Portal de Periódicos da CAPES

Episodic upwelling and dust deposition as bloom triggers in low-nutrient, low-chlorophyll regions

2011; American Geophysical Union; Volume: 116; Issue: C6 Linguagem: Inglês

10.1029/2010jc006704

2156-2202

Paulo H. R. Calil, Scott C. Doney, Keiya Yumimoto, Kenta Eguchi, Toshihiko Takemura,

Marine Biology and Ecology Research

Journal of Geophysical Research: OceansVolume 116, Issue C6 Free Access Episodic upwelling and dust deposition as bloom triggers in low-nutrient, low-chlorophyll regions Paulo H. R. Calil, Paulo H. R. Calil [email protected] Marine Chemistry and Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA Now at National Institute of Water and Atmospheric Research, Wellington, New Zealand.Search for more papers by this authorScott C. Doney, Scott C. Doney Marine Chemistry and Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USASearch for more papers by this authorKeiya Yumimoto, Keiya Yumimoto Meteorlogical Research Institute, Japan Meteorological Agency, Ibaraki, JapanSearch for more papers by this authorKenta Eguchi, Kenta Eguchi Research Institute for Applied Mechanics, Kyushu University, Fukuoka, JapanSearch for more papers by this authorToshihiko Takemura, Toshihiko Takemura Research Institute for Applied Mechanics, Kyushu University, Fukuoka, JapanSearch for more papers by this author Paulo H. R. Calil, Paulo H. R. Calil [email protected] Marine Chemistry and Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA Now at National Institute of Water and Atmospheric Research, Wellington, New Zealand.Search for more papers by this authorScott C. Doney, Scott C. Doney Marine Chemistry and Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USASearch for more papers by this authorKeiya Yumimoto, Keiya Yumimoto Meteorlogical Research Institute, Japan Meteorological Agency, Ibaraki, JapanSearch for more papers by this authorKenta Eguchi, Kenta Eguchi Research Institute for Applied Mechanics, Kyushu University, Fukuoka, JapanSearch for more papers by this authorToshihiko Takemura, Toshihiko Takemura Research Institute for Applied Mechanics, Kyushu University, Fukuoka, JapanSearch for more papers by this author First published: 30 June 2011 https://doi.org/10.1029/2010JC006704Citations: 39AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract [1] Summertime phytoplankton blooms in the oligotrophic North Pacific Ocean are supported by N2-fixing organisms that relieve the system of nitrate limitation. Phosphate and iron, however, limit their growth and need to be supplied for these organisms to thrive. We analyze two recent blooms in the region whose differences provide insight into their possible formation mechanisms. In 2008, a typical late summer bloom, with sporadic patches of higher-chlorophyll concentration, occurred near the island chain and the subtropical front. In 2010, an unusually large, contiguous bloom was observed in the western oligotrophic North Pacific, a region where blooms seldom, if ever, occur. Streaks of high chlorophyll in 2008 coincide with surface temperature fronts and regions of large horizontal stretching, as detected by Lagrangian diagnostics. Such regions are prone to the generation of vertical velocities via frontogenesis. Horizontal transport from upwelling regions or iron-rich island sediments is also important for the redistribution of nutrients. In the case of the 2010 bloom, we use a global aerosol transport model as well as space-borne lidar observations to argue that atmospheric dust deposition events prior to the bloom provided the necessary nutrient conditions for the growth of N2-fixing organisms. As sea surface temperature increased in the region, chlorophyll values increased significantly, showing that this bloom was likely a consequence of prior enrichment and that temperature is a key factor in bloom development in this important biome. Key Points Episodic upwelling associated with fronts impact chlorophyll distribution Dust deposition fertilize low nutrient, low chlorophyll regions Horizontal stirring controls the redistribution of tracers 1. Introduction [2] The oligotrophic regions of the ocean's subtropical gyres are important components of the machinery of the Earth's climate and biogeochemical dynamics. In the central portion of the North Pacific Subtropical Gyre (NPSG), the Hawaiian Ocean Time series (HOT) Station ALOHA (A Long-term Oligotrophic Habitat Assessment) has contributed significantly in revealing new insights into the interplay between physical and biogeochemical processes over multiple time scales [Karl and Lukas, 1996; Karl et al., 1996]. [3] The ocean around Station ALOHA (22°45′N, 158°W) is representative of a low-nutrient, low-chlorophyll (LNLC) region [Karl and Letelier, 2008]. The LNLC region arises mainly because of the large vertical stratification and deep pycnocline generated by the wind-forced anticyclonic subtropical