Information on marine CO2 system variability has been
limited along the northeast Pacific Inside Passage despite the region's rich
biodiversity, abundant fisheries, and developing aquaculture industry.
Beginning in 2017, the Alaska Marine Highway System M/V Columbia has served as a
platform for surface underway data collection while conducting twice weekly
∼1600 km transits between Bellingham, Washington, and Skagway,
Alaska. Marine CO2 system patterns were evaluated using measurements
made over a 2-year period, which revealed the seasonal cycle as the dominant
mode of temporal variability. The amplitude of this signal varied spatially
and was modulated by the relative influences of tidal mixing, net community
production, and the magnitude and character of freshwater input. Surface
water pHT (total hydrogen ion scale) and aragonite saturation state
(Ωarag) were determined using carbon dioxide partial pressure (pCO2) data with
alkalinity derived from a regional salinity-based relationship, which was
evaluated using intervals of discrete seawater samples and underway pH
measurements. High-pCO2, low-pHT, and corrosive Ωarag
conditions (Ωarag<1) were seen during winter and
within persistent tidal mixing zones, and corrosive Ωarag
values were also seen in areas that receive significant glacial melt in
summer. Biophysical drivers are shown to dominate pCO2 variability over
most of the Inside Passage except in areas highly impacted by glacial melt.
pHT and Ωarag extremes were also characterized based on
degrees of variability and severity, and regional differences were evident.
Computations of the time of detection identified tidal mixing zones as
strategic observing sites with relatively short time spans required to
capture secular trends in seawater pCO2 equivalent to the contemporary
rise in atmospheric CO2. Finally, estimates of anthropogenic CO2
showed notable spatiotemporal variability. Changes in total hydrogen ion
content ([H+]T), pHT, and Ωarag over the
industrial era and to an atmospheric pCO2 level consistent with a
1.5 ∘C warmer climate were theoretically evaluated. These
calculations revealed greater absolute changes in [H+]T and
pHT in winter as opposed to larger Ωarag change in summer.
The contemporary acidification signal everywhere along the Inside Passage
exceeded the global average, with specific areas, namely Johnstone Strait
and the Salish Sea, standing out as potential bellwethers for the emergence
of biological ocean acidification (OA) impacts. Nearly half of the contemporary acidification
signal is expected over the coming 15 years, with an atmospheric CO2
trajectory that continues to be shaped by fossil–fuel development.
Introduction
Atmospheric carbon dioxide (CO2) has increased over the industrial era
from 278 ppm in 1765 to 414 ppm in 2020 due to the emissions of CO2
from fossil fuel combustion and land use change, which combined have
mobilized a total of 690 ± 80 Gt of carbon (Friedlingstein et al.,
2021). So far over the industrial era, an estimated 180 ± 35 Gt of this
carbon pool has transferred into the ocean (Friedlingstein et al., 2021),
known as the oceanic anthropogenic CO2 component
(Sabine et al., 2004), and led to changes in the
marine CO2 system, including reduced carbonate ion content
([CO32-]) and pHT (total hydrogen ion scale) and increased
total hydrogen ion content ([H+]T) and CO2 partial pressure
(pCO2). These marine CO2 system changes are collectively referred
to as “ocean acidification” (Caldeira and Wickett, 2003; Doney et al.,
2009; Feely et al., 2004a, 2009), and two recent assessments
estimate an average pHT decline for the global surface ocean on the
order of 0.1 units over the industrial era (Jiang
et al., 2019; Lauvset et al., 2020). In conjunction with this pHT
decline, reductions in [CO32-] have simultaneously decreased the
saturation states (Ω) of carbonate biominerals, with aragonite as
the most soluble carbonate biomineral typically targeted in biological
studies investigating the effects of ocean acidification (OA). Ωarag is a ratio of the product of [CO32-] and calcium
content over the solubility product for aragonite, and this ratio dictates
the thermodynamic favorability of aragonite precipitation. If Ωarag is >1, precipitation is favored over dissolution.
Globally, average surface Ωarag is estimated to have declined
by 0.53 units (Lauvset et al., 2020). These assessments of
global average pHT and Ωarag decline over the industrial
era are based on calculations of anthropogenic CO2 content; however,
long-term change in both pHT and Ωarag resulting from
anthropogenic CO2 input is captured in multidecadal open-ocean time
series datasets (Bates et al., 2014; Doney et al., 2020; Franco et al.,
2021).
Along the continental margins, seawater conditions may not track the global
average surface ocean pHT decline, particularly in the northeast
Pacific where seawater is less buffered than in some other ocean regions,
thereby making it more sensitive to increasing anthropogenic CO2 (Feely et al., 2008, 2018; Lauvset et al., 2020; Cai et al.,
2020; Jiang et al., 2015). One estimate of pHT decline on this margin
suggests twice the global average based on fossil foraminifera shells
preserved in marine sediments (Osborne et al., 2020). Even
with the potential for larger pHT decline along the northeast Pacific
margin, putting this change into context can be challenging. Given that
pHT is negative log10 of [H+]T, the absolute change in
[H+]T varies based on the initial pHT for the same degree of
pHT decline (Fassbender et al., 2021, 2017). For example, a 0.1-unit pHT decrease with an initial
pHT of 7.6 will result in an absolute [H+]T change of 6.3 nmol kg-1, whereas the same degree of pHT decrease with an initial
pHT of 8.4 will drive a 1 nmol kg-1 [H+]T change. This
clarification is important because the absolute change in acidity can be
different despite the same relative change in pHT, and confusion may be
enhanced when considering that some continental margins likely have
experienced different relative pHT change compared to the global
surface ocean average (Osborne et al., 2020; Evans et al., 2019; Pacella et
al., 2018; Salisbury and Jönsson, 2018), including in some coastal
ecosystems currently being evaluated for their OA mitigation potential
(Ricart et al., 2021; Kroeker et al., 2021).
The magnitude of OA-driven marine CO2 system changes and an
ecosystem's mitigation potential are both critical areas of research because
negative impacts are already being felt by some vulnerable marine species.
Along the northeast Pacific continental margin, larval shellfish mortality
within hatcheries has been tied directly to low Ωarag (Barton et al., 2012), and some adaptation measures to avoid such
conditions have been developed (Barton et al., 2015). Other
shell-forming marine species in this region are also exhibiting impacts from
OA, including Dungeness crab (Bednarsek et al., 2020; Berger et al., 2021)
and pteropods (Bednarsek et al., 2017, 2021; Mekkes et
al., 2021). The general consensus is that calcifying species may be the most
directly impacted (Kroeker et al., 2013; Haigh et al., 2015; Marshall et
al., 2017), although sensitivity to OA appears to be very species, life
stage, and population specific (Doney et al., 2020) with the
potential for compensatory mechanisms to help sustain populations
(Peck et al., 2018; Bednarsek et al., 2021).
However, there is a high likelihood of enhanced vulnerabilities by other
co-occurring stressors like warming (Kroeker et al., 2013)
and reduced oxygen content (Gobler and Baumann, 2016). Biological
stressors, such as viral pathogens and harmful algal species, may also
become more prevalent or virulent in association with changes in marine
CO2 chemistry and warming (Raven et al., 2020; Asplund et al., 2013).
The sum of both the direct and indirect effects from OA and other
co-stressors threatens marine food webs (Jin et al., 2020),
harvested species (Ekstrom et al., 2015), and dependent
coastal communities (Mathis et al., 2015);
understanding this threat demands assessing how the marine CO2 system
has and will evolve through time.
Determining long-term trends in coastal settings is difficult because of
inherent high variability resulting from a number of processes unique to the
land–ocean interface. Physical forcing from upwelling-favorable winds or
tide-induced vertical mixing can result in surface water pCO2 that is
super-saturated with respect to the atmosphere, whereas high rates of
primary production draw down surface water pCO2 to well below
atmospheric levels. Additionally, freshwater input from land can act to
dilute total dissolved inorganic carbon (TCO2) and total alkalinity and
reduce pCO2 (Meire et al., 2015), or, alternatively,
increase pCO2 through the respiration of riverine organic matter
(Ward et al., 2017). These processes all occur on
different timescales, are not uniformly important across coastal settings,
and collectively act to make resolving the relatively small anthropogenic
CO2 signal difficult to disentangle from environmental variability.
Resolving environmental variability, even to the point of capturing seasonal
cycles, remains a challenge in many settings due to a lack of measurements
(Hales et al., 2008). In the northeast Pacific between British
Columbia (BC) and southeast Alaska (AK), modeling efforts have aided in
addressing this knowledge gap and have indicated the relative significance
of freshwater input (Siedlecki et al., 2017; Hauri
et al., 2020) and its source character (Pilcher et al.,
2016), as well as projected warming, deoxygenation, and acidification on
multidecadal timescales (Holdsworth et al., 2021). However,
observations remain essential to evaluate model output and confirm our
understanding of the governing processes that shape marine CO2 system
variability, particularly in nearshore settings.
Study region
The Inside Passage is a roughly 1600 km network of coastal waterways that
spans the semi-enclosed Salish Sea, the central and northern BC coast, and
southeast AK. This nearshore region is a key interface between the Pacific
coastal temperate rainforest (O'Neel et al., 2015; Bidlack et al., 2021)
and a highly productive continental shelf ecosystem (Ware and
Thomson, 2005; Jackson et al., 2015). The area experiences a wide array of
physical and biogeochemical drivers including intense tidal currents within
narrow passages that induce persistent vertical mixing
(Whitney et al., 2005; Dosser et al., 2021), strong
autumn and winter storms (Stabeno et al., 2004), high runoff
from rainfall and snowmelt and glacial-melt sources (Morrison et al.,
2012; Beamer et al., 2016; Edwards et al., 2020; Neal et al., 2010), high
terrestrial organic carbon input (Edwards et al., 2020; Oliver et al.,
2017; St. Pierre et al., 2021), and remotely forced influences such as El
Niño events and marine heat waves (Bond et al.,
2015; Jackson et al., 2018).
The coastal ocean from BC to southeast AK has large under-sampled areas
(Hales et al., 2008; Evans and Mathis, 2013) and coarse temporal
information on marine CO2 system variability based on direct
measurements (Evans and Mathis, 2013; Evans et al., 2012; Tortell et al.,
2012) except within the Salish Sea where seasonal and spatial patterns are
more constrained (Evans et al., 2019; Cai et al., 2021; Ianson et al.,
2016; Fassbender et al., 2018a; Feely et al., 2010; Lowe et al., 2019). We
reduced this information gap by outfitting a passenger ferry within the
Alaska Marine Highway System (AMHS) fleet, the M/V Columbia, with instrumentation to
monitor surface ocean conditions along the Inside Passage (Fig. 1). We
report here on surface underway measurements made from November 2017 to
October 2019, and we use this dataset to describe marine CO2 system
patterns and quantify the relative importance of key drivers in shaping the
observed variability. We also evaluate marine CO2 system extremes and
their timing along the ferry transit, which likely has implications for the
exposure histories of vulnerable species. Finally, we estimate the
anthropogenic CO2 content accrued over the industrial era; assess the
impact this perturbation has had on [H+]T, pHT, and Ωarag; and theoretically gauge the extent of acidification implied by
the Paris Agreement (UNFCC, 2015) to limit global warming to
preferably 1.5 ∘C relative to pre-industrial levels, what we refer
to as the “1.5 ∘C acidification level”. The so-called
“remaining 1.5 ∘C carbon budget” translates to an atmospheric
CO2 level which would be reached with all potential mitigation pathways
(Rogelj et al., 2018) and
therefore can be viewed as the best case scenario for the maximum
acidification owing to anthropogenic CO2 input.
Surface salinity along the Inside Passage expressed as the
coefficient of variation (CV; %; a) computed for 0.03∘×0.03∘ monthly grid cells from underway measurements made from the
M/V Columbia between 3 November 2017 and 2 October 2019. Areas of the highest salinity
CV are due to large freshwater input, and black circles with labels mark the
Alaska Marine Highway System terminals: Bellingham, WA (B); Ketchikan, AK
(K); Wrangell, AK (W); Petersburg, AK (P); Juneau, AK (J); Haines, AK (H);
Skagway, AK (Sk); and Sitka, AK (Si). Also shown is the surface chlorophyll
CV (%; b) from Moderate Resolution Imaging Spectroradiometer (MODIS)
Level 3, 4 km mapped data from February to October 2018 and 2019. Areas of
high chlorophyll CV reflect instances of biomass accumulation presumably
owing to periods of high primary productivity. Areas labeled in the figure
panel are the states of Washington (WA) and Alaska (AK), the province of
British Columbia (BC), Lynn Canal (LC), Sergius Narrows (SN), Wrangell
Narrows (WN), Dixon Entrance (DE), Johnstone Strait (JS), Vancouver Island
(VI), and the Salish Sea (SS).
MethodsUnderway instrumentation
The AMHS M/V Columbia transited the 1600 km Inside Passage on a weekly basis
(Sect. S1). Surface (∼2 m) seawater pCO2
data were obtained from measurements of CO2 mixing ratio (xCO2)
made using a General Oceanics 8050 (GO8050) pCO2 measuring system
following recommended protocols (Pierrot et al., 2009). Seawater
was drawn into the M/V Columbia through an intake located in the bow thruster engine
room and supplied to the GO8050 and ancillary sensors using a 12
HP self-priming centrifugal pump (AMT 429A-98 or similar) located
∼2 m from the seawater intake. Temperature measurements were
made at the seawater intake using an SBE 38 digital oceanographic
thermometer with an accuracy reported by Sea-Bird Electronics of
0.001 ∘C. Seawater was then circulated from the bow thruster room
up one deck to the car deck and then aft approximately 30 m along the
starboard side to where the GO8050 and ancillary sensors were installed. The
seawater circulation loop was split at this location between the GO8050 wet
box and an ancillary sensor loop consisting of an SBE 45 MicroTSG
thermosalinograph and an Aanderaa 4330F oxygen optode. The accuracy of the
temperature measurement from the SBE 45 was reported as 0.001 ∘C
when interfaced with the SBE 38, and the accuracy of the conductivity
measurement was 0.0003 S m-1. Salinity, computed from conductivity and
temperature, is reported here on the Practical Salinity Scale (PSS-78). The
accuracy of the Aanderaa 4330F oxygen optode reported by the manufacturer
Xylem was <1.5 %. All ancillary sensors were serviced annually,
and a multipoint calibration was conducted on the oxygen optode at the
Aanderaa facility in Norway. Oxygen data from the Aanderaa 4330F were output
in µmol L-1, salinity-corrected using the approach described in
Bittig et al. (2018), and then density-corrected to µmol kg-1. Oxygen data are reported here as the difference from
saturated values (ΔO2).
