El periodismo femenino en Egipto (1952-1965)

From Chapter 3, we have seen that the PBE mostly encompasses the in situ surface chlorophyll concen- trations at all of the stations. However, in the oligotrophic stations, the DIN concentrations are overes- timated by the ensemble. Although the ensemble has been reduced, the in situ chlorophyll at all of the

5.3. The effect of perturbing biogeochemistry, physics, and both 149

Figure 5.2: Ensemble range of annual mean of chlorophyll (a to d), DIN (e to h), and zooplankton (i to l) profiles between 1998 to 2007 at BATS (a, e, and i), ALOHA (b, f, and j), PAP (c, g, and k), and Cariaco (d, h, and l). Blue cross, red, yellow, black, and green dots denote the mean concentrations from the default run, in situ, the ensemble mean of PPE, PBE, and perturbing both biogeochemistry and physics. For station PAP, the annual mean is taken form 2004 for DIN and 2003 for chlorophyll. Station L4 profiles are not shown because in situ data is only available in the surface.

stations are still within the ensemble range. In terms of DIN, the ensemble at the oligotrophic stations still produces ranges that overestimates the in situ concentrations in most depth. For other stations, the ensemble ranges can still capture the in situ chlorophyll and DIN, similar to the full ensemble.

Perturbations to the vertical velocity used for the PPE, produce relatively little spread in the bulk prop- erties (e.g. the total concentration of DIN, chlorophyll, and zooplankton) at each depth. The immediate impact should be seen in the concentrations of DIN at upper ocean levels but in the biologically active top (∼75m) the PPE range does not change significantly although it widens at depth (Figure 5.2) showing that the vertical velocity variations are having an impact below the biologically active layers. However this does not have a big impact on bulk properties (Figure 5.2). These bulk properties have also been seen to be fairly insensitive when different ocean general circulation models have been coupled with the same ocean biogeochemical model (e.g., Sinha et al. (2010)).

The spreads generated by perturbing both physics and biology (i.e., PBPE) are mostly only slightly wider than for PBE alone, at least in the biologically active zone. Below this layer, the ensemble from PPE, especially in the oligotrophic stations produces larger spread (Figure 5.2(e) and (f)). However at Cariaco and PAP, the PBPE produces slightly larger DIN spread than PPE and PBE, even below the biologically active depths (Figure 5.2(g) and (h)). The spreads generated by perturbed physics alone are therefore mostly insufficient to encompass the in situ observations, especially for chlorophyll. In contrast the observed concentrations of chlorophyll at all five stations, from surface to deep water, are mostly within the PBE range (Figure 5.2(a)-(d)), suggesting that the full range of biological production through a strong nutrient gradient can be obtained by perturbing the biological processes equations.

Discrepancy between observations and ensemble simulation

If we consider individual stations, as discussed in Chapter 3, at Cariaco good agreement is found for DIN between the in situ and all the ensembles; PBE, PPE, and PBPE. However, at the oligotrophic stations BATS and ALOHA, a large mismatch between the observed and modelled DIN is apparent (Fig- ure 5.2(e)–(h)). Similar to the results in Chapter 3, the observed DIN in the top 150m (Figure 5.5(e)–(f)) are beyond the ranges produced by either PBE or PBPE. This discrepancy in the oligotrophic stations indicates that uptake process is insufficient at low nutrient concentrations, leaving high DIN concen- trations along with an underestimation of chlorophyll, (Yool et al., 2011; Cox and Kwiatkowski, 2013; Kwiatkowski et al., 2014) similar to the results in Chapter 3. Furthermore, at ALOHA the range pro- duced by the PBE and PBPE is not wide enough to encompass the in situ maximum chlorophyll depth (∼ 110m). The PBPE and PBE produce maximum ranges at ∼ 80m, therefore at greater depths, the light availability is even lower, and therefore low phytoplankton growth rate is simulated, despite higher DIN

5.3. The effect of perturbing biogeochemistry, physics, and both 151

concentrations at this depth. Meanwhile, at BATS, the oligotrophic conditions have not been simulated well by the ensemble, where at the top ∼ 75m, the DIN concentration is a magnitude higher than the in situ. However, the in situ chlorophyll concentration at BATS is within the ensemble range. This may be due to the 1-D model not simulating the input temperature well, making the nutrient uptake at BATS done inefficiently.

Compared to the previous run in Chapter 3, the in situ DIN concentrations at station PAP that are within the full ensemble (at the top ∼40 m, Figure 3.1) are now outside the PBE range, shown in Figure 5.2(g). The observed DIN profile does not show an increase in DIN concentrations with depth. However, the model ensembles do show nutrients increasing with depth, such that DIN concentration is underestimated near the surface (at <75m), and overestimated at depth >120m (Figure 5.2(g)). From Figure 5.3, the DIN concentrations are not sampled evenly at different depths. This might be due to the quality-controlled DIN concentrations only being available at certain times and depths, and lateral advection that occurred between 2003 and 2004, which increases the DIN concentration below ∼150m (Hartman et al., 2015). The in situ DIN profile shown in Figure 5.2(g) shows little variation with depth, unlike the ensemble simulation. However, the chlorophyll, shown in Figure 5.4, has more samples that are quality controlled at different depths and times.

Figure 5.3: Monthly mean of in situ DIN (nitrate plus nitrite) at station PAP. The samples are taken between 2002 to 2004, in a sensor frame at 30m, which samples within deep chlorophyll maxima .