gyre, which isolates the well-lit, nutrient-depleted surface ocean from the nutrient-rich waters below the euphotic zone. [4] Late summer phytoplankton blooms commonly observed in the vicinity of Station ALOHA are mostly supported by a rich diversity of diazotrophs including the cyanobacteria Trichodesmium spp., the symbiotic bacteria in diatoms Richelia, croccosphera and uncultured unicellular organisms [Dore et al., 2008]. These organisms are able to fix dissolved gaseous nitrogen into bioavailable inorganic form, thus relieving the system of nitrate limitation, which is particularly severe in the summertime because of enhanced stratification and inhibited vertical mixing. [5] Diazotrophs, however, depend heavily on micronutrients such as iron, required for nitrogenase [Sanudo-Wilhelmy et al., 2001], and phosphate. Atmospheric dust deposition can certainly be a source of iron and potentially other micronutrients [Jickells et al., 2005; Moore et al., 2004], but no study has hitherto demonstrated a clear relationship between deposition events and the summer blooms in this remote area of the ocean. New species of nitrogen fixing organisms, with different adaptive traits, are still being discovered [Moisander et al., 2010]. Thus it is difficult at this stage to make definitive statements about the response of the different groups of diazotrophs to environmental forcing. Nevertheless, temperature does seem be a critical constraint on the abundance and growth of nitrogen fixing organisms [Moisander et al., 2010]. Trichodesmium spp., for example, an important diazotroph frequently observed around Station ALOHA [Karl et al., 1997], is thought to have higher growth rates within the temperature range of 24–30°C, with maximum growth rates at 27°C [Breitbarth et al., 2007]. 2. Summer Blooms Around Station ALOHA [6] The Hawaiian archipelago is located approximately at the center of the NPSG. Annual sea surface temperature (SST) from the Moderate Resolution Imaging Spectroradiometer (MODIS; see section 3) averaged from 2002 to 2009 (Figure 1, top) reveal that, on average, waters warmer than 25°C encompass the island chain. The average position of the subtropical front (STF) is outlined by the 22°C isotherm. In general, there is a southward bent of isotherms east of 150°W, a consequence of the Sverdrup flow, which may enhance zonal SST gradients in different seasons (see section 4.1). Surface chlorophyll-a concentration from MODIS averaged from 2002 to 2009 (Figure 1, bottom) is very low (≈0.04 mg chl-a m−3) in the western portion of the gyre (west of 165°). This is the region where warm waters from the western North Pacific flow eastward along the subtropical counter current system [Kobashi and Kawamura, 2002]. The slightly higher surface chlorophyll-a concentration observed in the vicinity of the islands possibly results from a combination of physical, biological and geochemical factors. These factors affect the supply of nutrients and influence primary production in the region. Physical processes include the vertical displacement of isopycnals caused by the formation and self-interaction of mesoscale eddies, which are particularly strong in the lee of the islands [cf. Benitez-Nelson et al., 2007; Calil et al., 2008], and the large horizontal stirring associated with these features [Calil and Richards, 2010]. Biological factors include N2 fixation [Dore et al., 2008], recycling and vertical migration of diatom mats [Brzezinski et al., 1998]. Figure 1Open in figure viewerPowerPoint (top) Surface SST and (bottom) chlorophyll-a for the subtropical North Pacific ocean. The circle is centered at the location of Station ALOHA. The 22°C isotherm represents the average position of the subtropical front. The 25°C isotherm is highlighted because of its potential importance for N2-fixing organisms (see text). [7] Regional geochemical characteristics are also important. Previous studies have observed higher dissolved iron surface concentrations near the Hawaiian Islands (≈0.5 nM) than in the surrounding North Pacific [Noble et al., 2008; Brown et al., 2005]. The interaction of the eddies with the topography of the islands may influence the lateral transport of inorganic nutrients from iron-rich island sediments into the open ocean. Horizontal stirring is enhanced along specific regions of the flow field where passive tracers can be transported over long distances [Rypina et al., 2010; Lehahn et al., 2007]. This could be the reason why the region around Station ALOHA is phosphate limited rather than iron limited, while the western portion of the NPSG experiences iron-phosphate colimitation [Dore et al., 2008]. In fact, Shiozaki et al. [2010] suggest that the proximity to landmasses is a key factor in generating "N2 fixation hot spots" and that previous estimates of N2 fixation in the oligotrophic ocean are likely overestimates because most of the available measurements