Seawater entered the GO8050 wet box at ∼2.9 L min-1 and
∼10 psi and then was circulated into a water-jacketed (to
minimize warming) primary showerhead equilibrator with a liquid volume of
∼0.5 L and to a smaller secondary equilibrator with a liquid
volume of ∼0.1 L. The primary equilibrator was maintained at
ambient pressure on the car deck by a vent that was plumbed to the secondary
equilibrator and then to the primary equilibrator. The pressure difference
between inside the primary equilibrator and the car deck was monitored using
a Setra pressure transducer (model 239) with a 0.15 hPa uncertainty. The
secondary equilibrator serves to pre-equilibrate (make-up) air entering the
primary equilibrator through the vent due to any loss through the headspace
gas recirculation loop. A flow meter was present at the opening of the vent
in order to monitor make-up air flow into the equilibrator. The headspace
gas volume of the primary equilibrator was ∼0.8 L, and
seawater temperature was monitored within the primary equilibrator using a
Fluke thermometer (model 1523) and thermistor probe (model 5610) with an
uncertainty of 0.01 ∘C.
Atmospheric air was drawn from an intake on the foredeck to the GO8050 wet
box. Both the equilibrator headspace gas and atmospheric air were dried
using a condenser (Peltier thermoelectric cooling device) and Permapure
Nafion drying tubes in order to minimize the correction for water vapor
content associated with band-broadening within the infrared gas analyzer
located in the GO8050 dry box. The analyzer housed in the GO8050 dry box was
a LI-COR LI840A CO2–H2O gas analyzer with a root-mean-square noise
level for 1 Hz measurements of 1 ppm reported by the manufacturer. Dried
equilibrator headspace and atmospheric gases were supplied to the analyzer
from the wet box at ∼0.1 L min-1. In addition to the
analysis of equilibrator headspace and atmospheric gases, four standard
gases of known mixing ratio (150, 349, 449, and 850 ppm; Praxair) were also
plumbed to provide gas flow to the GO8050. Praxair standard gases were
evaluated by calibrating a LI840A using the Praxair gases, and then the
calibrated analyzer was used to measure the CO2 content of a World
Meteorological Organization (WMO) traceable standard gas cylinder from the
National Oceanic and Atmospheric Administration Earth System Research
Laboratories (ESRL) Greenhouse Gas Global Reference Network. The Praxair gas
standard calibrated LI840 was able to reproduce the certified ESRL standard
to within 0.1 %.
The GO8050 was controlled using National Instruments LabVIEW software run on
a PC laptop computer. The software controls data acquisition from the
GO8050, an interface box connecting the SBE 38 and the SBE 45, the Aanderaa
4330F, the primary equilibrator temperature and pressure sensors, a Vaisala
digital barometer (0.07 mbar accuracy) with a model 61002 Gill pressure port
and GPS antenna positioned adjacent to the atmospheric air intake, and the
LI840A, while also controlling a Valco Instruments Co. Inc. (VICI)
multi-port actuator that cycles between the gas streams plumbed to the dry
box. The software captured measurements from all ancillary sensors as well
as analyses of the four gas standards of known CO2 content, 12
measurements of atmospheric CO2, and 240 seawater CO2 measurements
in a cycle that was repeated every 8.5 h with a 2 min measurement
frequency. The seawater and atmospheric CO2 analyses were run in a
sequence of three atmospheric measurements and 60 seawater measurements that was
repeated four times between standardization. Analyses of each gas standard were
interpolated over the time record of the dataset and used to create
calibration functions needed to correct the raw LI840A xCO2 data.
Calibrated seawater xCO2 data in dry air were quality-controlled and
then converted to CO2 partial pressure (pCO2) in wet air at
the equilibrator temperature by using atmospheric pressure measured by the
LI840A plus the differential pressure recorded in the equilibrator corrected
for the removal of water vapor. Finally, seawater pCO2 in wet air was
adjusted to sea surface temperature using the offset between SBE 45
temperature recorded at the GO8050 and intake temperature from the SBE 38
located at the seawater intake with zero lag (0.3 ∘C ± 0.17 ∘C), as the lag between the temperature measurements at these
two locations was determined to be less than the measurement frequency.
Total uncertainty in our pCO2 measurements is the combined
uncertainties from calibration, equilibrator temperature, equilibrator
pressure, and the warming correction added in quadrature. At contemporary
atmospheric pCO2 levels near 400 µatm, these component
uncertainties would equate to 0.4, 0.06, 0.17, and 3.9 µatm. Typically,
underway pCO2 measurement uncertainties are reported as a function of
uncertainties in the equilibrator temperature and pressure and the water
vapor pressure (Wanninkhof et al., 2013). Considering our dried
gas stream that minimizes uncertainty from water vapor pressure, the
uncertainty from just the equilibrator temperature and pressure, with
inclusion of the calibration uncertainty, would equal 0.44 µatm.
However, taking into account uncertainty in the warming correction (while
still not addressing deviations from a constant pCO2 temperature
sensitivity) increases the pCO2 uncertainty to 3.92 µatm. While
we prefer the more typical assessment that points to a lower pCO2
uncertainty, we use a conservative ∼1 % pCO2
uncertainty below to estimate the uncertainties in derived marine CO2
system parameters.
In June 2019, a BioGeoChemical SUrface MOnitoring-system (BGC-SUMO) was
configured with the GO8050 to provide underway pH measurements on the total
hydrogen ion scale (pHT). The BGC-SUMO measures pHT, temperature,
and nitrate concentration, although the latter measurement was not
successful on this vessel. The pHT was measured using a Deep-Sea
DuraFET, consisting of an ion-sensitive field effect transistor (ISFET) and
a chloride ion-selective electrode as the reference
(Johnson et al., 2016). Thus, seawater is unmodified, and no
chemicals are added as it flows through the BGC-SUMO. The pHT sensor
was calibrated prior to deployment on the M/V Columbia, and its performance was
verified based on discrete samples taken alongside the sensor (n=9)
throughout the deployment (Takeshita et al., 2018). Based on this
comparison, we assume an uncertainty in pHT of 0.01. Maintenance on all
instrumentation configured aboard the M/V Columbia was conducted during service
stops in Ketchikan to prevent biofouling.
Discrete sample collection
Discrete seawater samples were collected on two ferry trips in November 2017
and August 2018. Samples were drawn from the seawater supply line
immediately upstream of the GO8050 into rinsed 350 mL amber soda-lime glass
bottles and analyzed for TCO2 and pCO2 within a month of
collection following methods described elsewhere (Evans
et al., 2019). Briefly, TCO2 and pCO2 were analyzed from the same
sample bottle in this order at the Hakai Institute's Quadra Island Field
Station using a Burke-o-Lator pCO2–TCO2 analyzer. The TCO2
measurement was achieved by acidification and gas stripping followed by
non-dispersive infrared detection using a LI-COR LI840A and consumed
∼60 mL of sample. TCO2 measurements were adjusted using
correction factors developed through the analysis of certified reference
materials (CRMs) from Andrew Dickson (Scripps Institute of Oceanography),
with typical correction factors between 0.99 and 1.01. Uncertainty in the
discrete TCO2 measurement was determined to be 0.3 % (Evans et al., 2019). The pCO2 measurement was achieved by headspace gas recirculation
between the LI840A and the sample bottle in a closed loop until
equilibration of the headspace gas with the seawater sample pCO2 was
obtained (roughly 6 min). Uncertainty in the discrete pCO2
measurement was determined to be 1 %. The TCO2 measurement was
subsequently headspace gas corrected (Wanninkhof and Thoning, 1993),
and then alkalinity (Alk) was computed using the pCO2 and head space
gas-corrected TCO2 data with a MATLAB version of CO2SYS
(Sharp et al., 2021) and the carbonic acid
dissociation constants of Waters et al. (2014), bisulfate
dissociation constant of Dickson et al. (1990), fluoride and
hydrogen association constants from Perez and Fraga (1987), and
boron / chlorinity ratio of Uppström (1974). Alk computed in this
way excluded contributions from organic acids, phosphate, and silicate.
CalculationsGap filling, marine CO2 calculations, and gridding
The record of underway measurements from the M/V Columbia contained a number of data
gaps related to service interruptions, the largest of which was between
October 2018 and March 2019 when the ferry went into dry dock for the
winter. However, there was a period from 25 August to 2 October 2019 when
only the direct pCO2 measurements were compromised due to an issue with
the LI-COR. Subsequently, pCO2 was estimated indirectly using pHT
measurements and a regional Alk–salinity relationship
(Evans et al., 2015). To fill these missing data,
pHT measurements were interpolated to the measurement time of the
GO8050. pCO2 was then computed using the time-matched pHT data
with the relationships described above and derived Alk. Missing measured
pCO2 observations in late 2019 were filled with the computed values.
Seawater pHT, [H+]T, and Ωarag were computed for
the entire dataset using the salinity, intake temperature, the gap-filled
pCO2 record, and Alk derived from salinity (Evans
et al., 2015) with the dissociation constants and relationships described
above using a MATLAB version of CO2SYS (Sharp et
al., 2021). Uncertainty in pHT, [H+]T, and Ωarag
derived from our pCO2 record coupled with salinity-based Alk
determinations was assessed using the error propagation routine from
Orr et al. (2018) updated in the most recent MATLAB version of
CO2SYS (Sharp et al., 2021). Combined standard
uncertainties for pHT, [H+]T, and Ωarag were
computed using the previously described 1 % pCO2 uncertainty, the
reported 17.21 µmol kg-1 uncertainty in the regional
Alk–salinity relationship (Evans et al., 2015), and
the default uncertainties for the dissociation constants within the error
propagation routine. pHT, [H+]T, and Ωarag
uncertainties were computed across the range of observed Alk values and with
pCO2 computed across a range TCO2:Alk ratios spanning 0.85 to 1
for each corresponding Alk value (Fig. S1). These
calculations were done at a constant temperature and with salinity ranging
from 10 to 32 corresponding with the range of Alk values. Importantly,
marine CO2 system data quality falls into two objectives as defined by
the Global Ocean Acidification Observing Network (Newton et al.,
2015; Tilbrook et al., 2019): (1) weather and (2) climate. The weather data
quality objective is thought sufficient for identifying spatial and temporal
patterns excluding long-term trends, which is considered more appropriate
for data reaching the stringent climate quality objective to assess. The
mean pHT, [H+]T, and Ωarag uncertainties from
our calculations are 0.01, 0.23 nmol kg-1, and 0.07 respectively. These
values meet the Global Ocean Acidification Observing Network weather data
quality objective. However, we note that uncertainties vary across the range
of values considered. For instance, Ωarag uncertainty is higher
at higher Ωarag values, whereas pHT uncertainty is higher
at lower salinity and Alk (Fig. S1).
To evaluate basic statistics along the M/V Columbia transit, including means,
coefficients of variation (CV), and lower 5th percentiles, as well as
assess seasonal drivers and the time of detection that are both described
below, observations were gridded by isolating and averaging data within
0.03∘ by 0.03∘ grid cells. It is important to note that
due to the 2018–2019 winter data gap, gridded averages likely over-represent
spring and summer relative to the autumn and winter conditions. This grid
size equalled roughly 6 km2 across the latitudinal range of the M/V
Columbia transit. Analyses using gridded data were only conducted on grid cells
containing more than 40 measurements.
Seasonal drivers
Assessing the drivers of seasonal pCO2 variations requires isolating
the thermodynamic and biophysical components of the variability. The process
to achieve this is described in Sect. S2, and results in
isolating the pCO2 temperature component (pCO2T component), the
pCO2 salinity component (pCO2S component), and remaining
variability from biophysical drivers. Seasonal amplitudes of each component
of pCO2 variability were assessed, and the ratio of the amplitude of
thermodynamic (T, S, or combined TS) to biophysical drivers (BT,
BS, BTS; where subscript denotes the removed terms) defines which
is more important for determining pCO2 variability on an annual basis
(Takahashi et al., 2002; Fassbender et al., 2018b).
Severity and time of detection
We determined the severity of derived pHT and Ωarag in
each grid cell based on the lower 5th percentile as in
Chan et al. (2017) and the timing of severe
conditions as the mode of all months of observations less than or equal to
the lower 5th percentile of each grid cell. We also assessed the
time of detection (ToD) within each grid cell of the M/V Columbia transit in order
to guide future observational efforts targeting the identification of
long-term change. ToD is similar to the time of emergence used in climate
studies (Henson et al., 2017) with the exception that it
includes measurement uncertainty (Carter et al., 2019b).
Both of these terms represent the time required for a secular trend, in our
case increasing seawater pCO2 from anthropogenic CO2 uptake, to
emerge from the “noise” in an environmental dataset. Monthly mean
pCO2 is computed from the observations occurring within each grid cell,
and then the observations are differenced from the monthly mean in order to
compute de-seasonalized anomalies (i.e., removing the large-amplitude seasonal
cycle from the noise). The standard deviation of the de-seasonalized
anomalies was combined in quadrature with the pCO2 measurement
uncertainty to represent the remaining environmental noise and compute
ToD as
ToD=2×noisepCO2growth rate,
where the pCO2 growth rate used here was 2.5 µatm yr-1 and
is approximately the average of annual values from the National Oceanic and
Atmospheric Administration ESRL over the 2014–2019 period. Importantly, we
consider ToD as a guiding metric. The growth of seawater pCO2 can vary
across coastal settings and may or may not be entirely driven by
anthropogenic CO2 input (Laruelle et al.,
2018; Salisbury and Jönsson, 2018). For example, changes in nutrient
input from runoff can alter the pCO2 growth rate from an expected
anthropogenic CO2-driven signal (Turk et al., 2019).
Therefore, we present ToD only to discuss how these data might be used to
target observing efforts and not as absolute values.
Anthropogenic CO2
Anthropogenic CO2 content was estimated using the ΔTCO2 approach (Takeshita et al., 2015; Pacella et al.,
2018; Evans et al., 2019), which is a simplification of the ΔC∗ method (Sabine et al., 2002; Gruber et al.,
1996), and assumes a constant TCO2 disequilibrium with the atmosphere
defined as
ΔTCO2,diseq=TCO2,obs-TCO2atmpCO2,currentyear-age,Alkder,Tobs,Sobs,
where
TCO2,obs is the observed TCO2 and
TCO2atmpCO2,current year-age,Alkder,Tobs,Sobs
is the TCO2 content that would result from equilibration with
the atmospheric pCO2 at the time of last contact with the atmosphere
(current year minus the age of the water mass), at the derived Alk and at
the observed temperature and salinity. The time of last contact with the
atmosphere represents the age of a water mass in years and is zero for most
surface measurements except in areas where deep water is mixed to the
surface. Water mass age was estimated by dividing the measured apparent
oxygen utilization (AOU, or the inverse of ΔO2) by the oxygen utilization rate (OUR). A value of
4.1 µmol kg-1 yr-1 for OUR was taken from the literature
for Pacific subarctic upper water (Feely et al., 2004b) and used in
this calculation. Using the ΔTCO2,diseq term and assuming patterns in derived
Alk and observed temperature and salinity are largely invariant, the
TCO2 for a given year can be estimated by
TCO2,year=TCO2atmpCO2,year-age,Alkder,Tobs,Sobs+ΔTCO2,diseq,
where
TCO2,year is the TCO2 content for a
specific year and TCO2atmpCO2,year-age,Alkder,Tobs,Sobs
is the TCO2 content that would be realized if that surface
water mass were in equilibrium with the atmospheric pCO2 that occurred
during a given year, corrected for the age of the water mass, and at the
contemporary derived Alk and observed temperature and salinity. The
anthropogenic CO2 content is then determined as the difference between
the TCO2 for a given year and the TCO2 content for the year 1765.
