INterdisciplinary CLimate INvestigation cEnter

Portuguese English Spanish

Past and future evolution of rainfall extremes

Themes and Scientific Questions

Society´s vulnerability to extreme events has been repeatedly mentioned in questions related to climate change, in particular in the IPCC reports. (Intergovernamental Panel of Climate Change, 2007). The IPCC analysis points towards a future climate of the Earth with global average surface temperature a few degrees Celsius above present values due to the enhancement of the greenhouse effect and that, in these conditions, the meteorological events that are considered extremes today, would be more frequent.

An analysis of the future, few decades or centuries ahead, has to follow a detailed knowledge of the recent past, also in the scale of decades – in the sense of detecting tendencies or changes that might have happened or are happening know. This has been the case for the IPCC analysis based in climate numerical simulations for different scenarios of socioeconomical evolution, energy solution and public attitude with respect to conservation of natural resources. However, as discussed by Randall et al (2007), the changes observed in the past in the extreme temperatures are reasonably well represented in numerical simulations while there is large uncertainty in the results that refer to changes in rainfall. This has an implication on the future climate simulations where the different simulation used have sometimes contradictory results with respect to changes in rainfall a rainfall extremes. The IPCC (2007) report indicates large areas in Brazil were less than 66% of the models disagree with respect of the change in rainfall for a season, a century from now.

The objective of this component is to define and describe the daily rainfall extremes that happened in the past and to associate to these the vulnerability of the population of large urban areas in Brazil. Identify daily rainfall series tendencies in the past and explore their evolution in scenarios of climate change in the future.


Vulnerability is understood as the capacity that people, their activities and infrastructure to undergo loss and damage when subject to physical events of different orders of magnitude. This, an extreme event has the potential of causing an impact depending on the system vulnerability.

Specifically in the case of rainfall there are several possibilities of events that may considered extremes. However, the same extreme rainfall event may cause larger or less impact depending on the vulnerability of the population and of the infrastructure. This is an analysis that has to be regional with a methodology that may be essentially global.

Two categories will be addressed: (1) intense and rapid rainfall associated to thunderstorms with their associated damaging winds, electrical discharges and flash floods; (2) rainfall events lasting several days associated to floods and landslides. The impacts observed in each of these categories are localized and depend on topography, land cover, soil drainage as well as on the maintenance of public areas with respect to their drainage capacity. The identification of an extreme event can also be made from the meteorological point of view or from the socioeconomical perspective or a combination of these two.

The methodology employed is based on the extraction of time series of damages and losses associated to rainfall and the definition of meteorological indicators of their occurrence that may be obtained from reanalysis and climate simulations that have good performance in tropical Brazil (e..g. Quadro et al 2010).

a. Data

  • Loss and damges associated with rainfall events
  • Reanalysis of past and present
  • Climate simulations of the past and future

b. Method

  • Use of indicators of potencial occurrence of thuderstorms and heavy rainfall events verified in the past and present, and extension into the future in the form of time series.
  • Temporal tendencies: signal analysis in non-linear and nonstationary time series (e.g. Wu et al 2007)

Goals and expected products

  • Identification of tendencies in disasters associated with extreme rainfall events in the past in São Paulo and 5 other Brazilian metropolitan areas.
  • Identification of tendencies in rainfall extremes in the past and in the future for São Paulo and 5 other metropolitan regions in Brazil.
  • Students: 4 IC, 2 Ms, 2 Drs.
  • Two short courses on extreme precipitation events
  • 2 Workshops on rainfall extremes in urban áreas, impacts and vulnerabilities.


Quadro, M., M. A. F. Silva Dias, D. L. Herdies, L. G. De Gonçalves, E. H. Berbery, 2010 Evaluation of the hydrological cycle over South America through the new generation reanalyses. Submitted Geophysical Research Letters. Randall, D.A., et al., 2007: Climate Models and their Evaluation. In: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [S. 49 Solomon, et al. (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA Wu,Z. N. E. Huang, S. R. Long, and C.-K. Peng, 2007: On the trend, detrending, and variability of nonlinear and nonstationary time series, PNAS, 104 (38) 14889-14894;doi:10.1073/pnas.0701020104