Observed structural relationships between ocean chlorophyll variability and its heating effects on the ENSO

Abstract

Ocean chlorophyll (Chl)-induced heating can affect the climate system through the penetration of solar radiation in the upper ocean. Currently, the ocean biology-induced heating (OBH) feedback effects on the climate in the tropical Pacific are still not well understood, and the mechanisms regarding how SST is modulated remain elusive. In this paper, chlorophyll (Chl) data from satellites are combined with physical fields from Argo profiles to estimate OBH-related fields, including the penetration depth (H_p) and the ocean mixed-layer (ML) depth (H_m). In addition, some directly related heating terms with H_m and H_p are diagnosed, including the absorbed solar radiation component within the ML (denoted as Q_abs), the rate of ML temperature changes that are directly induced by Q_abs (denoted as R_sr = Q_abs/(ρ₀C_pH_m)), and the portion of solar radiation that penetrates through the bottom of the ML (denoted as Q_pen). The structural relationships between these related fields are examined to illustrate how these heating terms are affected by H_p and H_m. The extent to which R_sr and Q_pen are modulated by H_p is strikingly different during ENSO cycles. In the western-central equatorial Pacific, inter-annual variations in H_p tend to be out of phase with those in H_m. A decrease (increase) in Q_abs from a positive (negative) H_p anomaly during El Niño (La Niña) tends to be offset by a negative (positive) H_m anomaly. Thus, R_sr is not closely related with H_p, even though Q_abs is highly correlated with H_p, indicating that the direct thermal effect through Q_abs is not a dominant factor that affects the SST. In contrast, the inter-annual variability of Q_pen in the region is significantly enhanced by that of H_p, with their high positive correlation. The H_p-induced differential heating in the ML and subsurface layers from the Q_pen and Q_abs terms modifies the thermal contrast, stratification and vertical mixing, which represent a dominant indirect ocean dynamical effect on the SST. The revealed relationships between these related fields provide an observational basis for gaining structural insights into the OBH feedback effects and validating model simulations in the tropical Pacific.

Appendix: Mean state and seasonal variations

Observations are used to analyze the structure and variability of ocean biology-related heating effects in the tropical Pacific, including Chl, H_p, H_m, short wave (SW) solar radiation, and the three heating terms. The three heating terms, which were derived from combined satellite and in situ data, are new fields that have not been shown before. In this Appendix, we present the mean state and seasonal variabilities of these related fields to support the inter-annual variability analyses in the main text.

1.1 A1. Chl field

Figure 15 displays the annual-mean field of Chl and its variations in the equatorial Pacific according to satellite measurements. The magnitude of the seasonal and inter-annual Chl variabilities is comparable to that of its total value. The main characteristics of Chl are clearly evident in Fig. 1b and Fig. 15a, as previously described (e.g., Ballabrera-Poy et al. 2007). For example, a pronounced ridge is seen in the equatorial regions, with high Chl values extending from the east to the west along the Equator. The Chl concentration is low (< 0.1 mg m⁻³) in the western equatorial Pacific in association with warm waters and high in the central and eastern equatorial regions, where a cold tongue develops. In the far-eastern coastal regions, the Chl concentration is especially high. Correspondingly, large gradients exist in the western equatorial region and far-eastern region.

Fig. 15

Estimated Chl fields during 2005–2015 from satellite observations: a horizontal distribution for the annual mean climatology and b seasonal variations along the Equator. The contour interval is 0.02 mg m⁻³ in a and b

The annual-mean Chl field and corresponding seasonal variability of Chl along the Equator are shown in Fig. 15. Longitudinally, Chl exhibits elevated values from west to east across the equatorial Pacific. Large seasonal variations in Chl are seen in the western-central equatorial Pacific, with a pronounced peak in summer. In the eastern equatorial region, low values are observed in spring and high values in summer. These seasonal variations in Chl are associated with those in physical conditions in the equatorial Pacific (e.g., upwelling and vertical mixing).

1.2 A2. Short wave (SW) solar radiation, H_p and H_m

Figure 16a exhibits the horizontal distributions of the average annual-mean shortwave (SW) solar radiation during 2005–2015; its corresponding seasonal variation is shown in Fig. 17a. Regions with large values are seen in the western-central equatorial Pacific and those with low values are seen in the eastern region. Seasonally, solar radiation that reaches the sea surface has a pronounced semiannual cycle along the Equator (Fig. 17a). The SW radiation penetrates the upper ocean.