Historical atmospheric CO2 mol fractions based on observations and
projected atmospheric CO2 for the shared socio-economic pathways (SSPs)
were obtained from Meinshausen et
al. (2020) using their data portal (http://greenhousegases.science.unimelb.edu.au, last access: May 2020) and converted to pCO2
assuming standard atmospheric pressure. pHT, [H+]T, and
Ωarag were computed for each year from 1765 onward using the
TCO2 estimated for a given year with the modern derived Alk and
observed temperature and salinity. It is important to acknowledge that
uncertainty in estimating anthropogenic CO2 content using this approach
is at least 5 µmol kg-1 based on similarities with the
ΔC∗ method
(Sabine et al., 2002). Uncertainty stems from a number of
sources, including the key assumptions of constant TCO2 disequilibria
and unchanged variation in the natural carbon cycle, temperature, and
salinity. Inadequacies in these assumptions can lead to biases in
anthropogenic CO2 (Matsumoto and Gruber, 2005), which in turn
influences estimations of past and future pHT, [H+]T, and
Ωarag.
Results and discussionTime and space variability
Over 244 000 seawater temperature, salinity, O2, and pCO2
measurements were made on the M/V Columbia during 135 north- and south-bound
transits of the Inside Passage over a 2-year period. These data revealed
substantial spatiotemporal variability in surface seawater conditions along
this 1600 km stretch of coastline. The spatial and temporal mosaic captured
by these measurements (Figs. 2 and 3) portrays two key features of the
Inside Passage: (1) the dominant mode of temporal variability is the
seasonal cycle, and (2) there is regional variability in the seasonal cycle
amplitude that is modulated by the relative influences of tidal mixing, net
community production, and the magnitude and character of freshwater input.
Between November and March, cold seawater spanned the entire Inside Passage,
with the coldest water in southeast AK generally near 4 ∘C, but
0.5 ∘C was observed near Juneau. Seasonal warming in most regions
began in April and occurred earlier in the Salish Sea, which was consistent
with satellite observations that identified earlier seasonal warming in this
region relative to coastal areas to the north (Jackson et al.,
2015). Surface salinity was fresher throughout the year in the Salish Sea,
although variability in salinity was larger in southeast AK (Figs. 1 and
2) where seasonal freshwater delivery to the coastal ocean contributes
41 % of the freshwater input to the Gulf of Alaska (Edwards
et al., 2020). The combined estimates of discharge from each major watershed
along the Inside Passage from Edwards et al. (2020) and
Morrison et al. (2012) indicate that over 570 km3 yr-1
of freshwater enters the northeast Pacific from southeast AK, an amount that
exceeds the Mississippi River discharge (Dai and Trenberth, 2002). Along
the BC portion of the Inside Passage, discharge is near 390 km3 yr-1 with almost a quarter of this amount originating from the Fraser
River. Despite lower runoff from BC, its influence on salinity manifests
earlier than the peak freshwater input in southeast AK (Fig. 2) due to the
high contribution of snowmelt to the late spring and early summer discharge
(Morrison et al., 2012). In southeast AK, seasonal
reduction in salinity began in May and reached the summer minima in August.
Low-salinity conditions were uniform over an area that encompassed Lynn
Canal and the inside waterways around Juneau (Figs. 1 and 2). The late
summer minimum in salinity reflects the significant contribution of glacial
melt in the most northern portion of the Inside Passage
(Neal et al., 2010; Edwards et al., 2020). Seasonal
variation in temperature and salinity was reduced in some confined
waterways, such as Johnstone Strait (Fig. 1), owing to the influence of
intense tidal mixing in these areas that dampens the seasonal cycle
amplitude (Dosser et al., 2021; Whitney et al., 2005).
Despite the emergence of a marine heat wave in the North Pacific in late 2019
(Amaya et al., 2020), spring and summer patterns in temperature
and salinity appeared similar between 2018 and 2019.
Sea surface temperature (SST; ∘C; a) and
salinity (b) measured between 3 November 2017 and 2 October 2019. The x
and y axes represent longitude and latitude, respectively, and with the
coastline and terminal positions shown as in Fig. 1 and time increasing
along the z axis.
Across most of the Inside Passage, ΔO2 and pCO2 showed an
inverse relationship (Fig. 3). Where ΔO2 values were
positive, pCO2 was undersaturated with respect to the atmosphere, and
this combination likely reflects the influence of primary productivity
exceeding rates of organic matter respiration, i.e., positive net community
production (NCP). Abiotic changes in ΔO2 result from changes in
temperature and salinity as well as bubble injection and wave breaking
(Juranek et al., 2019), although we contend that the latter two
drivers may be of lesser importance in the protected Inside Passage
waterways. Seasonal warming increases both ΔO2 and pCO2.
However, we observed an increase in ΔO2 with a corresponding
decrease in pCO2, which strongly suggests that O2 supersaturation
and pCO2 drawdown resulted from positive NCP (Tortell et al.,
2012; Juranek et al., 2019). In areas outside of the influence of tidal
mixing, the signals of O2 supersaturation and pCO2 drawdown were
initiated in response to the spring phytoplankton bloom, and generally were
sustained through summer until the autumn storm season commenced (Evans et
al., 2019; Fassbender et al., 2018a). An exception was Lynn Canal in
southeast Alaska (Fig. 1) where the relationship between O2 and
pCO2 diverged in summer (Fig. 3) when the seasonal change in salinity
was maximal (Fig. 2). The addition of cold glacial meltwater results in
undersaturated surface pCO2 (Cai et al., 2021; Pilcher et al.,
2016; Evans et al., 2014) while also increasing oxygen solubility and
subsequently decreasing ΔO2 (Fig. 3). The diverging character
between O2 and pCO2 within Lynn Canal in summer dissipated during
autumn when salinity increased in response to storm-induced vertical mixing.
Autumn marked the transition back to supersaturated pCO2 with respect
to the atmosphere throughout the Inside Passage. Inter-annual variability
was apparent in this dataset during the spring and summer months, as 2019
had slightly greater O2 supersaturation and pCO2 drawdown during
spring in the Salish Sea and on the central BC coast and throughout much of
the summer in southeast AK (Fig. 3).
ΔO2 (µmol kg-1; a) and
pCO2 (µatm; b) measured between 3 November 2017 and
2 October 2019. The x and y axes represent longitude and latitude,
respectively, and with the coastline and terminal positions shown as in
Fig. 1 and time increasing along the z axis.
Seasonal O2 supersaturation and pCO2 drawdown do not occur
uniformly along the Inside Passage but in distinct regions separated by
areas of tidal mixing that support sustained low-O2 and high-pCO2
conditions (Fig. 3) due to the near-continuous ventilation of sub-surface
waters (Whitney et al., 2005; Dosser et al., 2021; Evans et al.,
2012; Tortell et al., 2012). The most obvious region of intense tidal mixing
along this coastline was in Johnstone Strait between Vancouver Island and
mainland BC (Fig. 1), but other areas were also evident, including in the
narrow waterway north of Sitka known as Sergius Narrows (Fig. 1). As
mentioned above, the seasonal amplitude in the tidal mixing zones is reduced
because the water column may be completely mixed, and seasonal variation in
these areas may more reflect that of sub-surface water entering the mixing
zone laterally (Dosser et al., 2021). Since seasonality in these areas is
potentially more influenced by sub-surface source waters, the seasonal cycle
can be out of phase with adjacent areas outside of the tidal mixing zones.
This was most obvious in Johnstone Strait, where high-pCO2 conditions
outside of this area were generally seen during winter, whereas within this
region, the highest pCO2 was in autumn. The highest observed seawater
pCO2 was near 1200 µatm in Johnstone Strait during September.
Winter pCO2 values outside of the tidal mixing zones broadly ranged
between 450 and 800 µatm, being higher in regions with less direct
connection to the open continental shelf, such as in the semi-enclosed
Salish Sea, in areas of the central BC coast, and in southeast AK (Fig. 3). These areas receive high amounts of riverine organic matter (St. Pierre et al., 2021; Oliver et al., 2017; Johannessen et al., 2003) that may
be confined to the nearshore zone by winter-time downwelling circulation
(Thomson, 1981; Weingartner et al., 2009) and there subsequently
remineralized by the microbial community (St. Pierre et al., 2020), leading to elevated surface pCO2 nearshore that is
not seen in the offshore data along this coast (Evans and Mathis,
2013). In tidal mixing zones like Johnstone Strait, the highest pCO2 in
early autumn decreased through winter to a minimum by late spring, albeit
with values that were still supersaturated with respect to atmospheric
pCO2. This difference in timing likely reflects the seasonality of
sub-surface waters (Dosser et al., 2021), since, without a
short residence time (Pawlowicz et al., 2007), these waters would
experience a build-up of respiratory CO2 through the growing season as
organic matter rains out of the surface layer and is respired at depth by
the microbial community. We suggest this sub-surface respiration signal is
ventilated in the tidal mixing zones and is responsible for the early
autumn peak in surface pCO2.
Seawater pHT and Ωarag variability was evaluated by
employing an Alk–salinity relationship developed from observations spanning
a large portion of the region (Evans et al., 2015). Validation of this
relationship was done using Alk determined from seawater samples collected
during ferry ride-along cruises and processed as described above. These
cruises occurred in November 2017 and August 2018 and spanned the dynamic
range of observed salinity conditions (Fig. 2). During November,
comparison between discrete Alk and salinity-derived Alk was within 2
times the root-mean-square error of the salinity-based relationship
(Fig. S2). During August, larger divergence between discrete
and salinity-derived Alk occurred in low-salinity water within the
northernmost portion of the Inside Passage. Specifically in the area of Lynn
Canal, Alk determined from the salinity-based relationship overpredicted
bottle-determined Alk by at most 200 µmol kg-1. Salinity-based
Alk determination was further evaluated in 2019 by comparing estimated
pHT, computed from directly measured pCO2 and salinity-based Alk,
to directly measured pHT (Fig. S3). pHT was
measured from June to October 2019 over the period of lowest observed
salinities in southeast AK and revealed a similar pattern to the discrete
Alk comparison. Divergence between estimated and directly measured pHT
was greatest in seawater with salinity <22 and north of
57∘ N in the region around Juneau and up Lynn Canal (Fig. S3). In the analysis that follows, we continue to use salinity-based
Alk with our gap-filled pCO2 record to determine components of the
marine CO2 system along the Inside Passage, but we acknowledge that the
northernmost region during summer is likely more corrosive for aragonite
than our analysis suggests because of our Alk over-predictions in low-salinity water. It is also possible that local deviations from the
assumption of proportionality between salinity and calcium made within
CO2SYS may counteract a portion of this “missing” corrosive signal in the
low-salinity surface water of southeast AK (Beckwith et al.,
2019). The presence of proton-binding dissolved organic molecules may cause
additional complication by impacting the interpretation of low-salinity Alk
measurements used to generate the regional Alk–salinity relationship
(Sharp and Byrne, 2020). While the magnitude of how corrosive Ωarag is during the melt season in the northernmost area of the Inside
Passage may be less well-constrained, the drivers and timing of adverse
conditions should not deviate from what we describe below. The confounding
factors of variable freshwater Alk, interpretations of Alk measurements in
the presence of proton-binding dissolved organic molecules, and
potential for non-zero calcium end-members in this region all demand
further study in order to more accurately assess the magnitude of corrosive
summer Ωarag conditions in these glacial-melt-influenced
waters.
pHT (total scale; a) and Ωarag
(b) derived from measurements made from 3 November 2017 to 2 October 2019. The x and y axes represent longitude and latitude, respectively, and with
the coastline and terminal positions shown as in Fig. 1 and time
increasing along the z axis.
Patterns in seawater pHT and Ωarag were largely the
inverse of that for pCO2 (Fig. 4); areas exhibiting pCO2
undersaturation with respect to the atmosphere typically co-occurred with
high-pHT and high-Ωarag conditions, whereas regions with high
pCO2 have low pHT and Ωarag. Areas with both high
pHT and Ωarag have experienced recent positive NCP that
would also support O2 supersaturation and pCO2 drawdown (Figs. 3
and 4). The evidence of inter-annual variability discussed above for spring
and summer O2 and pCO2 was apparent for pHT and Ωarag, with 2019 exhibiting more frequent occurrences of Ωarag>3 compared to 2018. pHT and Ωarag
were lowest in most areas during winter and year-round within tidal mixing
zones. Winter Ωarag values were <1 in all regions that
lacked direct connection to the open continental shelf, specifically within
the Salish Sea, Johnstone Strait, inside waterways on the central and
northern BC coast, and in southeast AK. Corrosive conditions for aragonite
persisted throughout the year in Johnstone Strait, and in an area known as
Wrangell Narrows between Wrangell and Petersburg (Fig. 1). In the
northernmost area of the Inside Passage, a short period of Ωarag conditions >1 occurred between March and June,
resulting from the spring phytoplankton bloom as evidenced by co-occurring
O2 supersaturation and pCO2 drawdown (Fig. 3). Once the summer
melt season commenced, the Inside Passage-wide minimum in Ωarag
was observed in this region despite the over-prediction in Alk in low-salinity water mentioned above. Lynn Canal exhibited the most corrosive
conditions for aragonite along the 1600 km M/V Columbia transit due to the large
contribution of meltwater in this region (Fig. 2). Such corrosive
conditions in glacial-melt-influenced settings have been reported previously
in AK (Reisdorph and Mathis, 2013; Evans et al., 2014) as well as
in Svalbard (Ericson et al., 2019; Cantoni et al., 2020). Co-occurring
corrosive conditions for aragonite (Fig. 4) and undersaturated pCO2
with respect to the atmosphere (Fig. 3) are unique to cold glacial-melt-influenced coastal regions, which likely enables a positive feedback whereby
CO2 influx from the atmosphere either enhances or prolongs corrosive
summer Ωarag conditions (Evans et al., 2014; Ericson et al.,
2019; Cantoni et al., 2020).
Seasonal drivers
pCO2 variability is determined by thermodynamic and biophysical
forcings, the latter being the sum of the physical and biogeochemical
influences of vertical mixing, horizontal transport, NCP, sea–air CO2
exchange, and calcification. Seasonal variation in pCO2 reflects the
interaction of these terms, which often are competing. For instance, warming
and freshwater input have opposing influences on CO2 solubility such
that together they can dampen pCO2 variability
(Cai et al., 2021; Salisbury and Jönsson, 2018).