are from such areas (e.g., near isolated islands). [8] Climatological values (from 1988 to 2009) of key hydrographic indices measured at Station ALOHA (Figure 2) show that temperature has a larger influence on density in the surface ocean, in a climatological sense, than salinity. It is also seen that 14C-based primary productivity is characterized by surface (upper 50 m) maxima from late spring to early fall [Dore et al., 2008]. Two peaks in production are observed, one in late spring, following the shoaling of the mixed layer depth, and one in late summer, when surface waters are strongly stratified and extremely nutrient poor. Figure 2Open in figure viewerPowerPoint Climatology for physical and biological parameters at Station ALOHA. The climatological mixed layer depth is outlined as a black line. [9] Prior to 2010, summer blooms in the oligotrophic North Pacific had been reported to occur only to the east of 160°W [see, e.g., Dore et al., 2008, Figure 1; Wilson et al., 2008, Figure 2], namely, in the vicinity of Station ALOHA and at the approximate summertime location of the STF, whose latitude is approximately 30°N. In the STF shallow nutriclines and large, horizontal density and nutrient gradients make it easier for vertical and horizontal physical processes to supply substantial amounts of nutrients into the surface. In fact, recent observational studies show that intrusions driven by submesoscale instabilities seem to be ubiquitous in the STF [Shcherbina et al., 2010; Hosegood et al., 2008]. Blooms in this region usually sustain larger organisms such as diatoms [cf. Wilson et al., 2008; Brzezinski et al., 1998], which require large nutrient delivery. [10] In the nutrient-depleted, central portion of the oligotrophic ocean, surface frontal processes are not as common because the deeper thermocline and strong vertical stratification prevent the formation of strong horizontal density gradients at the surface. Nutrient levels and nutrient supply are also low in this region. As a consequence, smaller organisms, adapted to low-nutrient conditions are dominant. N2-fixing organisms are important throughout most of the blooms [Dore et al., 2008], with the emergence of larger organisms dependent on the amount of fixed nitrate. Exceptions to this rule are the blooms induced by the strong, locally formed cyclonic eddies in the lee of the island of Hawaii [Benitez-Nelson et al., 2007]. Surface waters in the summertime are consistently warmer than 25°C, a threshold above which diazotrophs are thought to have higher growth rates [Goebel et al., 2008; Breitbarth et al., 2007; White et al., 2007; Fennel et al., 2001]. It is of note, however, that these organisms depend heavily on iron and phosphorus, which must be somehow supplied prior to bloom events. [11] On space scales larger than the mesoscale ((O) (10–100 km)) and time scales of seasonal to decadal, the supply of nutrients into the surface ocean is modulated by large-scale characteristics of the flow field [cf. Di Lorenzo et al., 2008; Stammer et al., 2008; McGillicuddy et al., 2003]. Our focus here will be on episodic processes on smaller scales (e.g., meso- and possibly submesoscale) that generate large vertical velocities that could relieve the oligotrophic ocean of nutrient limitation, hence allowing bloom formation. [12] It is well known that shear or strain deformation associated with the mesoscale eddy flow may sharpen existing horizontal density gradients to the point where the thermal wind balance is disrupted (i.e., the balance between the vertical shear of the horizontal geostrophic velocities and the horizontal density gradients). As a result, vertical velocities are generated with the directional sense such to maintain the thermal wind balance by flattening steeply sloping isopycnal surfaces. This process is responsible in large part for the restratification of the surface ocean [Klein and Lapeyre, 2009; Lapeyre et al., 2006; Hoskins, 1982]. These vertical velocities, which are intertwined with regions where large horizontal straining act upon existing horizontal density gradients [Klein et al., 1998; Lévy et al., 2001], could potentially have a significant biogeochemical impact by bringing nutrients into the euphotic zone or by exporting organic material into the deep ocean (L. Guidi et al., Eddy-eddy interactions promote phytoplankton production and carbon export in the ocean, submitted to Geophysical Research Letters, 2010). In fact, subsurface injections of phosphorus and iron could stimulate the growth of diazotrophs [Karl and Letelier, 2008]. [13] In this paper we analyze several recent, open ocean blooms in the oligotrophic NPSG (Figure 3). In 2008, a variety of meso- and submesoscale patches with high surface chlorophyll-a concentrations were remotely observed from the Moderate Resolution Imaging Spectroradiometer (see section 3) near Station ALOHA, and at the STF, northeast of the islands, between 30 and 35°N. In 2010, an unusually large, contiguous bloom took place in the western NPSG, from 168°E to approximately 165°W, a region where summer phytoplankton blooms are rarely, if ever, observed. This bloom was preceded by an atmospheric dust deposition event. Figure 3Open in figure viewerPowerPoint Snapshots of 8 day average surface chlorophyll from MODIS for the (top) 2008 and (bottom) 2010 blooms during their period of maximum areal extent. [14] We propose that the 2010 bloom, which occurred in a region where phosphorus and iron colimit phytoplankton growth, was the consequence of a combination of events: the dust deposition event, episodic upwelling caused by frontal processes and the seasonal increase in temperature. The 2008 bloom, which occurred in a phosphate-limited region, was a consequence of episodic upwelling when the temperature was already above the estimated threshold of 25°C for diazotrophs (see section 4.1). 3. Data and Methods [15] The surface chlorophyll-a concentration and SST images are obtained from MODIS sensor on NASA's Aqua satellite. We use the 8 day composites of the Level 3 standard mapped image products, which are image representations of binned data products generated from MODIS. The chlorophyll-a concentration and the 11 micron nighttime SST are mapped onto geophysical coordinates at a nominal resolution of 4 km. Data is available at http://oceancolor.gsfc.nasa.gov/. For the calculation of the absolute magnitude of the SST gradient, ∣∇SST∣, a moving average of 5 grid points was applied to the 4 km resolution data in order to reduce the effect of speckling and "cloud fronts" in the composite images. While the smoothing does not fully resolve the problem, it gives more emphasis to the dynamically relevant features as the purpose of the calculation is to show how SST horizontal gradients vary seasonally. A linear interpolation was applied to the SST data set to fill missing points due to the clouds prior to the computation of the SST gradient. An upper limit in ∣∇SST∣ of 0.03°C km−1 was imposed in order to remove spuriously large values (e.g., missing data) associated with the differentiation of the SST field. [16] To estimate dust deposition over the North Pacific ocean, the global Spectral Radiation-Transport Model for Aerosol Species (SPRINTARS) [Takemura et al., 2005], available at http://sprintars.net/indexe.html, was used. The SPRINTARS model was developed jointly by the Center for Climate System Research of the University of Tokyo, National Institute for Environmental Studies and Frontier Research Center for Global Change. This model treats the major tropospheric aerosol components (black carbon, organics, sulphate, soil dust and sea salt), and includes explicit estimates of the aerosol direct and indirect effects. In this study, the horizontal resolution was set to T106 (≈100 km) with 56 vertical layers in a sigma coordinate. Meteorological boundary conditions are taken form the National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data (2.5° by 2.5° resolution, available at http://www.esrl.noaa.gov/psd/data/reanalysis/reanalysis.shtml). The dust emission scheme was based on Takemura et al. [2009]. [17] The Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) is a space-based lidar onboard the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) satellite [Winker et al., 2007], available at http://www-calipso.larc.nasa.gov/. CALIOP has a two-wavelength, polarization-sensitive backscatter lidar. Since the launch on 28 April 2006, CALIOP has been providing unique measurements of the global vertical distributions of aerosols and clouds with very high spatial resolution. In this study, we used the total attenuated backscatter coefficient at 532 nm, aerosol subtype and Vertical Feature Mask (VFM) [Liu et al., 2009]. [18] To investigate the origin of dust particles, we also performed a HYSPLIT trajectory model analysis. The HYSPLIT model [Draxler and Hess, 1998], available at http://ready.arl.noaa.gov/HYSPLIT.php, computes air parcel trajectories. We used the Global Data Assimilation System global reanalysis data as the meteorological inputs. SPRINTARS, CALIOP, and HYSPLIT have been previously used for analyses of trans-Pacific transport of Asian Dust [cf. Uno et al., 2009; Yumimoto et al., 2009, 2010]. Lyapunov Exponents [19] Horizontal stirring has been shown to be important in controlling the evolution of open ocean blooms and creating heterogeneity in tracer fields [Tzella and Haynes, 2007; Abraham and Bowen, 2002; Abraham, 1998]. Passive tracer dispersion occurs preferentially in regions of large stirring while the horizontal transport of tracers across these regions is strongly constrained. From an Eulerian perspective, such regions would be those where shear (σs) and strain (σn) deformation, where σs = δxv + δyu and σn = δxu − δyv, are larger than the relative