Fig. 16

Horizontal distributions of the annual-mean fields during 2005–2015: a shortwave radiation from MERRA, b H_m from the Argo product, and c H_p from the GlobColour Project. The contour interval is 10 W m² in a, 5 m in b, and 1 m in c

Fig. 17

Seasonal variations along the Equator for climatological fields during 2005–2015: a shortwave radiation, b H_m, and c H_p. The contour interval is 10 W m⁻² in a, 5 m in b, and 1 m in c

The corresponding results for the depth of the ML (H_m) are displayed in Figs. 16b and 17b. Regions with a deep ML are located in the western-central equatorial Pacific and those with a shallow ML are seen in the western and eastern regions. In both the western and eastern sides of the tropical Pacific and the Intertropical Convergence Zone (ITCZ), the ML is relatively shallow because of the weak surface friction velocity and the stabilizing surface buoyancy flux. Values larger than 50 m are found in the central tropical regions, where surface winds are energetic with strong surface buoyancy losses. Seasonal variations are clearly evident. In the eastern equatorial region, the ML is shallow in spring and deep in fall. In the western equatorial basin, the ML is deep in winter but shallow in the late spring.

Chl is used to derive the penetration depth (H_p), and its annual-mean structure is shown in Fig. 15a. A pronounced trough is seen in the equatorial regions, with low H_p values extending from east to west along the Equator. Deep penetration depths with values larger than 20 m are observed in the western equatorial Pacific, while shallow penetration depths are found in the east. Large gradients are located in the western equatorial Pacific and far-eastern regions. Seasonally, the solar radiation exhibits deep penetration in spring and shallow penetration in fall.

1.3 A3. Three heating terms (Q_abs, Q_pen and R_sr)

As mathematically expressed above, SW, H_p and H_m all determine the distributions of penetrative solar radiation between the mixed layer and the underlying subsurface layers. The annual-mean structures of these three heating terms are shown in Fig. 18, and the corresponding seasonal variations are displayed in Fig. 19. The magnitude of the total Q_abs and Q_pen fields is one order larger than that of their inter-annual variabilities. Most of the solar radiation is absorbed within the ML (Fig. 18a), with some penetrating through the bottom of the ML (Fig. 18b). These terms exhibit coherent relationships with SW, H_m and H_p. For example, the structure and magnitude of Q_abs is similar to those of SW; Q_pen has low values in the western-central equatorial Pacific and high values in the western and eastern equatorial Pacific. Q_abs, Q_pen and R_sr all have a clear signature for H_m, so H_m is major factor that affects the penetration of solar radiation. Seasonally, Q_abs has a pronounced semiannual cycle along the equator (Fig. 19a), similar to the SW solar radiation reaching the sea surface. Interestingly, seasonal variations in Q_pen and R_sr (Fig. 19b, c) exhibit a slight shift in time compared to Q_abs (Fig. 19c). The seasonal variations in these heating terms indicate that these terms are predominantly determined by H_m. No clear signature is seen for the effect of H_p on these heating terms in terms of the mean field and seasonal variations.

Fig. 18

Horizontal distributions of the annual-mean fields during 2005-2015 for ocean Chl-related heating terms: a Q_abs, b Q_pen, and c R_sr. The contour interval is 10 W m⁻² in a, 5 W m⁻² in b and 1 °C month⁻¹ in c

Fig. 19

Seasonal variations along the Equator for the climatological heating terms during 2005–2015: a Q_abs, b Q_pen, and c R_sr. The contour interval is 10 W m⁻² in a, 5 W m⁻² in b and 1 °C month⁻¹ in c

Thus, H_m is a major factor that controls the distribution of solar radiation within the ML and subsurface layers. As such, if the ML is deeper, SW solar radiation is absorbed more within the ML and penetrates less into the subsurface layers. If the ML is deep enough, all the radiation would be absorbed within the ML, with little penetration through the bottom of the ML. In such a situation, H_p would barely influence the penetrative solar radiation in the upper ocean.