As described (Sect. S2), the ratio of the seasonal amplitude of
pCO2 at Tmean (B; Fig. S4) to the pCO2T component (T; Fig. S5) provides information on whether
biophysical processes or seasonal warming are more important for shaping
pCO2 variability within a region. Takahashi et al. (2002)
describe this as T/B (or as a difference, T–B; see their Fig. 9), where if
T/B (or T–B) is greater than 1 (or positive), seasonal temperature change is
the dominant process determining pCO2 variability. The global analysis
by Takahashi et al. (2002) suggests that in the area closest to BC and
southeast AK, temperature and biophysical processes play equal roles in
determining pCO2. A more recent analysis by Fassbender et
al. (2018b) produced similar results for the northeast Pacific with balanced
roles of temperature and biophysical processes evident closest to the coast.
However, both of these analyses were conducted with large global grids that
did not resolve the coastal margin and did not differentiate the role of
freshwater given the open-ocean focus.
In the nearshore zone spanning BC and southeast AK, it is essential to
account for salinity variation when assessing pCO2 variability. As
pointed out by Sarmiento and Gruber (2006), variations in pCO2
that result from changes in salinity cannot be evaluated based solely on the
salinity sensitivity (Takahashi et al., 1993) because this only accounts for
changes in solubility and not the corresponding change in TCO2 and Alk
from a decrease in salinity. Instead, the contribution of changes in
salinity to the pCO2 variability can be evaluated by incorporating
TCO2 and Alk buffer factors into the calculation (Sect. S2). Changes in salinity would result from both freshwater input (decrease)
and vertical mixing (increase) and are expressed here as the pCO2S component (S; Fig. S6). As mentioned, the pCO2S component and pCO2T component can be in opposition such that their
corresponding influences on pCO2 are counterbalanced (Fig. S7). However, there are times and locations when these factors are
not balanced. Lynn Canal (Fig. 1) during the summer months is an important
example of an area and time period when salinity variability exceeds the
influence of seasonal warming (Fig. 5). Subtracting both the pCO2S component and the pCO2T component (TS) from the observed pCO2
leaves remaining variability associated with NCP, calcification, and gas
exchange (Fig. S8). Given that calcification is only
episodically important in this region and gas exchange is generally slow
(on the order of months), this remaining pCO2 variability largely reflects the
influence of NCP, or CO2 removal and addition by organic matter
production and degradation.
Ratios of the seasonal amplitudes of thermodynamic and
biophysical drivers of pCO2 variability. (a) The ratio of
the amplitude of the temperature component (T) to the amplitude of the
remaining biophysical components (BT). (b) The ratio of the
amplitude of the salinity component (S) to the amplitude of the remaining
biophysical components (BS). Areas with major river outflows are
highlighted in this panel (Taku River (T), Stikine River (St), Nass River
(N), Skeena River (Sk), and the Fraser River (F)). (c) The
ratio of the amplitude of the combined temperature and salinity components
(TS) to the amplitude of the remaining biophysical components (BTS).
As illustrated in Fig. 5, the biophysical component dominates over the
temperature component in shaping pCO2 variability on an annual basis
everywhere along the Inside Passage. Excluding Lynn Canal, the salinity
component is also less important than the biophysical component, even in
areas adjacent to major river outflows. At the outflows of major rivers,
such as the Fraser and Stikine (Fig. 5), the salinity component is an
important contributor but still roughly 30 % less than the amplitude of
the biophysical component. In Lynn Canal, salinity variance exceeded all
other locations along the Inside Passage (Figs. 1 and 2), which resulted
in a dominant contribution to the pCO2 variability (Fig. 5).
Temperature counterbalanced some of the salinity component in Lynn Canal,
such that this was the only area where there was near equivalence between
the combined thermodynamic components and the biophysical drivers in
determining pCO2. These computations show the spatial complexity in the
balance between thermodynamic and biophysical drivers in the nearshore zone
and that the influence of salinity must be considered with temperature in
settings with significant freshwater input. The importance of salinity in
shaping marine CO2 system variability in this region has been discussed
previously in modeling studies by Siedlecki et al. (2017) and
Hauri et al. (2020), as well as by Pilcher
et al. (2016), who evaluated the role of variability in freshwater Alk
end-members in enhancing nearshore atmospheric CO2 uptake. However, the
contributions of thermodynamic versus biophysical drivers to the observed
variability have not been evaluated to the extent shown here, which indicated
the dominance of biophysical drivers over most of the Inside Passage.
The northernmost reach of the Inside Passage is heavily influenced by
changes in salinity resulting from the volume of glacial melt water entering
this area (Neal et al., 2010; Edwards et al., 2020).
Reisdorph and Mathis (2013) first described the influence of meltwater
on marine CO2 chemistry in this region, and subsequent observational
and modeling work has considered the de-coupling that can occur between
pCO2 and Ωarag in locales of significant cold glacial melt
discharge (Evans et al., 2014; Ericson et al., 2019; Hauri et al.,
2020; Cantoni et al., 2020). Given that atmospheric CO2 uptake is
promoted in glacially influenced regions, these areas may be important
amplifiers of OA (Cantoni et al., 2020; Ericson et al., 2019; Evans et al.,
2014). Increasing glacial discharge, changes in glacial meltwater Alk as
glaciers further recede and the flow path over land to the ocean increases,
increasing glacial river temperatures, and increasing organic matter
decomposition in glacial rivers are all factors that would modulate the
extremely corrosive conditions found within these nearshore environments, as
well as the decoupling between pCO2 and Ωarag. Given the
potential for intensifying positive feedback with further increasing
atmospheric pCO2 and enhanced sea–air CO2 exchange, thereby
amplifying the already extreme Ωarag conditions, additional
research effort should target these areas in order to understand which
feedbacks dominate from seasonal to inter-annual timescales.
Characterizing regional extremes
Identifying regional extremes in the marine CO2 system is important for
characterizing environmental variability, identifying where unfavorable
conditions for vulnerable marine species occur more often or more intensely,
and pin-pointing areas that may experience faster rates of change from
anthropogenic CO2 input (Feely et al., 2018; Hare et al., 2020).
Marine CO2 system extremes were characterized here based on pHT
and Ωarag variability and severity
(Chan et al., 2017). Importantly, these two
descriptors for extremes may not manifest the same way in a region or with
the same timing; a region may have severely low pHT but also low
variability and experience severely low pHT at a different time of
year than in adjacent areas. pHT and Ωarag extremes can
also be temporally mismatched within a region.
Panels (a–c) show the pHT coefficient of
variation (CV), severity, and the timing of severe pHT conditions,
respectively. Panels (d–f) show the same parameters for Ωarag.
Extremes, as defined by the variability, were regionally similar for
pHT and Ωarag; both the Salish Sea and select areas in
southeast AK exhibited large variability relative to other areas along the
Inside Passage (Fig. 6). Contrasting these highly variable areas,
Johnstone Strait, Sergius Narrows, and Wrangell Narrows (Fig. 1) all had
low variability owing to the influence of persistent tidal mixing. Using
severity to portray extremes provided a nearly inverse picture, with
Johnstone Strait exhibiting both severe pHT and Ωarag, while Lynn Canal had severe Ωarag values but
less severe pHT. Notably, Ωarag severity was only above 1
in surface water most exposed to the open northeast Pacific between AK and
BC, an area known as Dixon Entrance (Fig. 1), and in Sitka; areas exposed
to the open continental shelf generally had less severe and variable
pHT and Ωarag compared to more confined waters along the
Inside Passage.
There were also differences in the timing of severe pHT and Ωarag conditions across and within regions (Fig. 6). The majority of
Inside Passage waters experienced severe pHT and Ωarag
between November and February when seawater pCO2 was highest; however,
severe conditions occurred in some areas earlier in the year. In Johnstone
Strait, most severe pHT and Ωarag occurred in September,
whereas most severe pHT and Ωarag conditions occurred in
June in areas proximal to the Skeena and Stikine outflows due to the
influence of the snowmelt freshet (Fig. 5). In Lynn Canal, most severe
Ωarag values were in August coinciding with the peak input of
glacial melt and in November for pHT owing to storm-induced vertical
mixing. The variation in timing of most severe conditions along the Inside
Passage may have regionally distinct biological implications when these
coincide with times when more sensitive life stages of vulnerable species
are present. In addition, the different characterization of extremes based
on variability and severity has potential implications for adaptation
trajectories as, for example, vulnerable organisms in Johnstone Strait would
experience a sustained corrosive and moderately stable low-pHT
environment, whereas in much of southeast AK, vulnerable organisms would be
subjected to large swings in marine CO2 system parameters over the
year. While some research considers long-term exposure to variable marine
CO2 conditions to be a factor enhancing physiological tolerance to OA
(Kapsenberg and Cyronak, 2019), other research suggests that organisms
living in persistently low-pH environments might be more locally adapted
(Chan et al., 2017). Here we provide information
on the locations of both of these types of settings such that future work
can move to examine how species fare along this gradient within the Inside
Passage.
Standard deviation of pCO2 anomalies (µatm;
a) and the time of detection (years; b) to resolve the secular trend
of increasing seawater pCO2 that tracks the contemporary rise in
atmospheric CO2.
Characterizing extremes is also useful for guiding observing efforts by
identifying locations that either minimize the amount of time anticipated to
observe an OA-driven change in the marine CO2 system or capture a key
process that may be driving large-amplitude signals in a region
(Turk et al., 2019). Areas with low natural variability
require fewer years to resolve a secular trend as opposed to regions of high
variability (Sutton et al.,
2019). Figure 7 shows the standard deviation of de-seasonalized pCO2
observations (anomalies), representing the environmental noise along the
Inside Passage, with the resulting ToD (computed following Eq. 1). The
Salish Sea, the area near the Stikine River, and the northern portion of
Lynn Canal all have very long ToD, and observing efforts in these regions
would be better suited for targeting the processes discussed previously that
shape the variability. As these processes may themselves be subject to
climate change (Bidlack et al., 2021),
tracking their evolving influence on marine CO2 system parameters will
provide valuable information on the dynamics organisms are subjected to in
these highly variable environments. On the other hand, Johnstone Strait, the
area south of Ketchikan, and Sergius Narrows all have much shorter ToD
(Fig. 7). These areas would be ideal for placing observing assets aimed at
resolving long-term secular trends. It is also important to understand that
the ToD estimates computed here are “forced” values (Turk
et al., 2019), in that they are based on seawater pCO2 increasing at a
similar pace to the present atmospheric CO2 increase
(Sutton et al., 2019). In
some cases, there may be an observed trend in a time series that differs
from the forced trend (Laruelle et al., 2018), and this can
reflect either the role of other processes independent of anthropogenic
CO2 increase that modulates the trend in seawater pCO2
(Turk et al., 2019; Salisbury and Jönsson, 2018) or
faster increases in pCO2 resulting from anthropogenic CO2 addition
in weakly buffered settings (Feely et al., 2018). Areas,
such as Johnstone Strait, with low variability that results in short ToD,
and with severe pHT and Ωarag stemming from higher
TCO2:Alk ratios and weaker buffering, likely will exhibit faster rates
of change than estimated by forced ToD values. Identifying and
establishing these areas as sentinel sites for tracking OA would optimize
coastal observing efforts aimed at resolving long-term secular trends.
Estimating past and future conditions
We consider below how the marine CO2 system along the Inside Passage
has evolved over the industrial era as well as what additional change might
be anticipated if greenhouse gas emissions are reduced to reach the
preferable Paris Agreement target of 1.5 ∘C warming
(UNFCC, 2015); referred to here as the 1.5 ∘C
acidification level. It is important to note that our evaluation is
theoretical and only considers the role of anthropogenic CO2 and not
the influence of other forcings like increasing temperature or changing
freshwater input. Following the approach outlined above, anthropogenic
CO2 for Inside Passage surface waters was determined and showed notable
spatiotemporal variability (Fig. 8). This variability was strongly
influenced by freshwater input and water mass age, of which the latter
ranged from 0 to 35 years (Fig. S9). Maximal age estimates
were confined to the areas of persistent tidal mixing and were similar to
estimates from other studies for the age of upwelled water present on the
northeast Pacific continental shelf (Feely et al., 2008; Murray et al.,
2015). This agreement is encouraging considering the potential for
inaccurately representing OUR in the calculation due to a likely higher
oxygen utilization in the confined nearshore regions (Johannessen
et al., 2014; Pawlowicz et al., 2007). It is therefore worth considering how
inaccurately estimating water mass age translates to uncertainty in
anthropogenic CO2 and characterization of preindustrial
[H+]T, pHT, and Ωarag. Inaccurate water mass age
estimates are more influential in older water masses at the surface and over
time periods when atmospheric pCO2 is changing faster (i.e., when computing
contemporary ΔTCO2,diseq). Keeping
this in mind, we consider how a 50 % uncertainty in the age of surface
water in Johnstone Strait impacts our estimation of preindustrial
conditions. An overestimate of the age by 50 % results in a lower
ΔTCO2,diseq by 11.2 µmol kg-1 and a higher anthropogenic CO2 by 11.6 µmol kg-1. This overestimate adjusts the contemporary [H+]T,
pHT, and Ωarag acidification signals (i.e., the difference
between contemporary and preindustrial values) by 0.96 nmol kg-1,
-0.03, and -0.08, respectively. An underestimate of the age by 50 % would
increase ΔTCO2,diseq by 9.6 µmol kg-1 and decrease anthropogenic CO2 by 10 µmol kg-1. The contemporary [H+]T, pHT, and Ωarag acidification signals would adjust by -1.4 nmol kg-1, 0.04,
and 0.07, respectively. Despite the presence of variability in the age
estimate for Johnstone Strait (Fig. S9), we suggest it is
unlikely that the age is underestimated. Rather, this variability is more
likely a function of seasonal variation in sub-surface water mass oxygen
utilization (Johannessen et al., 2014) and the application of a
constant OUR; although, given that the presence of older, upwelled water is
seasonal along this coastline (Feely et al., 2016), some variability in the
age of surface water masses within persistent tidal mixing zones is
expected. We therefore use the absolute values of the shifts in
[H+]T, pHT, and Ωarag resulting from an
overestimate of the water mass age by 50 % as uncertainty bounds when
considering the contemporary and 1.5 ∘C acidification levels.
Contemporary anthropogenic CO2 content (µmol kg-1; a) and the estimated first year when Ωarag was
<1 (b). The x and y axes represent longitude and latitude,
respectively, and with the coastline and terminal positions shown as in
Fig. 1 and time increasing along the z axis.
