Global correlations between winds and ocean chlorophyll

M. Kahru1, S. T. Gille1, R. Murtugudde2, P. G. Strutton3, M. Manzano-Sarabia4, H. Wang1 and B. G. Mitchell1

1Scripps Institution of Oceanography, University of California San Diego, La Jolla, California, USA

22207 CSS Bldg/ESSIC, University of Maryland, College Park, MD 20742, USA

3Institute for Marine and Antarctic Studies, University of Tasmania, Hobart, Tasmania, Australia

4Facultad de Ciencias del Mar, Universidad Autónoma de Sinaloa, Mazatlán, Sinaloa, México

 Kahru, M., S. T. Gille, R. Murtugudde, P. Strutton, M. Manzano-Sarabia, H. Wang and B. G. Mitchell, Global correlations between winds and ocean chlorophyll, J. Geophys. Res., doi:10.1029/2010JC006500, 2010. PDF

Global time series of satellite derived winds and surface chlorophyll concentration (Chl-a) show patterns of coherent areas with either positive or negative correlations. The correlation between Chl-a and wind speed is generally negative in areas with deep mixed layers and positive in areas with shallow mixed layers. These patterns are interpreted in terms of the main limiting factors that control phytoplankton growth, i.e. either nutrients that control phytoplankton biomass in areas with positive correlation between Chl-a and wind speed, or light, that controls phytoplankton biomass in areas with negative correlation between Chl-a and wind speed. More complex patterns are observed in the equatorial regions due to regional specificities in physical-biological interactions. These correlation patterns can be used to map out the biogeochemical provinces of the world ocean in an objective way.

 

Figure 1. A, correlation coefficient (R) between monthly anomalies of wind speed and Chl-a. N = 144 (corresponding to the number of months of data available), the critical value, Rcrit (P<0.05) = 0.164, Rcrit (P<0.01) = 0.214, i.e. the correlation in the red and blue areas is statistically significant (P<0.05). The selected areas in Western Equatorial Pacific (WEPAC, 2ºN-2ºS, 160ºE-160ºW) and Eastern Equatorial Pacific (EEPAC, 2ºN-2ºS, 140ºW-100ºW) along the equator are shown. The black curves are contour lines of R = 0. TI = Tropical Indian Ocean, NP = North Pacific, TP = Tropical Pacific, SO = Southern Ocean, SP = South Pacific, NA = North Atlantic, TA = Tropical Atlantic, SA = South Atlantic. B, Correlation coefficient between monthly wind speed and sea-surface temperature anomalies, N = 258, the critical value, Rcrit (P<0.05) = 0.122, Rcrit (P<0.01) = 0.160. The correlation in the red and blue areas is statistically significant (P<0.05).

Figure 2. The upper mixed layer depth in spring (March in the Northern Hemisphere and September in the Southern Hemisphere) with the simplified contours separating regions with positive correlations between Chl-a and wind speed anomalies from those with negative correlations (as in Fig. 1A). The monthly mixed layer depth climatology is based on the 0.2ºC threshold criterion [de Boyer Montégut et al., 2004].

Figure 3. Correlations between monthly anomalies of Chl-a and anomalies of the eastward wind pseudostress (uU, m2/s2, A) and northward wind pseudostress (vU, m2/s2, B).

Figure 4. Results of the k-means cluster analysis using the correlation coefficients between the following 7 pairs of variables: (1) monthly Chl-a anomaly and eastward wind pseudostress anomaly, (2) monthly Chl-a anomaly and northward wind pseudostress anomaly, (3) monthly Chl-a anomaly and wind speed anomaly, (4) monthly Chl-a anomaly and eastward wind pseudostress, (5) monthly Chl-a anomaly and northward wind pseudostress, (6) monthly Chl-a anomaly and wind speed and (7) monthly SST anomaly and wind speed anomaly.

Figure 5. Time series of monthly anomalies of wind speed (U, m/s, left axis) and Chl-a (%, right axis) in the northern WEPAC box (2ºN-0ºN, 160ºE-160ºW). R = 0.705, P<0.01. The Multivariate ENSO Index (MEI, http://www.esrl.noaa.gov/psd/data/climateindices/list/) is shown for comparison.

Figure 6. Scatter plots of the monthly Chl-a anomaly versus eastward wind velocity (u, A and C) and northward wind velocity (v, B and D) in the WEPAC area north of the equator (A and B) and south of the equator (C and D). The correlations are slightly stronger if anomalies are used instead of the wind velocities themselves but for simplicity we use the wind velocities here. The black circles are the months during the strong El Niño event of June, 1997 to June, 1998.

Figure 7.   Monthly mean time series in the southern EEPAC box. A, wind speed (U) anomaly, mixed layer depth (MLD) and Chl-a anomaly. Chl-a bloom events (↑) occurred at dips in U anomaly in 1998, 1999/2000, 2003, 2006 and 2007. B, eastward wind velocity (u), MLD anomaly and Chl-a anomaly. C, meridional wind divergence (MWD), SST and Chl-a anomalies. Significant correlations: R (U anomaly, Chl-a anomaly) = -0.438; R (MLD, Chl-a anomaly) = -0.370; R (MLD anomaly, Chl-a anomaly) = -0.516; R (MWD anomaly, Chl-a anomaly) = 0.371; R (SST anomaly, Chl-a anomaly) = -0.671; Rcrit = 0.214 (P<0.01); N = 144.

Figure 8. Cross-spectral squared coherency functions between time series of 5-day anomalies of wind speed and 5-day anomalies of Chl-a in the northern (2ºN-0ºN, 160ºE-160ºW) and southern (0ºS-2ºS, 160ºE-160ºW) WEPAC area. The horizontal short dashes indicate the 95% critical value for testing the null hypothesis of zero coherence [von Storch and Zwiers, 1999].