Stuck in the middle: thermal regimes of coastal lagoons and estuaries in a warming world


               <jats:p>N/A</jats:p>


Introduction
Water temperatures impact water chemistry and ecosystem productivity and diversity, making water temperature the ecological master factor [1]. In freshwater environments, climate change and other anthropogenic disturbances are causing widespread river, stream, and lake warming and a loss or fragmentation of cold-water habitat [2,3]. Models project accelerated future river warming [4] and cascading impacts to cold-water biodiversity. In marine environments, oceanographers have documented intensifying marine heat waves [5] and multi-decadal oceanic warming, with implications for ecosystems, ocean circulation, and global sea levels [6].
These topics have become major research fields, with over 5000 papers per year published on river and stream temperature alone [7], and research networks with coordinated large-scale marine or freshwater temperature monitoring [e.g. 8,9]. These networks have focused on either end of the salinity spectrum (oceans or rivers) and have generally overlooked transitional, coastal waters. This perspective highlights the importance and distinctive aspects of the thermal regimes of coastal lagoons and estuaries in temperate zones and the need to understand their sensitivity to climate change through the establishment of large-scale, coordinated monitoring and modeling studies.

Importance of estuarine and lagoon thermal regimes
Estuaries and coastal lagoons exhibit tidal dynamics and are influenced by both seawater and freshwater inputs. Due in part to their atmospheric and marine forcing, these water bodies are characterized by pronounced spatiotemporal variability in water temperature and salinity, which allows them to support rich, complex ecosystems that provide essential services globally [10]. Estuaries and lagoons also function as critical aquatic corridors for many migratory species, with perhaps the most attention paid to anadromous salmonids. Coastal waters that warm in response to climate change or other environmental perturbations can act as thermal dams that impose migratory barriers. Transition through these zones is already physiologically taxing due to the sharp salinity gradients, and further stresses imposed by high water temperatures can result in altered migration timing or temperature-induced mortalities [e.g . 11]. Coastal lagoons and estuaries can also support intensive aquaculture operations that are rapidly expanding worldwide, including for temperature-sensitive species [12]. Rising water temperatures could threaten these operations through several mechanisms, including increased eutrophication. Climate change alterations to the thermal regimes of estuaries and lagoons warrant more attention as they threaten economic and environmental sustainability in coastal regions and could negatively impact food security.

Distinctive aspects of estuarine and lagoon thermal regimes
The extensive thermal monitoring and modeling presently underway in freshwater or marine environments is not sufficient to understand the thermal sensitivity of estuaries and coastal lagoons to climate change. Temperatures in coastal waters are characterized by drivers and dynamics that are distinct from, but depend on, those in freshwater and marine settings. For example, estuarine thermal regimes are complex, as they are governed by atmospheric heat fluxes, tide-and storm-driven exchange with the ocean, estuarine hydraulics, and freshwater hydrology (figure 1). In the following sub-sections, we briefly elaborate on three distinctive and important thermal drivers for estuaries and lagoons, although we acknowledge there are other important thermal processes in coastal zones. We primarily contrast and storms (c) depicted. Assuming the ocean is thermally moderated compared to the inflowing river, the intruded seawater at high tides or surges can warm (cool) the estuary during winter (summer). Instrument deployment in (c) indicates the sensor deployment for figure 2(f). estuaries and lagoons with freshwater systems [7], as the differences between relatively small coastal water bodies and large marine water bodies with high thermal inertia are evident. As illustrative examples, we present new data collected from the estuary of the Chéticamp River in Nova Scotia, Canada (figure 2). The Chéticamp River (276 km 2 watershed) has a mixed semi-diurnal tidal range of approximately 1 m at its mouth, and a humid continental climate. The river is a popular fishing destination with an intensively monitored freshwater thermal regime given its population of Atlantic salmon (Salmo salar).