Interestingly, areas identified previously as pHT and Ωarag extrema based on severity, due to either persistent tidal mixing
or glacial melt input, were not locales containing the highest anthropogenic
CO2 content. Instead, the highest values were in regions that experienced
the greatest O2 supersaturation and pCO2 drawdown during summer
(Figs. 3 and 8). The highest estimated values were near 66 µmol kg-1 and similar to other estimates for coastal northeast Pacific
surface water (Carter et al., 2019a; Feely et al., 2016); however, much
lower values were evident in some locations. The Salish Sea is considered to
have more moderate anthropogenic CO2 levels (Feely et al., 2010; Hare
et al., 2020; Evans et al., 2019), but the freshest areas observed in
southeast AK exhibited very low anthropogenic CO2 content. Such
weakly buffered areas are likely locations that have been corrosive for
aragonite prior to the industrial era. To consider this possibility, we
estimated the first year when Ωarag was <1 along the
Inside Passage (Fig. 8) by calculating the anthropogenic CO2 content
accrued each year along with the resulting change in Ωarag over
the industrial era. The most weakly buffered areas, either due to being very
fresh or because of ventilation of sub-surface water with high TCO2
relative to Alk, are shown by this calculation to have been corrosive at
least on a seasonal basis since the start of the industrial era (Fig. 8).
These naturally corrosive hot spots are being amplified by anthropogenic
CO2 addition such that, for example, Johnstone Strait now experiences
under-saturation throughout the year (Fig. 4). The shift to corrosive
winter conditions has occurred over more recent decades outside of the
mixing zones. Winter surface water in the Salish Sea likely transitioned to
Ωarag<1 beginning around 1950, consistent with the
emergence of corrosive winter values found in a previous study
(Evans et al., 2019). Similar winter transition timing
was evident for Inside Passage waters on the central BC coast, although it has
appeared more recently within the last few decades over large portions of
southeast AK (Fig. 8).
Contemporary [H+]T values minus the estimated
values for 1765 (nmol kg-1; a) and the estimated values for 2035
minus the 1765 values (nmol kg-1; b). The x and y axes represent
longitude and latitude, respectively, and with the coastline and terminal
positions shown as in Fig. 1 and time increasing along the z axis.
Matthews et al. (2021) determined the allowable future
CO2 emissions that limit global warming to the preferred 1.5 ∘C level stated in the Paris Agreement (UNFCC, 2015), the so-called
“remaining carbon budget”, to be 440 GtCO2 from 2020 onwards. Using
the relationships described in Friedlingstein et al. (2021), full
emission of the remaining carbon budget can be equated to a rise in
atmospheric CO2. With 3664 GtCO2 equalling 1 GtC, and every 2124 GtC of emissions increasing atmospheric CO2 by 1 ppm, the remaining
carbon budget would drive an increase in atmospheric CO2 of 56 ppm. At
1 atmosphere of pressure, combining the contemporary atmospheric CO2
mole fraction with the remaining carbon-budget-forced atmospheric CO2
increase results in an atmospheric pCO2 of 468 µatm. The time at
which this atmospheric pCO2 would be realized is trajectory-dependent,
with both the sustainable development pathway (SSP1) and the fossil-fuel
development pathway (SSP5) reaching this atmospheric pCO2 level at
different times (Meinshausen et
al., 2020). SSP5 reaches, and surpasses, this value quickly, by roughly
2035. SSP1 takes longer and reaches this level by 2063. Here we use this
atmospheric CO2 target to consider the theoretical 1.5 ∘C
acidification level and report the year 2035 as the fastest trajectory
following the central estimate for the year when 1.5 ∘C warming
would be reached if the current rate of warming continues
(IPCC, 2018). However, what is important is
that this is an anticipated extent of OA reachable with either trajectory.
If our society follows SSP1, this is theoretically the most acidification we
should expect without considering amplifying processes. However, if we
follow a SSP5-type scenario, acidification will surpass what we estimate
below.
The contemporary Ωarag values minus the
estimated values for 1765 (a) and the estimated values for 2035 minus the
1765 values (b). The x and y axes represent longitude and latitude,
respectively, and with the coastline and terminal positions shown as in
Fig. 1 and time increasing along the z axis.
The estimated average percentage increase in [H+]T along the
Inside Passage is 40 ± 18 % over the 254 years since the start of the
industrial era, although important spatiotemporal variability in
acidification was evident (Fig. 9). This average extent of acidification
equates to a 0.14 ± 0.02 unit drop in pHT. The Inside Passage
average pHT change exceeds the global average
(Lauvset et al., 2020; Jiang et al., 2019);
however, similar to [H+]T, pHT change varied spatially and
temporally with values ranging from -0.06 to -0.20 (Fig. S10). Spatially, [H+]T and pHT change has been greatest in
the more weakly buffered and moderately anthropogenic CO2-concentrated
waters of the Salish Sea and Johnstone Strait. The largest [H+]T
change was evident in Johnstone Strait (Fig. 9), with a maximum near 7 nmol kg-1, due to the inherently lower background pHT level
(Fassbender et al., 2021). For comparison, open-ocean surface water
has experienced an average [H+]T change of 1.6 nmol kg-1
(Fassbender et al., 2021). The pHT change within these two
settings appeared only marginally different, being 0.17 versus 0.1, but with
an over 4-fold increase in [H+]T in Johnstone Strait. On a
seasonal basis over most of the Inside Passage, and despite a higher
anthropogenic CO2 signal during summer (Fig. 8), the change in
[H+]T and pHT appeared larger in winter. Note that seasonal
variation in the acidification signal exceeded the 0.96 nmol kg-1 and
0.03 [H+]T and pHT uncertainties, respectively. This
seasonality in acidification manifests because of seasonal differences in
TCO2:Alk ratios (Fig. S11) that alter the CO2
system response to anthropogenic CO2 increase. During winter, the
TCO2:Alk ratio is closer to unity such that seawater is more
weakly buffered and the percent change in pH following a percentage change
in TCO2 has a greater magnitude. Larger percent changes in pCO2
and Ωarag are also expected during this season, and this
pattern follows the modeled (Fassbender et al.,
2018b; Kwiatkowski and Orr, 2018) and observed (Landschützer
et al., 2018) changes in the seasonality of surface marine CO2
parameters at the global scale.
At an atmospheric pCO2 of 468 µatm, an additional 17 ± 7 % increase in [H+]T, on average, would theoretically be
expected for the 1.5 ∘C acidification level. This implies that
nearly half of the acidification experienced thus far over the industrial
era will likely occur over the coming 15 years if society maintains the
current emissions trajectory. However, a change in emissions trajectory that
follows a sustainable development pathway would enable this acidification
signal to occur over a longer period of time. It is anticipated that
acidification will be further amplified during winter along the Inside
Passage and particularly within the semi-enclosed and more weakly buffered
waterways (Fig. 8). Johnstone Strait and the Salish Sea will likely
continue to experience the largest changes in [H+]T (Fig. 9),
and these areas may serve as bellwethers for the emergence of biological OA
impacts in a similar manner as how high-latitude settings are viewed
(Fabry et al., 2009). Efforts to examine biological impacts in situ should
target these regions where we estimate the largest contemporary and
1.5 ∘C acidification levels. In addition, studies challenging
organisms to adverse [H+]T and pHT levels within experimental
settings may benefit from our estimates of the 1.5 ∘C
acidification level, as these could serve as a near-term treatment for
diagnosing OA impacts.
Unlike for pHT, change in Ωarag over the industrial era
has a seasonal maximum in summer (Fig. 10). This characteristic can also
be explained by considering the seasonality in seawater TCO2 and Alk as
well as relative versus absolute changes in Ωarag. Despite the
TCO2:Alk ratio being closer to unity during winter, and the percent
change in Ωarag following a percentage change in TCO2
being larger during that season (Fig. S11), summer Ωarag values are much higher than winter values (Fig. 4). A 14 %
change in an Ωarag value near 1 is a smaller absolute change
than a 9 % change in an Ωarag value of 3. Considering the
1.5 ∘C acidification level, the change in summer Ωarag values at times may exceed 0.8 units, consistent with an overall
reduction in seasonality as anthropogenic CO2 content continues to
increase (Kwiatkowski and Orr, 2018). To our knowledge, differences
in the season during which maximum absolute changes in [H+]T and
pHT versus Ωarag occur have not been widely acknowledged in
the literature and point to the need for careful consideration of the
specific marine CO2 system parameter an organism may be most sensitive
to (Waldbusser et al., 2014). Seasonally specific changes in the
most impactful marine CO2 system parameter for a sensitive species may
or may not align with periods of maximal vulnerability (Hales
et al., 2016). Considering how the marine CO2 system is being modified
by increasing anthropogenic CO2 on a seasonal as well as long-term
basis, and which specific variable is most impactful for an organism, are
both essential elements for understanding the implications of OA.
Conclusions
Through partnership with the Alaska Marine Highway System, we have reduced
the information gap on marine CO2 system variability along the
northeast Pacific Inside Passage. This study has documented spatiotemporal
variability in surface water along this 1600 km passageway and shown that the
dominant mode of temporal variability is the seasonal cycle and that the
amplitude of this signal varies spatially and is modulated by the relative
influences of tidal mixing, net community production, and the magnitude and
character of freshwater input. While this effort advanced our understanding
of marine CO2 system variability in this region, winter observations
were limited to a single year, and inter-annual variability was not
adequately constrained. Enhancing winter observations and further evaluating
the magnitude of inter-annual variations are both important next steps for
marine CO2 system research in this region.
We have highlighted that within the northernmost area of the Inside Passage,
deviations in freshwater end-members from the broader regional
alkalinity–salinity relationship and the proportionality between salinity
and calcium require further study given the potential for positive feedback
with atmospheric CO2 uptake and for modifications in freshwater
outflows that can alter the coastal OA signals. The impact of proton-binding
organic molecules on the interpretation of alkalinity measurements in this
region is also a large unknown. Analysis of seasonal drivers indicated that
the biophysical component has a dominant role in shaping variability along
most of the Inside Passage but that the combined influences of temperature
and salinity balance the biophysical component in glacial-melt-impacted
areas of southeast AK where these uncertainties are expected to be the
largest. Further research along these lines will be valuable for the
research community striving to understand marine CO2 system patterns in
areas exhibiting large-amplitude variation in salinity.
We considered the characterization of pHT and Ωarag
extremes and recognized that there are regional differences in the
manifestation of extremes based on variability versus severity that likely
have biological implications. Vulnerable organisms experiencing a sustained
corrosive and moderately stable low-pHT environment may have a
differing adaptation trajectory than organisms subjected to large swings in
marine CO2 system parameters over the year. Our diagnosis of these
locations should be useful for future studies examining organismal and
ecosystem adaptation trajectories within the context of OA. We also used our
variability assessment to determine the time of detection and point out that
this information can help optimize coastal observing efforts aimed at
establishing sentinel sites to resolve long-term secular trends or further
evaluating the drivers of large-amplitude variability.
Finally, we estimated the anthropogenic CO2 content in surface water
and considered change over the industrial era and to an atmospheric
CO2 level that corresponds with the exhaustion of the remaining
1.5 ∘C carbon budget. It was shown that some areas, including the
tidally mixed Johnstone Strait and within Lynn Canal, have likely
experienced seasonal Ωarag values <1 over the entire
industrial era. Other areas have transitioned to winter Ωarag
values <1 more recently. Seasonal differences were also identified
in the absolute changes in [H+]T, pHT, and Ωarag.
Absolute [H+]T and pHT changes appeared larger during
winter when conditions are more weakly buffered while Ωarag
change was greater during summer. This difference should be a consideration
when evaluating biological OA impacts. Looking to the future, the
theoretical 1.5 ∘C acidification level suggests that significant
marine CO2 system changes may develop over the coming 15 years if
society continues on a fossil-fuel development emissions trajectory. These
estimates of the 1.5 ∘C acidification level should be useful as a
near-term treatment for regional challenge studies aiming to diagnose
species responses to OA. Time series observations must be expanded and
maintained along the Inside Passage to determine if the 1.5 ∘C
acidification level is realized to the extent suggested by this study.
Code availability
MATLAB routines developed as part of this study are available upon request
to the corresponding author.
Data availability
Three datasets were generated through this study: (1) the record of
directly measured surface pCO2, (2) the gap-filled pCO2 record
including measurements of pH from the BGC-SUMO, and (3) the measurements of
TCO2 and pCO2 on discrete samples collected during the ferry
ride-along cruises. The directly measured surface pCO2 record can be
found within the Surface Ocean CO2 Atlas data holdings
(https://www.socat.info/) as well as within the Ocean Carbon and
Acidification Data Portal at the National Centers for Environmental
Information
(10.25921/jq11-2268, Evans et al., 2020).
The gap-filled pCO2 record including BGC-SUMO pH data can be found with
the discrete sample dataset in the Hakai Institute's data portal
(10.21966/m0es-7520, Evans et al., 2021).
The supplement related to this article is available online at: https://doi.org/10.5194/bg-19-1277-2022-supplement.
Author contributions
WE and AB procured State of Alaska Department of Transportation approval for
the equipment installation aboard the M/V Columbia and the funding for this
project. WE, GTL, and CDH oversaw the installation and operation of the GO8050
pCO2 Monitoring System and ancillary sensors. YT, WE, and CDH oversaw
the installation and operation of the BGC-SUMO. WE participated in
ride-along cruises and collected discrete samples for validation. WE
conducted the analysis and wrote the manuscript. All authors contributing to
revising and editing the manuscript for submission.
Competing interests
The contact author has declared that neither they nor their co-authors have any competing interests.
Disclaimer
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Furthermore, nothing in this report is endorsed by or reflects the views of the State of Alaska Department of Transportation and Public Facilities.
Acknowledgements
We gratefully acknowledge funding support from the Alaska Ocean Observing
System, the Alaska Coastal Rainforest Center at the University of Alaska
Southeast, and the Tula Foundation. Yuichiro Takeshita, and work at the Monterey Bay
Aquarium Research Institute, was supported by the David and Lucile Packard
Foundation and NSF OCE-1736864. This project was made possible through
partnership with the State of Alaska Department of Transportation, and we
thank the crew of the M/V Columbia, who helped to maintain the integrity of the
dataset. We also thank Katie Pocock and Carrie Weekes for processing the
discrete samples used to assess the regional Alk–salinity relationship. We
are grateful for the constructive comments from Andrea Fassbender and the two
anonymous reviewers that have helped to improve this contribution. This is
PMEL contribution number 5298 and CICOES contribution number 2021-1157.
Financial support
This project was not supported by specific awards but by agency contributions. Specifically, funding contributions from the Tula Foundation, the Alaska Ocean Observing System, and the Alaska Coastal Rainforest Center allowed for infrastructure to be purchased and used for this project.
Review statement
This paper was edited by Jack Middelburg and reviewed by two anonymous referees.
ReferencesAmaya, D. J., Miller, A. J., Xie, S.-P., and Kosaka, Y.: Physical drivers
of the summer 2019 North Pacific marine heatwave, Nat. Commun., 11, 1903,
10.1038/s41467-41020-15820-w, 2020.Asplund, M. E., Baden, S. P., Russ, S., Ellis, R. P., Gong, N., and
Hernroth, B. E.: Ocean acidification and host-pathogen interactions: blue
mussels, Mytilus edulis, encountering Vibrio tubiashii, Environ. Microbiol., 16, 1029–1039, 2013.Barton, A., Hales, B., Waldbusser, G., Langdon, C., and Feely, R. A.: The
Pacific oyster, Crassostrea gigas, shows negative correlation to naturally elevated carbon
dioxide levels: Implications for near-term ocean acidification effects,
Limnol. Oceanogr., 57, 698–710, 2012.
