Spatio-temporal variability of droughts over past 80 years in Madeira Island

• Published in: Journal of Hydrology: Regional Studies
• Main Authors: Luis Angel Espinosa, Maria Manuela Portela, Rui Rodrigues
• Format: Article in Journal/Newspaper
• Language: English
• Published: Elsevier 2019
• Subjects: Physical geography; GB3-5030; Geology; QE1-996.5; North Atlantic; North Atlantic oscillation
• Online Access: https://doi.org/10.1016/j.ejrh.2019.100623; https://doaj.org/article/0d52ba000372434199413aff80fd4ce1

Study region: Madeira Island, Portugal (a small island of the North Atlantic Ocean). Study focus: Droughts from January 1937 to December 2016, were studied using monthly rainfall at 41 rain gauges covering most of the island. The gaps of the original rainfall data were filled by Multiple Imputation by Chained Equations (MICE). The drought conditions were assessed by means of the Standardized Precipitation Index (SPI), specifically the SPI6. To study the drought variability, some clustering techniques and principal components analysis (PCA) were applied to the SPI field. New hydrological insights for the region: Three homogeneous regions (northern slope, southern slope, and central region), each of them with different temporal climatic variability may be identified. Furthermore, for each region, the droughts were characterized in terms of magnitude and duration, and based on a kernel occurrence rate estimator (KORE) as well as on frequency of the drought periods. Finally, two climatic drivers, namely the El Niño-Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO) were teleconnected to the drought variability at the identified homogeneous regions for the period under study. It was not possible to establish a clear relationship between ENSO/NAO and drought occurrence. Nevertheless, the results showed that the spatio-temporal drought variability in Madeira Island has been subjected to noticeable changes in recent years (2001–2016) with a considerable higher number of periods under drought conditions than in the past. Keywords: Drought, Standardized Precipitation Index (SPI), MICE, Principal components analysis (PCA), Occurrence rate, El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO)