit dataset (CRU TS 2.0) as a representation of the observed climatestate. Secondly, the model is also forced by the model output of a global circulationmodel, the ECHAM4-T42 GCM. This model itself is driven by observed monthlymeans of the global Sea Surface Temperatures (SST) and the sea ice coverage for theperiod of 1903 to 1994 (GISST). Thus, the climate model and the hydrological model representa realistic simulated realisation of the hydro-climatologic state of the last century.Since four simulations of the ECHAM4 model with the same forcing, but with differentinitial conditions are carried out, an analysis of variance (ANOVA) gives an impressionof the impact of the varying SST on the hydro-climatology, because the variancecan be separated into a SST-explained and a model internal variability (noise). Also regionalmultivariate analyses, like Empirical Orthogonal Functions (EOF) and CanonicalCorrelation Analysis (CCA) provide information of the complex time-space variability.In particular the Amazon region and the South of Brazil are significantly influencedby the ENSO-variability, but also the Pacific coastal areas of Ecuador and Peru are affected. Additionally, different ENSO-indices, based on SSTanomalies (e.g. NINO3.4, NINO1+2), and its influence on the South Americanhydro-climatology are analysed. Especially, the Pacific coast regions of Ecuador, Peruand Chile show a very different behaviour dependant on those indices.