A Bayesian Network Approach to Identity Climate Teleconnections Within Homogeneous Precipitation Regions in Ecuador
Reliable precipitation predictions require an understanding of climate teleconnections over precipitation. In Ecuador, these teleconnections were studied with correlation methods, but multivariate studies with several climatic indexes simultaneously has been less study. The objective of this work is to carry out a multivariate study using Bayesian networks to identify the influence of climate indexes in homogenous precipitation regions in Ecuador. The climate teleconnections, defined as the correlation between precipitation satellite data and climate indexes, as well as the regionalization of seasonality of precipitation were used to learn a Bayesian network in R software. It was characterized the structure and strength of the relationship between the teleconnections and the precipitation. Additionally, three types of belief propagation were used: regions to climate index, climate index to regions, and interactions between indexes. This was useful to determine whether the influence of a climate index is homogeneous throughout the country or varies by region, as well as to identify interactions between different indexes. The results of this study contribute to a better understanding of precipitation in Ecuador, and to promote making evidence-based water resource decisions.
KeywordsClimate teleconnections Bayesian networks Probability propagation
This study has been financed by the Corporación Ecuatoriana para el Desarrollo de la Investigación y la Academia (CEDIA) through the project CEPRA XII “Spatial representation of climatic teleconnections in the precipitation of Ecuador”.
- 4.Pratiwi, R., Sukardjo, S.: Effects of rainfall on the population of Shrimps Penaeus Monodon Fabricius in Segara Anakan lagoon, Central Java, Indonesia. 2(3), 156–169 (2018). https://doi.org/10.11598/btb.2018.25.3.830
- 14.Das, M., Ghosh, S.K.: A probabilistic approach for weather forecast using spatio-temporal inter-relationships among climate variables. In: 9th International Conference on Industrial and Information Systems, ICIIS 2014 (2015). https://doi.org/10.1109/ICIINFS.2014.7036528
- 22.Ballari, D., Giraldo, R., Campozano, L., Samaniego, E.: Spatial functional data analysis for regionalizing precipitation seasonality and intensity in a sparsely monitored region: unveiling the spatio-temporal dependencies of precipitation in Ecuador. Int. J. Climatol. 38(8), 3337–3354 (2018). https://doi.org/10.1002/joc.5504CrossRefGoogle Scholar
- 24.López, D.A.G.: Algoritmo de Discretización de Series de Tiempo Basado en Entropía y su Aplicación en Datos Colposcópicos (2007). http://cdigital.uv.mx/bitstream/123456789/32352/1/garcialopezdaniel.pdf
- 25.Scutari, M.: Package ‘bnlearn’ (2019). https://cran.r-project.org/web/packages/bnlearn/bnlearn.pdf
- 30.Carvalho, A.: Scoring functions for learning Bayesian networks. INESC-ID Technical report 54/2009, pp. 1–27 (2009). https://pdfs.semanticscholar.org/6efe/f4bacfb14cfe4c1ababae751904431b75cc9.pdf