Barton, A., Waldbusser, G. G., Feely, R. A., Weisberg, S. B., Newton, J. A.,
Hales, B., Cudd, S., Eudeline, B., Langdon, C. J., Jefferds, I., King, T.,
Suhrbier, A., and McLaughlin, K.: Impacts of coastal acidification on the
Pacific Northwest shellfish industry and adaptation strategies implemented
in response, Oceanography, 28, 146–159, 2015.Bates, N. R., Astor, Y. M., Church, M. J., Currie, K., Dore, J. E.,
Gonzalez-Davila, M., Lorenzoni, L., Muller-Karger, F., Olafsson, J., and
Santana-Casiano, J. M.: A time-series view of changing ocean chemistry due
to ocean uptake of anthropogenic CO2 and ocean acidification,
Oceanography, 27, 126–141,
10.5670/oceanog.2014.16, 2014.Beamer, J. P., Hill, D. F., Arendt, A., and Liston, G. E.: High-resolution
modeling of coastal freshwater discharge and glacier mass balance in the
Gulf of Alaska watershed, Water Resour. Res., 52, 3888–3909, 10.1002/2015WR018457, 2016.Beckwith, S. T., Byrne, R. H., and Hallock, P.: Riverine Calcium End-Members
Improve Coastal Saturation State Calculations and Reveal Regionally Variable
Calcification Potential, Front. Mar. Sci., 6, 169, 10.3389/fmars.2019.00169, 2019.Bednarsek, N., Feely, R. A., Tolimieri, N., Hermann, A. J., Siedlecki, S.
A., Waldbusser, G. G., McElhany, P., Alin, S. R., Klinger, T., Moore-Maley,
B., and Pörtner, H. O.: Exposure history determines pteropod
vulnerability to ocean acidification along the US West Coast, Sci. Rep., 7, 4526, 10.1038/s41598-41017-03934-z, 2017.Bednarsek, N., Feely, R. A., Beck, M. W., Alin, S., Siedlecki, S., Calosi,
P., Norton, E. C., Saenger, C., Štrus, J., Greeley, D., Nezlin, N. P.,
Roethler, M., and Spicer, J. I.: Exoskeleton dissolution with
mechanoreceptor damage in larval Dungeness crab related to severity of
present-day ocean acidification vertical gradients, Sci. Total
Environ., 716, 136610, 10.1016/j.scitotenv.2020.136610, 2020.Bednarsek, N., Newton, J. A., Beck, M. W., Alin, S. R., Feely, R. A.,
Christman, N. R., and Klinger, T.: Severe biological effects under
present-day estuarine acidification in the seasonally variable Salish Sea,
Sci. Total Environ., 765, 142689, 10.1016/j.scitotenv.2020.142689, 2021.Berger, H. M., Siedlecki, S. A., Matassa, C. M., Alin, S. R., Kaplan, I. C.,
Hodgson, E. E., Pilcher, D. J., Norton, E. C., and Newton, J. A.:
Seasonality and Life History Complexity Determine Vulnerability of Dungeness
Crab to Multiple Climate Stressors, AGU Adv., 2, e2021AV000456, 10.1029/2021AV000456, 2021.
Bidlack, A. L., Bisbing, S. M., Buma, B. J., Diefenderfer, H. L., Fellman,
J. B., Floyd, W. C., Giesbrecht, I., Lally, A., Lertzman, K. P., Perakis, S.
S., Butman, D. E., D'Amore, D. V., Fleming, S. W., Hood, E. W., Hunt, B. P.
V., Kiffney, P. M., McNicol, G., Menounos, B., and Tank, S. E.:
Climate-Mediated Changes to Linked Terrestrial and Marine Ecosystems across
the Northeast Pacific Coastal Temperatre Rainforest Margin, BioScience, 71,
biaa171, doi.org/10.1093/biosci/biaa1171, 2021.Bittig, H. C., Körtzinger, A., Neill, C., van Ooijen, E., Plant, J. N.,
Hahn, J., Johnson, K. S., Yang, B., and Emerson, S. R.: Oxygen Optode
Sensors: Principle, Characterization, Calibration, and Application in the
Ocean, Front. Mar. Sci., 4, 429, 10.3389/fmars.2017.00429, 2018.Bond, N. A., Cronin, M. F., Freeland, H., and Mantua, N.: Causes and impacts
of the 2014 warm anomaly in the NE Pacific, Geophys. Res. Lett.,
42, 3414–3420, 10.1002/2015GL063306, 2015.Cai, W. J., Xu, Y.-Y., Feely, R. A., Wanninkhof, R., Jönsson, B. F.,
Alin, S. R., Barbero, L., Cross, J. N., Azetsu-Scott, K., Fassbender, A. J.,
Carter, B. R., Jiang, L.-Q., Pepin, P., Chen, B., Hussain, N., Reimer, J.
J., Xue, L., Salisbury, J. E., Martín Hernández-Ayón, J.,
Langdon, C., Li, Q., Sutton, A. J., Chen, C.-T. A., and Gledhill, D. K.:
Controls on surface water carbonate chemistry along North American ocean
margins, Nat. Commun., 11, 2691, 10.1038/s41467-41020-16530-z,
2020.
Cai, W. J., Feely, R. A., Testa, J. M., Li, M., Evans, W., Alin, S. R., Xu,
Y.-Y., Pelletier, G., Ahmed, A., Greeley, D. J., Newton, J. A., and
Bednarsek, N.: Natural and Anthropogenic Drivers of Acidification in Large
Estuaries, Annu. Rev. Mar. Sci., 13, 11–33, 2021.Caldeira, K. and Wickett, M. E.: Anthropogenic carbon and ocean pH, Nature,
425, 365, 10.1038/425365a, 2003.Cantoni, C., Hopwood, M. J., Clarke, J. S., Chiggiato, J., Achterberg, E.
P., and Cozzi, S.: Glacial drivers of marine biogeochemistry indicate a
future shift to more corrosive conditions in an Arctic fjord, J. Geophys. Res.-Biogeo., 125, e2020JG005633, 10.1029/2020JG005633,
2020.
Carter, B. R., Feely, R. A., Wanninkhof, R., Kouketsu, S., Sonnerup, R. E.,
Pardo, P. C., Sabine, C. L., Johnson, G. C., Sloyan, B. M., Murata, A.,
Mecking, S., Tilbrook, B., Speer, K., Talley, L. D., Millero, F. J.,
Wijffels, S. E., Macdonald, A. M., Gruber, N., and Bullister, J. L.: Pacific
Anthropogenic Carbon Between 1991 and 2017, Global Biogeochem. Cy.,
33, 597–617, 2019a.Carter, B. R., Williams, N. L., Evans, W., Fassbender, A. J., Barbero, L.,
Hauri, C., Feely, R. A., and Sutton, A. J.: Time-of-detection as a metric
for prioritizing between climate observation quality, frequency, and
duration, Geophys. Res. Lett., 46, 3853–3861, 10.1029/2018GL080773, 2019b.Chan, F., Barth, J. A., Blanchette, C. A., Byrne, R. H., Chavez, F. P.,
Cheriton, O., Feely, R. A., Friederich, G., Gaylord, B., Gouchier, T.,
Hacker, S., Hill, T., Hofmann, G., McManus, M. A., Menge, B. A., Nielsen, K.
J., Russell, A., Sanford, E., Sevadjian, J., and Washburn, L.: Persistent
spatial structuring of coastal ocean acidification in the California Current
System, Sci. Rep., 7, 2526, 10.1038/s41598-41017-02777-y, 2017.
Dai, A. and Trenberth, K. E.: Estimates of Freshwater Discharge from
Continents: Latitudinal and Seasonal Variations, J. Hydrometeorol., 3, 660–687, 2002.Dickson, A., Wesolowski, D. J., Palmer, D. A., and Mesmer, R. E.: Dissociation
constant of bisulfate ion in aqueous sodium chloride solutions at 250 ∘C,
J. Phys. Chem. 94, 7978–7985, 10.1021/j100383a042, 1990.Doney, S. C., Fabry, V. J., Feely, R. A., and Kleypas, J. A.: Ocean
Acidification: The Other CO2 Problem, Annu. Rev. Mar. Sci.,
1, 169–192, 2009.Doney, S. C., Busch, D. S., Cooley, S. R., and Kroeker, K. J.: The Impacts
of Ocean Acidification on Marine Ecosystems and Relient Human Communities
Annual Review of Environment and Resources, Annu. Rev., 45, 83–112,
10.1146/annurev-environ-012320-083019, 2020.
Dosser, H. V., Waterman, S., Jackson, J. M., Hannah, C. G., Evans, W., and
Hunt, B. P. V.: Stark Physical and Biogeochemical Differences and
Implications for Ecosystem Stressors in the Northeast Pacific Coastal Ocean,
J. Geophys. Res.-Oceans, 126, e2020JC017033, 2021.Edwards, R. T., D'Amore, D. V., Biles, F. E., Fellman, J. B., Hood, E.,
Trubilowicz, J. W., and Floyd, W. C.: Riverine Dissolved Organic Carbon and
Freshwater Export in the Eastern Gulf of Alaska, J. Geophys. Res.-Biogeo., 126, e2020JG005725, 10.1029/2020JG005725, 2020.Ekstrom, J. A., Suatoni, L., Cooley, S. R., Pendleton, L. H., Waldbusser, G.
G., Cinner, J. E., Ritter, J., Langdon, C., van Hooidonk, R., Gledhill, D.,
Wellman, K., Beck, M. W., Brander, L. M., Rittschof, D., Doherty, C.,
Edwards, P. E. T., and Portela, R.: Vulnerability and adaptation of US
shellfisheries to ocean acidification, Nat. Clim. Change, 5, 207–214,
10.1038/nclimate2508, 2015.Ericson, Y., Falck, E., Chierici, M., Fransson, A., and Kristiansen, S.:
Marine CO2 system variability in a high arctic tidewater-glacier fjord
system, Tempeljorden, Svalbard, Cont. Shelf Res., 181, 1–13, 2019.Evans, W., Hales, B., Strutton, P. G., and Ianson, D.: Sea-air CO2
fluxes in the western Canadian coastal ocean, Prog. Oceanogr., 101, 78–91, 10.1016/j.pocean.2012.1001.1003, 2012.Evans, W. and Mathis, J. T.: The Gulf of Alaska coastal ocean as an
atmospheric CO2 sink, Cont. Shelf Res., 65, 52–63, 2013.Evans, W., Mathis, J. T., and Cross, J. N.: Calcium carbonate corrosivity in an Alaskan inland sea, Biogeosciences, 11, 365–379, 10.5194/bg-11-365-2014, 2014.Evans, W., Mathis, J. T., Ramsay, J., and Hetrick, J.: On the Frontline:
Tracking Ocean Acidification in an Alaskan Shellfish Hatchery, PLoS One, 10,
e0130384, 10.1371/journal.pone.0130384, 2015.Evans, W., Pocock, K., Hare, A., Weekes, C., Hales, B., Jackson, J.,
Gurney-Smith, H., Mathis, J. T., Alin, S. R., and Feely, R. A.: Marine
CO2 Patterns in the Northern Salish Sea, Front. Mar. Sci.,
5, 536, 10.3389/fmars.2018.00536, 2019.Evans, W. and Lebon, G. T., Harrington, C. D., and Bidlack, A.: Surface underway measurements partial pressure of carbon dioxide (pCO2) in the water and atmosphere, sea surface salinity, sea surface temperature, oxygen and other parameters during the Alaska Marine Highway System M/V Columbia 135 service route transits along British Columbia coast, southeast Alaska coast, Gulf of Alaska and North Pacific Ocean from 2017-10-26 to 2019-10-04 (NCEI Accession 0209049), [indicate subset used], NOAA National Centers for Environmental Information, [data set] 10.25921/jq11-2268, 2020.Evans, W., Lebon, G. T., Harrington, C. D., Takeshita, Y., and Bidlack, A.: Marine CO2 system variability along the Inside Passage of the Pacific Northwest coast of North America determined from an Alaskan ferry [data set], 10.21966/m0es-7520, 2021.
Fabry, V. J., McClintock, J. B., Mathis, J. T., and Grebmeier, J. M.: Ocean
Acidification at High Latitudes: The Bellwether, Oceanography, 22, 160–171,
2009.Fassbender, A. J., Sabine, C. L., and Palevsky, H. I.: Nonuniform ocean
acidification and attenuation of the ocean carbon sink, Geophys. Res. Lett., 44, 8404–8413, 10.1002/2017GL074389, 2017.Fassbender, A. J., Alin, S. R., Feely, R. A., Sutton, A. J., Newton, J. A., Krembs, C., Bos, J., Keyzers, M., Devol, A., Ruef, W., and Pelletier, G.: Seasonal carbonate chemistry variability in marine surface waters of the US Pacific Northwest, Earth Syst. Sci. Data, 10, 1367–1401, 10.5194/essd-10-1367-2018, 2018a.Fassbender, A. J., Rodgers, K. B., Palevsky, H. I., and Sabine, C. L.:
Seasonal Asymmetry in the Evolution of Surface Ocean pCO2 and pH
Thermodynamic Drivers and the Influence of Sea-Air CO2 Flux, Global Biogeochem. Cy., 32, 1476–1497, 2018b.Fassbender, A. J., Orr, J. C., and Dickson, A. G.: Technical note: Interpreting pH changes, Biogeosciences, 18, 1407–1415, 10.5194/bg-18-1407-2021, 2021.Feely, R. A., Sabine, C. L., Lee, K., Berelson, W., Kleypas, J., Fabry, V.
J., and Millero, F. J.: Impact of Anthropogenic CO2 on the CaCO3
System in the Oceans, Science, 305, 362–366, 2004a.
Feely, R. A., Sabine, C. L., Schlitzer, R., Bullister, J. L., Mecking, S.,
and Greeley, D.: Oxygen Utilization and Organic Carbon Remineralization in
the Upper Water Column of the Pacific Ocean, J. Oceanogr., 60,
45–52, 2004b.
Feely, R. A., Sabine, C. L., Hernandez-Ayon, M., Ianson, D., and Hales, B.:
Evidence for Upwelling of Corrosive “Acidified” Water onto the Continental
Shelf, Science, 320, 1490–1492, 2008.Feely, R. A., Doney, S. C., and Cooley, S. R.: Ocean Acidification: Present
Conditions and Future Changes in a High-CO2 World, Oceanography, 22,
36–47, 2009.
Feely, R. A., Alin, S. R., Newton, J., Sabine, C. L., Warner, M., Devol, A.,
Krembs, C., and Maloy, C.: The combined effects of ocean acidification,
mixing, and respiration on pH and carbonate saturation in an urbanized
estuary, Eastuar. Coast. Shelf S., 88, 442–449, 2010.