vorticity, ζ, where ζ = δxv − δyu, u and v are the horizontal components of the velocity field and δx, δy the horizontal derivatives with respect to longitude and latitude, respectively, e.g., the edges of mesoscale vortices [cf. Legal et al., 2007]. Tracers, however, will only be affected by stirring if they remain in such regions for some time. Hence, Lagrangian diagnostics are more appropriate to detect stirring regions that substantially affect tracer distribution. [20] A useful tool to identify such regions is the Finite Size Lyapunov Exponent (FSLE) [d'Ovidio et al., 2004] (see Appendix A for details on the technique). FSLEs are a measure of the horizontal stretching of a given velocity field and their calculation is based on the separation vector between two adjacent water parcels. They can be calculated backward or forward in time with unstable manifolds, marked by large backward FSLE values, representing regions of rapid particle separation (i.e., large stretching), and the stable manifolds marked by large forward FSLE values, representing regions of maximum confluence. [21] The FSLEs are calculated as λ = log(δf/δ0), where δ0 and δf the initial and final distance of particle pairs, respectively, and τ the time the particle pair took to reach δf. We use the horizontal geostrophic velocities obtained from the AVISO merged sea surface height anomaly data, with spatial resolution of approximately 30 km. The initial (δ0) and final (δf) synthetic float separations are 1 and 60 km, respectively. δf was chosen so that it is representative of the mesoscale straining effect. Maximum integration time is 100 days. [22] The regions where unstable and stable manifolds cross each other are known as hyperbolic points. Passive tracers located in these regions are stretched along the unstable manifolds and compressed along the stable manifolds [Lehahn et al., 2007]. Hence, preexisting surface tracer gradients tend to be sharpened along and are generally aligned with stretching regions [Abraham and Bowen, 2002]. If the tracer in question is potential density, the sharpening of its horizontal gradients by the mesoscale deformation field will tend to disrupt the thermal wind balance and make these regions prone to vertical motion (see section 3.2). [23] An illustrative example of the different advection characteristics at meso- and submesoscales in a given flow field is shown in Figure 4 (middle). Using AVISO velocities, backward (negative) and forward (positive) FSLEs displayed in Figure 4 were calculated for 24 July 2008, during the 2008 bloom (see section 4.1). Three circular patches of synthetic floats were released on 24 June 2008 and advected for 30 days, using the time-varying velocity field, at the center of a cyclone located at 157.3°W, 25.05°N, an anticyclone located at 156.2°W, 22.6°N and in the region between these two eddies at 157.6°W, 23.5°N, a snapshot of the SST and surface geostrophic velocities for the same date is shown in Figure 4 (top). Note the initially circular patch of synthetic floats seeded at the location between the two eddies is split into two patches (one to the west and one to the east), which are elongated into narrow filaments. From the FSLEs, it is clear that the region between the two eddies encompasses a hyperbolic point with the unstable manifold (blue line) in the west-east direction and the stable manifold (red line) in the northwest-south direction. Figure 4Open in figure viewerPowerPoint (top) SST from MODIS and surface geostrophic velocity vectors from AVISO for 24 July 2008. (middle) Geostrophic advection of relative vorticity, an approximation of the thermal wind vorticity advection term in equation (2). Downward motion ensues where this term is positive, while upward motion results otherwise. (bottom) Forward and backward FSLEs (day−1) for 2 July 2008. Synthetic floats were released on 24 June 2008 as circular magenta patches in three nearby regions with different dynamical characteristics (see text). Their final positions are plotted in black. [24] Synthetic floats released inside the eddies exhibit some distortion typical of eddy propagation and interaction with the surrounding flow but remain coherent structures. The patch released in between the eddies, on the other hand, is thinned in the zonal direction, a consequence of the confluent tendency along the stable manifold and the stretching tendency along the unstable manifold. As seen in Figure 4, the flow field is composed of a number of hyperbolic points, where unstable and stable manifolds cross each other, which would have a similar impact on passive tracer distribution. Horizontal Stirring and Vertical Motion [25] While the spatial pattern of a bloom, once created, can be explained to a large extent by horizontal stirring, the question still remain as to how blooms are triggered. The inorganic nutrients that allow blooms to form could be (1) vertically transported from the top of the nutricline, (2) horizontally transported from sediments associated with the landmass or regions with higher nutrient concentrations, or (3) deposited from the atmosphere. In the first case, the dynamical connection between horizontal stirring and vertical motion occurs via surface frontogenesis. As horizontal gradients are sharpened by the straining of the eddying flow, the thermal wind balance is disrupted. Vertical motion ensues in order to restore the balance. The link between horizontal deformation and vertical motion is clearly seen in the traditional form of the quasi-geostrophic omega equation [Holton, 1992] where N is the buoyancy frequency, f0 is the Coriolis frequency, w the vertical velocity, vg = (ug,vg) the geostrophic velocity vector, ρ′ is the density anomaly and ∇h is the horizontal gradient operator. Equation (1) shows that the quasi-geostrophic forcing of vertical motion is a sum of two terms related to the effect of the horizontal geostrophic flow on density and vorticity gradients. While equation (1) seems like a convenient way to estimate the sense of vertical circulations, its major drawback is the fact that the two terms tend to cancel each other [Tintoré et al., 1991; Trenberth, 1978]. In order to overcome that Trenberth [1978] approximated the right-hand side of equation (1) as [26] In general, the dominant forcing of vertical motion in equation (2) is the term associated with the advection of the geostrophic relative vorticity by the thermal wind [Pallàs-Sanz and Viúdez, 2005; Trenberth, 1978]. If we assume that the vertical shear of the geostrophic current is parallel to surface current, this term is further simplified as vhg · ∇hζg. Downward motion ensues if vhg · ∇hζg > 0, while upward motion results otherwise. [27] A snapshot of this term, using the geostrophic velocities from AVISO for 24 July 2008, is shown in Figure 4 (middle). Regions of alternating positive and negative vertical motion are generated primarily at the edges of relatively strong mesoscale features. It is important to note, however, that the usually discarded deformation term in equation (2) has been shown to be important in regions where there is a superposition of horizontal variations of the geostrophic flow and horizontal gradients in density [Martin, 1998]. Thus, Figure 4 (middle) would be somewhat modified in regions where the deformation term is large. The FSLEs provide information about the deformation of the horizontal geostrophic flow. Although regions of large stretching are ubiquitous (see Figure 4, bottom and also Figure 11), they would only be frontogenetical or frontolytical, hence prone to vertical motion as forced by the deformation term in equation (2), when approximately perpendicular to the direction of the density gradient (i.e., when horizontal stretching enhances or diminishes existing density gradients). 4. Results and Discussion The 2008 Bloom [28] From June to October 2008, meso- and submesoscale patches of high ((O)(0.1–0.2 mg chl-a m−3)) surface chlorophyll-a concentration appeared sporadically around the Hawaiian Islands in the region where summertime blooms are typically observed (see Figure 3). These patches occurred between the longitudes of 160°W and 135°W, at two different latitudes: near Station ALOHA (22.75°N), including the lee of the smaller islands, and near the southern limb of the North Pacific STF, which is around 30°N at this time of the year. [29] Because frontal activity is an important indicator of the physical supply of nutrients, it is worth investigating the spatial and temporal pattern of SST variability both at the large scale and mesoscale. The seasonal evolution of the SST for 2008 (Figure 5) elucidates some of the regional differences. The 25°C isotherm is displaced northward from 24°N to between 30 and 35°N in the western part of the region (west of 150°W) while its northernmost latitude east of 150°W is, approximately, 20°N. Figure 5Open in figure viewerPowerPoint Seasonal climatology of SST in the NPSG for 2008. Black contour shows the evolution of the 25°C isotherm. Color bar was chosen to highlight SST evolution in the central portion of the North Pacific. Data source: MODIS SST at 4 km resolution. The rectangles in Figure 5a represent the LEE (magenta), ALOHA (green), and NORTHEAST (white) regions as defined in the text and used in Figure 7. [30] Given the horizontal gradients in nutrient concentrations in the STF, with higher values in the colder water to the north, it is expected that horizontal advection might be as important as vertical processes in supplying a larger background nutrient concentration in the frontal area. During the spring and summer the front is oriented approximately northwest to southeast from 150°W to 135°W. This is the region northeast of Hawaii where blooms are common [Wilson et al., 2008]. The SST characteristics in th