Long-Term Tendencies and Low-Frequency Variability of Canary Upwelling Intensity and Their Causes

In this paper, trends and low-frequency variability characteristics in the intensity of the Canary Upwelling (CU) have been estimated using satellite data on the wind field and surface level (allowing us to calculate the necessary parameters of Ekman and gradient currents, as well as the waters rise rate), starting from 1993 and up to 2023. It is shown, that the contribution of geostrophic transport to the formation of vertical waters movements in the CU area is approximately half as much as compared with the Ekman drift transport and is comparable to the contribution of the vorticity of the wind field. On average, geostrophic currents forming in the upper ocean layer in the upwelling region lead to a decrease in the intensity of waters rise. The linear trend in the total rate of rise of the upper water layer, averaged over the entire CU area, is about 1.3×10 m/s over 30 years. The significance of the trend does not exceed 85%. This is a consequence of the presence of intense interannual and inter-decadal variability of wind fields and geostrophic currents, as well as the noisiness of the gradient current by subgrid (mesoscale) fluctuations. Received assessment of the long-term trend of total vertical movement of the subsurface waters averaged over the entire CU region is in a good agreement with the estimation of the vertical velocity at the upper mixed layer bottom calculated using oceanic re-analysis data with the spatial resolution of 1/12°×1/12°. The statistical significance of the trend assessed using the re-analysis data is smaller than 80%. At the same time, a significant parabolic trend is distinguished in the time course of the average annual wind speed values and the Ekman upwelling index, which makes a significant contribution to the total variance of the primary series (more than 15%). This can be considered as a manifestation of the Atlantic Multidecadal Oscillation, whose periodicity is approximately twice the length of the analyzed series.