Density-dependent stratification and dynamics
Variable-density water bodies are often stratified with a zone of less dense (i.e. warmer and/or less saline) water overlying a zone of denser water (figure 1(a)). Such stratification is not found in well-mixed streams and rivers but is common in lakes, estuaries, and lagoons. In the absence of contamination from road salt, freshwater lake stratification is driven by small density differences arising from water temperature. For example, at atmospheric pressure, freshwater reaches its maximum density of 1000 kg m −3 at 4 • C, but must warm to 45 • C before its density drops by 1%. In contrast, in estuarine settings, water density can range by 30 kg m −3 (3%) by isothermally varying salinity between freshwater and ocean values. Thus, estuaries and lagoons along marine coastlines can be stratified by salinity alone and are more likely to be stratified than freshwater bodies with the same hydraulics and morphology. Importantly, stratification from salinity gradients often also imposes strong vertical temperature gradients. Figure 2(d) illustrates this by presenting pre-storm water temperature and salinity data at the top and bottom of the surface water column close to the mouth of the Chéticamp estuary. There is up to 5.3 • C difference between these sensors in 2021 despite being separated by only ∼2-3 m depending on the tidal stage. Such thermal offsets are common in estuary and lagoon salt wedges (figure 1(a)) and reflect vertical variation in salinity and oxygen.
These spatial complexities in temperature are critical for supporting the rich, diverse ecosystems in coastal waters. Cold-water species that can tolerate a salinity range and are mobile may be able to survive in thermal niches in colder estuary zones as is common in rivers [13]. However, changes in dissolved oxygen and salinity across stratified zones may pose challenges to such thermoregulation, and this remains a source of ongoing research [14]. Complex density-dependent stratification and associated spatiotemporal thermal complexity for coastal water bodies may be beneficial for ecosystems, but it is challenging to model. The 1D (longitudinal) numerical thermal advection and dispersion models [15] frequently employed for short-term temperature forecasting or long-term warming projections in wellmixed rivers are not well suited for estuaries and lagoons. However, fully 3D, transient temperature models of variable-density coastal water bodies [e.g. 16] are computationally expensive, require considerable parameterization or calibration data, and are difficult to scale spatially. This likely explains the relative lack of thermal modeling for estuaries and lagoons compared to rivers. Further research is required to understand how these critical migratory corridors will respond to changing river flows, atmospheric heat fluxes, and marine climate change [11].

Tidal forcing and dynamics
Another distinctive aspect of coastal thermal regimes is the combined atmospheric and tidal forcing they are subject to and the associated multi-frequency  (panel (c)). The specific conductance and water temperatures are distinct at the top and bottom but converge during and immediately following a coastal storm (see grey shading). dynamics they exhibit. During the summer, rivers are typically warmer than the ocean given their lower thermal inertia. Thus, estuaries and lagoons tend to be warmer at low tide, when freshwater flow dominates, than at high tide, when cooler seawater intrudes (figures 1(a) and (b)), although the baseflow dominance may influence this relationship. Figure 2(d) presents water temperatures from the Chéticamp estuary in the time domain, for which both diurnal and semi-diurnal thermal signals are apparent. Diurnal signals can be caused by diurnal tidal constituents, but diurnal radiative forcing from the atmosphere is often the primary driver. In contrast, semi-diurnal temperature signals are driven by tides, at least along marine coastlines with semidiurnal tidal level constituents. The different dominant frequencies are even more apparent when the temperature time series are converted to the frequency domain (figure 2(e)).
Quantifying the relative dominance of diurnal vs. semi-diurnal forcing can help reveal the marine and freshwater advective thermal controls on coastal water bodies and their sensitivity to environmental change. We propose the dimensionless thermal form factor F t , which is computed as the ratio of the diurnal temperature amplitude ∆T D to the semidiurnal amplitude ∆T s , where both can be estimated through dynamic harmonic regression or Fourier transforms (figure 2(e)): This is the thermal analogue to the tidal form factor used by coastal engineers and oceanographers to classify tide level oscillations (p. 77 of [17]). A high F t indicates that diurnal signals dominate (i.e. solar radiation or diurnal tides) the thermal regime, while a low F t indicates semi-diurnal (tidal exchange) signals dominate. Figures 2(d) and (e) shows that one of the lowest F t values (highest semi-diurnal influence) of the selected temperature time series was obtained for the temperature logger installed in an intertidal spring. This is because the logger records groundwater discharge temperatures at low tide but measures surface water temperature at high tide when the hydraulics facilitate more thermal mixing within the water column and when groundwater discharge slows or ceases due to altered hydraulic gradients. Such dynamics point to the critical role of groundwater for enhancing thermal heterogeneity in coastal ecosystems and the importance of tide-driven exchange in both groundwater and surface water systems.
The semi-diurnal temperature signals (figures 2(d) and (e)) indicate that oscillating coastal hydrodynamics strongly influence the thermal regimes of coastal water bodies. Deterministic modeling of coastal thermal regimes must then be underpinned by coupled modeling of complex coastal hydrodynamics, providing further explanation for the relative lack of quantitative assessments of coastal water thermal sensitivity to climate change compared to river thermal sensitivity assessments. Further, long-term thermal sensitivity studies of coastal waters must consider how the impacts of sea-level rise and bathymetry change on tidal pumping may influence thermal dynamics.

