| dc.description.abstract | This research identified correlations between extreme weather events and indices that
characterize the ocean-atmosphere processes using the wavelet coherence method.
The climatic daily data of precipitation and maximum and minimum temperature for
the period 1985 to 2015 were provided by the National Institute of Meteorology and
Hydrology (INAMHI). A total of 177 precipitation stations and 53 temperature stations
were subjected to quality control, data homogenization and filling of missing data using
the Climatol computational package. After data processing, the values of the study
variables were interpolated using two methods: Inverse Distance Weighting (IDW) for
precipitation and co-kriging for temperature. The results obtained from this process
were used to calculate ten climate extreme indices (five related to precipitation and
five related to temperature). Subsequently, using the TREND software, which contains
the Mann-Kendall statistical test, the trends of the extremes were calculated according
to the nine types of climate that coexist in Ecuador. As a result, stationary trends were
obtained for the CDD, PRCTOT and R99pTOT indices, i.e., they show oscillatory
behavior; for the CWD and R95pTOT indices, increasing trends were obtained for
certain types of climate. In the case of temperature, the trends obtained were
statistically significant, showing an increase or decrease according to the climatic
extreme and type of climate analyzed. Finally, cross-correlations between climate
extremes and large-scale indices were performed using the WaveletComp package
(within R) to generate Morlet wavelet power spectra and bivariate analysis with
crossed wavelet. The results show the phase relationship between the analyzed time
series and the observed predominant cycles showing the interaction between
extremes and climatic indices. The highest correlations were observed between the
indices related to the ENSO phenomenon (BEST, SOI, Niño 1+2, Niño 4, Niño 3.4,
PDO and ONI) with correlations higher than 0.8 in different time lapses for return
periods higher than 1.5 years and the climatic extremes related to precipitation and
temperature. Regarding the results of the cross-correlations with temperature
extremes, it is evident that the best correlations are between the large-scale indices
related to ENSO and precipitation. Finally, the correlations obtained with the other
analysis indices were low or negative. | |
