Trends in sea surface temperature off the coast of Ecuador and the major processes that contribute to them

Abstract

Monthly-averaged sea surface temperatures (SSTs) from seven adjoining sub-regions spanning the coast of Ecuador have been examined for long-term trends and the major processes that contribute to them. The study area extends from the coast out to 84°W longitude, and from 4°S to 2°N latitude. The period of observation extends from 1900 through 2014, a period of 115years.Linear trend analysis shows that the slopes are positive in all sub-regions and statistically significant in 5 out of 7 of them. Based on the trends SSTs increased from +0.10°C to +1.42°C over the entire period with the greatest increases occurring in the sub-regions next to the coast and to a lesser extent in the northern and southern sub-regions.Asecond non-linear trend analysis was conducted using Ensemble Empirical Mode Decomposition (EEMD). The results from EEMD for the various sub-regions indicate that SSTs increased significantly from 1900 to at least 1940 in the majority of cases. After 1950, SSTs tended to fluctuate with a minimum that often occurred during the 1960s or 1970s. Since circa 1990, SSTs have experienced little change. Because the data tended to be spatially homogeneous, with two sub-regions being possible exceptions, a mean data set was calculated and likewise subjected to EEMD. The results show that SSTs have increased by approximately 1°C since 1900 but the process has not been uniform. The increase in temperature between 1900 and at least 1940 approaches 1°C, similar to the trends found in the majority of sub-regions. Between ~1950 and 1990, temperatures decreased slightly until the mid-1960s and then gradually increased until about 1990. Since ~1990 there has been no significant change in SST. Overall, the results of the global mean analysis are generally consistent with the sub-regional results.Several sources of variability have been tentatively identified that may contribute to the long-term changes in temperature that we have observed. Both spectral analysis and EEMD suggest the importance of the Pacific Decadal Oscillation as possibly the greatest contributor. The two major maxima in the PDO index that occurred during the past century (~1940 and ~1990) are generally consistent with maxima and/or inflection points that occur both in the sub-regional and regional long-term trends. It has been difficult to associate specific ENSO episodes with the long-term trends. However, ENSO and the PDO are closely related. Recent modeling studies have shown that the PDO essentially owes its existence to ENSO. Thus, ENSO appears to be a major contributor to the long-term trends through its primary contribution to the PDO. Finally, spectral analysis also revealed the existence of an oscillation whose period is ~73years, easily of sufficient length to influence the long-term trends. Climate-related oscillations in the range of 70–100years are often associated with the Gleissberg sunspot cycle and this could be a contributing factor.The long-term trends are not necessarily a dominant feature in our data. However, had the observations been limited to the period from 1900 to ~1940, the trends would have been, without a doubt, dominant. Overall, the likely influence of the PDO on the observed long-term trends in SST must be emphasized. Finally, over the past 25years there has been no significant change in surface temperature over the study area and this may be due in part to the cooling influence of the PDO since ~1990.