Coastal storm forcing and dynamics
Rivers and streams are thermally impacted by high precipitation events, but they are not subjected to thermal perturbations from intruded seawater during coastal storms such as tropical cyclones. In contrast, coastal waters can be thermally influenced by both atmospheric and marine extreme events. Coastal storms can generate waves or storm surges that push high volumes of seawater into coastal water bodies ( figure 1(c)). Such events have at least two thermal effects: (1) pronounced thermal advection as the intruded seawater can be relatively warm or cool depending on the time of year and the event conditions, and (2) disturbed stratification from altered salinity distributions, with resultant thermal homogenization. Figure 2(f) illustrates the thermal disturbance of a coastal storm in the Chéticamp estuary with homogenizing (salinity and temperature) effects lasting from 2-7 July 2021 (grey shading), indicating that hydrodynamic forcing can interact with and overcome stratification from density differences (section 3.1). The specific conductance data reveal initial freshening from high precipitation and freshwater flow and then a sharp spike in salinity due to the surge and waves. The high flows and likely mixing from wind and waves caused the temperatures at the top and bottom of the surface water column to converge. These thermal effects are eventually overwhelmed following a storm as the natural stratification, flow regime, and thermal regime is restored ( figure 1(f)). Despite their transient nature, coastal storms can significantly influence the salinity and temperature distribution of a coastal water body. Surges can also interact with tides (section 3.2) as surges can impact zones further upriver when they coincide with high tide. The impacts of these storms on temperature-sensitive coastal ecosystems requires more consideration, particularly in the context of projected increases in the frequency and intensity of coastal storms [6].

Summary and call to action
Estuaries serve as critical migratory corridors and host rich and diverse ecosystems. Past studies have considered global-or continental-scale warming in rivers [2], lakes [3], and the ocean [6], as well as coastal atmospheric warming [18]. In contrast, the thermal sensitivities of coastal water bodies to atmospheric and marine climate change are relatively understudied, at least in a coordinated way across spatial scales. Coastal thermal dynamics and patterns are distinct from those for freshwater bodies as they are influenced by complicating factors, including density-dependent thermal stratification from salinity, high-frequency forcing from tides, and short-term disturbances from coastal storms. Collectively, these complicate the dynamics of coastal thermal regimes compared to freshwater. This complexity is likely one reason for the general paucity of studies using process-based models to investigate how coastal waters will thermally respond to atmospheric and marine climate change.
Large-scale monitoring studies of coastal water body warming are also generally lacking, with one rare exception in Australia [19]. Given their relatively low thermal inertia, coastal water bodies often warm at a higher rate than the ocean [e.g. 20]. Thus, existing widespread large-scale ocean thermal monitoring networks, while invaluable for climate change understanding, may be ineffectual for tracking and elucidating warming of semi-enclosed coastal waters.
We call for increased awareness, government support, and coordinated research efforts to investigate the thermal sensitivity of coastal water bodies to climate change. Such efforts could begin in large estuaries with global analysis of satellite thermal imagery, which has been invaluable for studying widespread ocean and lake thermal change. We also call for large-scale (continental and global) coordinated in situ thermal monitoring of coastal lagoons and estuaries which would bridge the established and coordinated aquatic temperature networks in freshwater and marine environments.
Finally, we call for more coordinated processbased modeling efforts through the combination of research disciplines (e.g. hydrology, oceanography, coastal engineering). For large estuaries or lagoons that influence broader continental shelf waters, these models could be nested within larger climate models. For smaller coastal bodies, such models be run independently with climate model output applied at the boundaries of coastal hydrodynamic models that include heat and solute (salt) transport.
Large research networks could tackle the grand challenge of understanding how changes to coastal storms, hydrology, sea levels, and atmospheric heat fluxes will collectively result in estuarine and lagoon thermal changes and ecosystem impacts. The topic of warming estuaries or lagoons is only briefly addressed in the IPCC AR6 ( [21], p. 20), and the IPCC SROCC ( [6], p. 493-494). More attention is paid to changing estuary salinity or the cascading effects of high temperatures, including eutrophication or changing oxygen. Increased research and management attention tied to warming estuaries and lagoons could help identify vulnerable hot spots, inform future thermal restoration or 'renovation' [22] efforts, and underpin future climate change synthesis reports.

Data availability statement
The data that support the findings of this study are openly available at the following URL/DOI: https:// doi.org/10.5683/SP3/PEV45A [23]. Data will be available from 01 March 2023.