Combining fractal analysis and time series mining to identify regional climate extremes

• Santiago Augusto Nunes USP
• Luciana A. S. Romani Embrapa
• Ana M. H. Avila UNICAMP
• Elaine P. M. de Sousa USP
• Agma J. M. Traina USP

Abstract

In the last few decades, huge amounts of climate data from ground-based stations and other different sensors have been gathered and stored by several private and public institutions. The analysis of these data has become an important task due to worldwide climate changes and the consequent social and economic effects. In this work, we propose an approach to analyze multiple climate time series in order to identify intrinsic temporal patterns. By dealing with multiple time series as multidimensional data streams, we can integrate different climate variables and discover general behavior changes over time. Experimental studies on real climate time series collected from different regions of Brazil show the applicability of our approach.

References

Alexander, L., Zhang, X., Peterson, T., Caesar, J., Gleason, B., Tank, A., Haylock, M., Collins, D., Trewin, B., Rahimzadech, F., Tagipour, A., Kumar, K. R., Revadekar, J., Griffiths, G., Vincent, L., Stephenson, D., Burn, J., Aguilar, E., Brunet, M. Taylor, M., New, M., Zhai, P., Rusticucci, M., and Vasquez-Aguirre, J. (2006). Global observed changes in daily climate extremes of temperature and precipitation. Journal of Geophysical Research, 111:1–22.

Baioco, G. B., Traina, A. J. M., and Traina, C. (2007). Mamcost: Global and local estimates leading to robust cost estimation of similarity queries. In Proceedings of SSDBM, pages 6–16, Banff, Canada.

Barbará, D. and Chen, P. (2000). Using the fractal dimension to cluster datasets. In Proceedings of KDD, pages 260–264, Boston, USA.

Berndt, D. J. and Clifford, J. (1994). Using dynamic time warping to find patterns in time series. In Proceedings of AAAI94 Workshop on Knowledge Discovery in Databases, pages 359–370, Seattle, USA.

Chakrabarti, D. and Faloutsos, C. (2002). F4: large-scale automated forecasting using fractals. In Proceedings of CIKM, pages 2–9, McLean, USA.

Faloutsos, C. and Kamel, I. (1994). Beyond uniformity and independence: Analysis of r-trees using the concept of fractal dimension. In Proceedings of PODS, pages 4–13, Minneapolis, USA.

Ganguly, A. R. and Steinhaeuser, K. (2008). Data mining for climate change and impacts. In Proceedings of ICDM, pages 385–394, Pisa, Italy.

Grimm, A. and Tedeschi, R. G. (2008). Enso and extreme rainfall events in south america. Journal of Climate, 22:1589–1609.

IPCC (2007). Intergovernmental panel on climate change. [link]. accessed: March, 2009.

Kaufman, L. and Rousseeuw, P. J. (1990). Finding Groups in Data: an Introduction to Cluster Analysis. John Wiley and Sons.

Nunes, S. A., Romani, L. A. S., Avila, A. M. H., Traina, C., Sousa, E. P. M., and Traina, A. J. M. (2010). Análise baseada em fractais para identificação de mudanças de tendências em múltiplas séries climáticas. In Proceedings of SBBD Short Papers, pages 65–72, Belo Horizonte, Brasil.

Schroeder, M. (1991). Fractals, Chaos, Power Laws. W. H. Freeman and Company.

Sousa, E. P. M., Traina, C., Traina, A. J. M., and Faloutsos, C. (2007a). Measuring evolving data streams’ behavior through their intrinsic dimension. New Generation Computing Journal, 25:33–59.

Sousa, E. P. M., Traina, C., Traina, A. J. M., Wu, L., and Faloutsos, C. (2007b). A fast and effective method to find correlations among attributes in databases. DMKD, 14(3):367 – 407.

Traina, C., Sousa, E. P. M., and Traina, A. J. M. (2005). Using fractals in data mining. In Kantardzic, M. M. and Zurada, J., editors, New Generation of Data Mining Applications, volume 1, pages 599–630 (Chapter 24). Wiley/IEEE Press.

Traina, C., Traina, A. J. M., Wu, L., and Faloutsos, C. (2000). Fast feature selection using fractal dimension. In Proceedings of SBBD, pages 158–171, João Pessoa, Brasil.

Velasco, I. and Fritsch, J. (1987). Mesoscale convective complexes in the americas. J. Geophys. Res., 92:9591–9613.