Feely, R. A., Alin, S. R., Carter, B., Bednarsek, N., Hales, B., Chan, F.,
Hill, T. M., Gaylord, B., Sanford, E., Byrne, R. H., Sabine, C. L., Greeley,
D., and Juranek, L.: Chemical and biological impacts of ocean acidification
along the west coast of North America, Eastuar. Coast. Shelf S.,
183, 260–270, 2016.Feely, R. A., Okazaki, R. R., Cai, W. J., Bednarsek, N., Alin, S. R., Byrne,
R. H., and Fassbender, A.: The combined effects of acidification and hypoxia
on pH and aragonite saturation in the coastal waters of the California
current ecosystem and the northern Gulf of Mexico, Cont. Shelf Res., 152, 50–60, 10/1016/j.csr.2017.11.002, 2018.Franco, A. C., Ianson, D., Ross, T., Hamme, R. C., Monahan, A. H.,
Christian, J. R., Davelaar, M., Johnson, W. K., Miller, L. A., Robert, M.,
and Tortell, P. D.: Anthropogenic and Climatic Contributions to Observed
Carbon System Trends in the Northeast Pacific, Global Biogeochem. Cy.,
35, e2020GB006829, 10.1029/2020GB006829, 2021.Friedlingstein, P., Jones, M. W., O'Sullivan, M., Andrew, R. M., Bakker, D. C. E., Hauck, J., Le Quéré, C., Peters, G. P., Peters, W., Pongratz, J., Sitch, S., Canadell, J. G., Ciais, P., Jackson, R. B., Alin, S. R., Anthoni, P., Bates, N. R., Becker, M., Bellouin, N., Bopp, L., Chau, T. T. T., Chevallier, F., Chini, L. P., Cronin, M., Currie, K. I., Decharme, B., Djeutchouang, L., Dou, X., Evans, W., Feely, R. A., Feng, L., Gasser, T., Gilfillan, D., Gkritzalis, T., Grassi, G., Gregor, L., Gruber, N., Gürses, Ö., Harris, I., Houghton, R. A., Hurtt, G. C., Iida, Y., Ilyina, T., Luijkx, I. T., Jain, A. K., Jones, S. D., Kato, E., Kennedy, D., Klein Goldewijk, K., Knauer, J., Korsbakken, J. I., Körtzinger, A., Landschützer, P., Lauvset, S. K., Lefèvre, N., Lienert, S., Liu, J., Marland, G., McGuire, P. C., Melton, J. R., Munro, D. R., Nabel, J. E. M. S., Nakaoka, S.-I., Niwa, Y., Ono, T., Pierrot, D., Poulter, B., Rehder, G., Resplandy, L., Robertson, E., Rödenbeck, C., Rosan, T. M., Schwinger, J., Schwingshackl, C., Séférian, R., Sutton, A. J., Sweeney, C., Tanhua, T., Tans, P. P., Tian, H., Tilbrook, B., Tubiello, F., van der Werf, G., Vuichard, N., Wada, C., Wanninkhof, R., Watson, A., Willis, D., Wiltshire, A. J., Yuan, W., Yue, C., Yue, X., Zaehle, S., and Zeng, J.: Global Carbon Budget 2021, Earth Syst. Sci. Data Discuss. [preprint], 10.5194/essd-2021-386, in review, 2021.Gobler, C. J. and Baumann, H.: Hypoxia and acidification in ocean
ecosystems: coupled dynamics and effects on marine life, Biol. Lett.,
12, 20150976, 10.1098/rsbl.2015.0976, 2016.Gruber, N., Sarmiento, J. L., and Stocker, T. F.: An improved method for
detecting anthropogenic CO2 in the oceans, Global Biogeochem. Cy., 10, 809–837, 1996.Haigh, R., Ianson, D., Holt, C. A., Neate, H. E., and Edwards, A. M.:
Effects of Ocean Acidification on Temperature Coastal Marine Ecosystems and
Fisheries in the Northeast Pacific, PLoS ONE, 10, e0117533, 10.1371/journal.pone.0117533, 2015.
Hales, B., Cai, W.-J., Mitchell, B. G., Sabine, C. L., and Schofield, O.:
North American Continental Margins: A Synthesis and Planning Workshop,
Report of the North American Continental Margins Working Group for the U.S.
Carbon Cycle Scientific Steering Group and Interagency Working Group, edited
by: Hales, B., Cai, W.-J., Mitchell, B. G., Sabine, C. L., and Schofield,
O., U.S. Carbon Cycle Science Program, Washington DC, 110 pp., 2008.Hales, B., Suhrbier, A., Waldbusser, G. G., Feely, R. A., and Newton, J. A.:
The Carbonate Chemistry of the “Fattening Line”, Willapa Bay, 2011–2014,
Estuar. Coast. Shelf S., 40, 173–186, 10.1007/s12237-12016-10136-12237, 2016.Hare, A., Evans, W., Pocock, K., Weekes, C., and Gimenez, I.: Contrasting
marine carbonate systems in two fjords in British Columbia, Canada: seawater
buffering capacity and the response to anthropogenic CO2 invasion, PLoS
ONE, 15, e0238432, 10.1371/journal.pone.0238432, 2020.Hauri, C., Schultz, C., Hedstrom, K., Danielson, S., Irving, B., Doney, S. C., Dussin, R., Curchitser, E. N., Hill, D. F., and Stock, C. A.: A regional hindcast model simulating ecosystem dynamics, inorganic carbon chemistry, and ocean acidification in the Gulf of Alaska, Biogeosciences, 17, 3837–3857, 10.5194/bg-17-3837-2020, 2020.Henson, S. A., Beaulieu, C., Ilyina, T., John, J. G., Long, M.,
Séférian, R., Tjiputra, J., and Sarmiento, J. L.: Rapid emergence of
climate change in environmetal drivers of marine ecosystems, Nat. Commun., 8, 14682 ,10.1038/ncomms14682, 2017.Holdsworth, A. M., Zhai, L., Lu, Y., and Christian, J. R.: Future Changes in
Oceanography and Biogeochemistry Along the Canadian Pacific Continental
Margin, Front. Mar. Sci., 8, 602991, 10.3389/fmars.2021.602991,
2021.Ianson, D., Allen, S. E., Moore-Maley, B. L., Johannessen, S. C., and
Macdonald, R. W.: Vulnerability of a semienclosed estuarine sea to ocean
acidification in contrast with hypoxia, Geophys. Res. Lett., 43,
5793–5801, 10.1002/2016GL068996, 2016.IPCC, 2018: Summary for Policymakers, in: Global Warming of 1.5 ∘C. An IPCC Special Report on the impacts of global warming of 1.5 ∘C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty, edited by: Masson-Delmotte, V., Zhai, P., Pörtner, H.-O., Roberts, D., Skea, J., Shukla, P. R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., Connors, S., Matthews, J. B. R., Chen, Y., Zhou, X., Gomis, M. I., Lonnoy, E., Maycock, T., Tignor, M., and Waterfield, T., World Meteorological Organization, Geneva, Switzerland, 32 pp., 2018.
Jackson, J., Thomson, R. E., Brown, L. N., Willis, P. G., and Borstad, G.
A.: Satellite chlorophyll off the British Columbia Coast, 1997–2010, J. Geophys. Res.-Oceans, 120, 4709–4728, 2015.Jackson, J., Johnson, G. C., Dosser, H. V., and Ross, T.: Warming From
Recent Marine Heatwave Lingers in Deep British Columbia Fjord, Geophys. Res. Lett., 45, 9757–9764, 10.1029/2018GL078971, 2018.
Jiang, L.-Q., Feely, R. A., Carter, B. R., Greeley, D., Gledhill, D. K., and
Arzayus, K. M.: Climatological distribution of aragonite saturation state in
the global oceans, Global Biogeochem. Cy., 29, 1656–1673, 2015.Jiang, L.-Q., Carter, B. R., Feely, R. A., Lauvset, S. K., and Olsen, A.:
Surface ocean pH and buffer capacity: past, present and future, Sci. Rep., 9, 18624, 10.1038/s41598-019-55039-4, 2019.Jin, P., Hutchins, D. A., and Gao, K.: The Impacts of Ocean Acidification on
Marine Food Quality and Its Potential Food Chain Consequences, Front. Mar. Sci., 7, 543979, 10.3389/fmars.2020.543979, 2020.
Johannessen, S. C., Macdonald, R. W., and Paton, D. W.: A sediment and
organic carbon budget for the greater Strait of Georgai, Estuar. Coast. Shelf S., 56, 845–860, 2003.
Johannessen, S. C., Masson, D., and Macdonald, R. W.: Oxygen in the deep
Strait of Georgia, 1951–2009: The roles of mixing, deep-water renewal, and
remineralization of organic carbon, Limnol. Oceanogr., 59, 211–222,
2014.
Johnson, K. S., Jannasch, H. W., Coletti, L. J., Elrod, V. A., Martz, T. R.,
Takeshita, Y., Carlson, R. J., and Connery, J. G.: Deep-Sea DuraFET: A
Pressure Tolerant pH Sensor Designed for Global Sensor Networks, Anal. Chem., 88, 3249–3256, 2016.Juranek, L., Takahashi, T., Mathis, J., and Pickart, R.: Significant
Biologically Mediated CO2 Uptake in the Pacific Arctic During the Late
Open Water Season, J. Geophys. Res.-Oceans, 124, 821–843,
doi.org/10.1029/2018JC014568, 2019.Kapsenberg, L. and Cyronak, T.: Ocean Acidification refugia in variable
environments, Glob. Change Biol., 25, 3201–3214, 10.1111/gcb.14730, 2019.
Kroeker, K. J., Kordas, R., L., Crim, R., Hendriks, I. E., Ramajo, L.,
Singh, G. S., Duarte, C. M., and Gattuso, J. P.: Impacts of ocean
acidification on marine organisms: quantifying sensitivities and
interactions with warming, Glob. Change Biol., 19, 1884–1896, 2013.Kroeker, K., Kindinger, T., Hirsh, H., Ward, M., Hill, T., Jellison, B., Koweek, D., Lummis, S., Rivest, E., Waldbusser, G., and Gaylord, B.: Reviews and Syntheses: Spatial and temporal patterns in metabolic fluxes inform potential for seagrass to locally mitigate ocean acidification, Biogeosciences Discuss. [preprint], 10.5194/bg-2021-137, in review, 2021.
Kwiatkowski, L. and Orr, J. C.: Diverging seasonal extremes for ocean
acidification during the twenty-first century, Nat. Clim. Change, 8,
141–145, doi.org/10.1038/s41558-017-0054-0, 2018.Landschützer, P., Gruber, N., Bakker, D. C. E., Stemmler, I., and Six,
K. D.: Strengthening seasonal marine CO2 variations due to increasing
atmospheric CO2, Nat. Clim. Change, 8, 146–150, 10.1038/s41558-017-0057-x,
2018.Laruelle, G. G., Cai, W. J., Hu, X., Gruber, N., Mackenzie, F. T., and
Regnier, P.: Continental shelves as a variable but increasing global sink
for atmospheric carbon dioxide, Nat. Commun., 9, 454,
10.1038/s41467-41017-02738-z, 2018.Lauvset, S. K., Carter, B. R., Perez, F. F., Jiang, L.-Q., Feely, R. A.,
Velo, A., and Olsen, A.: Processes Driving Global Interior Ocean pH
Distribution, Global Biogeochem. Cy., 34, e2019GB006229,
10.1029/2019GB006229, 2020.Lowe, A., Bos, J., and Ruesink, J.: Ecosystem metabolism drives pH
variability and modulates long-term ocean acidification in the Northeast
Pacific coastal ocean, Sci. Rep., 9, 963,
10.1038/s41598-41018-37764-41594, 2019.Marshall, K. N., Kaplan, I. C., Hodgson, E. E., Hermann, A., Busch, D. S.,
McElhany, P., Essington, T. E., Harvey, C. J., and Fulton, E. A.: Risks of
ocean acidification in the California Current food web and fisheries:
ecosystem model projections, Glob. Change Biol., 23, 1525–1539, 10.1111/gcb.13594,
2017.Mathis, J. T., Cooley, S. R., Lucey, N., Colt, S., Ekstrom, J., Hurst, T.,
Hauri, C., Evans, W., Cross, J. N., and Feely, R. A.: Ocean acidification
risk assessment for Alaska's fishery sector, Prog. Oceanogr., 136,
71–91, 10.1016/j.pocean.2014.07.001, 2015.Matsumoto, K. and Gruber, N.: How accurate is the estimation of
anthropogenic carbon in the ocean?, An evaluation of the ΔC* method, Global Biogeochem. Cy., 19, GB3014, 10.1029/2004GB002397, 2005.Matthews, H. D., Tokarska, K. B., Rogelj, J., Smith, C. J., MacDougall, A.
H., Haustein, K., Mengis, N., Sippel, S., Forster, P. M., and Knutti, R.: An
integrated approach to quantifying uncertainties in the remaining carbon
budget, Commun. Earth Environ., 7, 10.1038/s43247-020-00064-9,
2021.Meinshausen, M., Nicholls, Z. R. J., Lewis, J., Gidden, M. J., Vogel, E., Freund, M., Beyerle, U., Gessner, C., Nauels, A., Bauer, N., Canadell, J. G., Daniel, J. S., John, A., Krummel, P. B., Luderer, G., Meinshausen, N., Montzka, S. A., Rayner, P. J., Reimann, S., Smith, S. J., van den Berg, M., Velders, G. J. M., Vollmer, M. K., and Wang, R. H. J.: The shared socio-economic pathway (SSP) greenhouse gas concentrations and their extensions to 2500, Geosci. Model Dev., 13, 3571–3605, 10.5194/gmd-13-3571-2020, 2020.Meire, L., Søgaard, D. H., Mortensen, J., Meysman, F. J. R., Soetaert, K., Arendt, K. E., Juul-Pedersen, T., Blicher, M. E., and Rysgaard, S.: Glacial meltwater and primary production are drivers of strong CO2 uptake in fjord and coastal waters adjacent to the Greenland Ice Sheet, Biogeosciences, 12, 2347–2363, 10.5194/bg-12-2347-2015, 2015.Mekkes, L., Renema, W., Bednaršek, N., Alin, S. R., Feely, R. A.,
Huisman, J., Roessingh, P., and Peijnenburg, K. T. C. A.: Pteropods make
thinner shellfs in the upwelling region of the California Current Ecosystem,
Sci. Rep., 11, 1731, 10.1038/s41598-41021-81131-41599, 2021.Morrison, J., Foreman, M. G. G., and Masson, D.: A Method for Estimating
Monthly Freshwater Discharge Affecting British Columbia Coastal Waters,
Atmos. Ocean, 50, 1–8, 10.1080/07055900.2011.637667, 2012.Murray, J. W., Roberts, E., Howard, E., O'Donnell, M., Bantam, C.,
Carrington, E., Foy, M., Paul, B., and Fay, A.: An inland sea high
nitrate-low chlorophyll (HNLC) region with naturally high pCO2,
Limnol. Oceanogr., 60, 957–966, 10.1002/lno.10062, 2015.Neal, E. G., Hood, E., and Smikrud, K.: Contribution of glacier runoff to
freshwater discharge into the Gulf of Alaska, Geophys. Res. Lett.,
37, L06404, 10.1029/2010GL042385, 2010.Newton, J. A., Feely, R. A., Jewett, E. B., Williamson, P., and Mathis, J.:
Global Ocean Acidification Observing Network: Requirements and Governance
Plan, http://goa-on.org/docs/GOA-ON_2nd_edition_final.pdf (last access: May 2021), 2015.
O'Neel, S., Hood, E., Bidlack, A. L., Fleming, S. W., Arimitsu, M. L.,
Arendt, A., Burgess, E., Sergeant, C. J., Beaudreau, A. H., Timm, K.,
Hayward, G. D., Reynolds, J. H., and Pyare, S.: Icefield-to-Ocean Linkages
across the Northern Pacific Coastal Temperate Rainforest Ecosystem,
BioScience, 65, 499–512, 2015.Oliver, A. A., Tank, S. E., Giesbrecht, I., Korver, M. C., Floyd, W. C., Sanborn, P., Bulmer, C., and Lertzman, K. P.: A global hotspot for dissolved organic carbon in hypermaritime watersheds of coastal British Columbia, Biogeosciences, 14, 3743–3762, 10.5194/bg-14-3743-2017, 2017.Orr, J. C., Epitalon, J.-M., Dickson, A. G., and Gattuso, J.-P.: Routine
uncertainty propagation for the marine carbon dioxide system, Mar. Chem., 207, 84–107, 10.1016/j.marchem.2018.10.006, 2018.Osborne, E. B., Thunell, R. C., Gruber, N., Feely, R. A., and
Benitez-Nelson, C. R.: Decadal variability in twentieth-century ocean
acidification in the California Current Ecosystem, Nat. Geosci., 13, 43–49,
10.1038/s41561-41019-40499-z, 2020.Pacella, S. R., Brown, C. A., Waldbusser, G. G., Labiosa, R. G., and Hales,
B.: Seagrass habitat metabolism increases short-term extremes and long-term
offset of CO2 under future ocean acidification, P. Natl. Acad. Sci. USA, 15, 3870–3875, 10.1073/pnas.1703445115, 2018.
Pawlowicz, R., Riche, O., and Halverson, M.: The Circulation and Residence
Time of the Strait of Georgia using a Simple Mixing-box Approach,
Atmos. Ocean, 45, 1730-193, 2007.Peck, V. L., Oakes, R. L., Harper, E. M., Manno, C., and Tarling, G. A.:
Pteropods counter mechanical damage and dissolution through extensive shell
repair, Nat. Commun., 9, 264, 10.1038/s41467-41017-026692-w, 2018.
Perez, F. F. and Fraga, F.: Association constant of fluoride and hydrogen
ions in seawater, Mar. Chem., 21, 161–168, 1987.Pierrot, D., Neill, C., Sullivan, K., Castle, R., Wanninkhof, R., Lüger,
H., Johannessen, T., Olsen, A., Feely, R. A., and Cosca, C. E.:
Recommendations for autonomous underway pCO2 measuring systems and
data-reduction routines, Deep Sea Res. Pt. II, 56, 512–522, 10.1016/j.dsr2.2008.12.005, 2009.Pilcher, D. J., Siedlecki, S. A., Hermann, A. J., Coyle, K. O., Mathis, J.
T., and Evans, W.: Simulated impact of high alkalinity glacial runoff on
CO2 uptake in the coastal Gulf of Alaska, Geophys. Res. Lett., 45, Pages 880–890, 10.1002/2017GL075910, 2016.Raven, J. A., Gobler, C. J., and Juel Hansen, P.: Dynamic CO2 and pH
levels in coastal, estuarine, and inland waters: Theoretical and observed
effects on harmful algal blooms, Harmful Algae, 91, 101594,
doi/l10.1016/j.hal.2019.1003.1012, 2020.
Reisdorph, S. C. and Mathis, J. T.: The dynamic controls on carbonate
mineral saturation states and ocean acidification in a glacially dominated
estuary, Estuar. Coast. Shelf S., 144, 8–18, 2013.Ricart, A. M., Ward, M., Hill, T. M., Sanford, E., Kroeker, K. J.,
Takeshita, Y., Merolla, S., Shukla, P., Ninokawa, A. T., Elsmore, K., and
Gaylord, B.: Coast-wide evidence of low pH amelioration by seagrass
ecosystems, Glob. Change Biol., 27, 2580–2591, 10.1111/gcb.15594, 2021.Rogelj, J., Shindell, D., Jiang, K., Fifita, S., Foster, P., Ginzburg, V., Handa, C., Kheshgi, H., Kobayashi, S., Kriegler, E., Mundaca, L., Séférian, R., and Vilariño, M.: Mitigation pathways compatible with 1.5 ∘C in the context of sustainable development, in: Global Warming of 1.5 ∘C, An IPCC Special Report on the impacts of global warming of 1.5 ∘C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty (93–174), edited by: Masson-Delmotte, V., Zhai, P., Pörtner, H.-O., Roberts, D., Skea, J., Shukla, P. R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., Connors, S., Matthews, J. B. R., Chen, Y., Zhou, X., Gomis, M. I., Lonnoy, E., Maycock, T., Tignor, M., and Waterfield, T., IPCC/WMO, https://www.ipcc.ch/site/assets/uploads/sites/2/2019/05/SR15_Chapter2_Low_Res.pdf (last acces: May 2021), 2018.Sabine, C. L., Feely, R. A., Key, R. M., Bullister, J. L., Millero, F. J.,
Lee, K., Peng, T.-H., Tilbrook, B., Ono, T., and Wong, C. S.: Distribution
of anthropogenic CO2 in the Pacific Ocean, Global Biogeochem. Cy., 16, 1083, 10.1029/2001GB001639, 2002.Sabine, C. L., Feely, R. A., Gruber, N., Key, R. M., Lee, K., Bullister, J.
L., Wanninkhof, R., Wong, C. S., Wallace, D. W. R., Tilbrook, B., Millero,
F. J., Peng, T.-H., Kozyr, A., Ono, T., and Rios, A. F.: The Oceanic Sink
for Anthropogenic CO2, Science, 305, 367–371, 2004.Salisbury, J. E., and Jönsson, B. F.: Rapid warming and salinity changes
in the Gulf of Maine alter surface ocean carbonate parameters and hide ocean
acidification, Biogeochemistry, 141, 401–418, 10.1007/s10533-10018-10505-10533, 2018.
Sarmiento, J. L. and Gruber, N.: Ocean Biogeochemical Dynamics, Princeton
University Press, Princeton, ISBN: 9780691017075, 2006.Sharp, J. D. and Byrne, R. H.: Interpreting measurements of total
alkalinity in marine and estuarine waters in the presence of proton-binding
organic matter, Deep Sea Res. Pt. I, 165,
103338, 10.1016/j.dsr.2020.103338, 2020.Sharp, J. D., Pierrot, D., Humphreys, M. P., Epitalon, J.-M., Orr, J. C.,
Lewis, E. R., and Wallace, D. W. R.: CO2SYSv3 for MATLAB (Version v3.2.0),
Zenodo, 10.5281/zenodo.3950562, 2021.
Siedlecki, S. A., Pilcher, D. J., Hermann, A. J., Coyle, K., and Mathis, J.:
The Importance of Freshwater to Spatial Variability of Aragonite Saturation
State in the Gulf of Alaska, J. Geophys. Res.-Oceans, 122,
8482–8502, 2017.St. Pierre, K. A., Oliver, A. A., Tank, S. E., Hunt, B. P. V., Giesbrecht,
I., Kellogg, C. T. E., Jackson, J. M., Lertzman, K. P., Floyd, W. C., and
Korver, M. C.: Terrestrial exports of dissolved and particulate organic
carbon affect nearshore ecosystems of the Pacific coastal temperate
rainforest, Limnol. Oceanogr., 65, 2657–2675, 10.1002/lno.11538, 2020.St. Pierre, K. A., Hunt, B. P. V., Tank, S. E., Giesbrecht, I., Korver, M. C., Floyd, W. C., Oliver, A. A., and Lertzman, K. P.: Rain-fed streams dilute inorganic nutrients but subsidise organic-matter-associated nutrients in coastal waters of the northeast Pacific Ocean, Biogeosciences, 18, 3029–3052, 10.5194/bg-18-3029-2021, 2021.
Stabeno, P. J., Bond, N. A., Hermann, A. J., Kachel, N. B., Mordy, C. W.,
and Overland, J. E.: Meteorology and oceanography of the Northern Gulf of
Alaska, Cont. Shelf Res., 24, 859–897, 2004.Sutton, A. J., Feely, R. A., Maenner-Jones, S., Musielwicz, S., Osborne, J., Dietrich, C., Monacci, N., Cross, J., Bott, R., Kozyr, A., Andersson, A. J., Bates, N. R., Cai, W.-J., Cronin, M. F., De Carlo, E. H., Hales, B., Howden, S. D., Lee, C. M., Manzello, D. P., McPhaden, M. J., Meléndez, M., Mickett, J. B., Newton, J. A., Noakes, S. E., Noh, J. H., Olafsdottir, S. R., Salisbury, J. E., Send, U., Trull, T. W., Vandemark, D. C., and Weller, R. A.: Autonomous seawater pCO2 and pH time series from 40 surface buoys and the emergence of anthropogenic trends, Earth Syst. Sci. Data, 11, 421–439, 10.5194/essd-11-421-2019, 2019.Takahashi, T., Sutherland, S. C., Sweeney, C., Poisson, A., Metzl, N.,
Tilbrook, B., Bates, N. R., Wanninkhof, R., Feely, R. A., Sabine, C. L.,
Olafsson, J., and Nojiri, Y.: Global sea-air CO2 flux based on
climatological surface ocean pCO2, and seasonal biological and
temperature effects, Deep-Sea Res. Pt. II, 49, 1601–1622, 2002.Takeshita, Y., Frieder, C. A., Martz, T. R., Ballard, J. R., Feely, R. A., Kram, S., Nam, S., Navarro, M. O., Price, N. N., and Smith, J. E.: Including high-frequency variability in coastal ocean acidification projections, Biogeosciences, 12, 5853–5870, 10.5194/bg-12-5853-2015, 2015.Takeshita, Y., Johnson, K. S., Martz, T., Plant, J. N., and Sarmiento, J.
L.: Assessment of Autonomous pH Measurements for Determining Surface
Seawater Partial Pressure of CO2, J. Geophys. Res.-Oceans, 123, 4003–4013, 2018.
Thomson, R. E.: Oceanography of the British Columbia coast, Canadian Special
Publication of Fisheries and Aquatic Sciences 56, Department of Fisheries
and Oceans, Ottawa, ISBN: 0-660-10978-6, 1981.Tilbrook, B., Jewett, E. B., DeGrandpre, M. D., Hernandez-Ayon, J. M.,
Feely, R. A., Gledhill, D. K., Hansson, L., Isenee, K., Kurz, M. L., Newton,
J. A., Siedlecki, S. A., Chai, F., Dupont, S., Graco, M., Calvo, E.,
Greeley, D., Kapsenberg, L., Lebrec, M., Pelejero, C., Schoo, K., and
Telszewski, M.: An Enhanced Ocean Acidification Observing Network: From
People to Technology to Data Synthesis and Information Exchange, Front. Mar. Sci., 6, 337, 10.3389/fmars.2019.00337, 2019.Tortell, P. D., Merzouk, A., Ianson, D., Pawlowicz, R., and Yelland, D. R.:
Influence of regional climate forcing on surface water pCO2, ΔO2/Ar
and dimethylsulfide (DMS) along the southern British Columbia coast,
Cont. Shelf Res., 47, 119–132, 10.1016/j.csr.2012.07.007, 2012.Turk, D., Wang, H., Hu, X., Gledhill, D., Wang, Z. A., Jiang, L., and Cai,
W. J.: Time of Emergence of Surface Ocean Carbon Dioxide Trends in the North
American Coastal Margins in Support of Ocean Acidification Observing System
Design, Front. Mar. Sci., 6, 91, 10.3389/fmars.2019.00091, 2019.UNFCC: Adoption of the Paris Agreement FCCC/CP/2015/L.2019/Rev.2011.
2011–2032 (UNFCCC, Paris, France, 2015), https://unfccc.int/sites/default/files/resource/docs/2015/cop21/eng/l09r01.pdf (last access: May 2021), 2015.Uppström, L. R.: The boron/chlorinity ratio of deep-sea water from the
Pacific Ocean, Deep-Sea Res., 21, 161–162,
10.1016/0011-7471(74)90074-6, 1974.Waldbusser, G. G., Hales, B., Langdon, C. J., Haley, B. A., Schrader, P.,
Brunner, E. L., Gray, M. W., Miller, C. A., and Gimenez, I.:
Saturation-state sensitivity of marine bivalve larvae to ocean
acidification, Nat. Clim. Change, 5, 273–280, 10.1038/nclimate2479, 2014.Wanninkhof, R. and Thoning, K.: Measurement of fugacity of CO2 in
surface water using continuous and discrete sampling methods, Mar. Chem., 44, 189–204, 1993.Wanninkhof, R., Bakker, D., Bates, N., Olsen, A., Steinhoff, T., and Sutton,
A.: Incorporation of Alternative Sensors in the SOCAT Database and
Adjustments to Dataset Quality Control Flags, https://www.ncei.noaa.gov/access/ocean-carbon-data-system/oceans/Recommendationnewsensors.pdf (last access: May 2021),
2013.Ward, N. D., Bianchi, T. S., Medeiros, P. M., Seidel, M., Richey, J. E.,
Keil, R. G., and Sawakuchi, H. O.: Where Carbon Goes When Water Flows:
Carbon Cycling across the Aquatic Continuum, Front. Mar. Sci., 4, 7,
10.3389/fmars.2017.00007, 2017.
Ware, D. M. and Thomson, R. E.: Bottom-Up Ecosystem Trophic Dynamics
Determine Fish Production in the Northeast Pacific, Science, 308, 1280–1284,
2005.
Waters, J., Millero, F. J., and Woosley, R. J.: Corrigendum to “The free
proton concentration scale for seawater pH”, [MARCHE: 149(2013) 8-22],
Mar. Chem., 165, 66–67, 2014.
Weingartner, T. J., Eisner, L., Eckert, G. L., and Danielson, S. L.:
Southeast Alaska: oceanographic habitats and linkages, J. Biogeogr., 36, 387–400, 2009.
Whitney, F. A., Crawford, W. R., and Harrison, P. J.: Physical processes
that enhance nutrient transport and primary productivity in the coastal and
open ocean of the subartic NE Pacific, Deep-Sea Res. Pt. II, 52, 681–706,
2005.