Sage Journals: Discover world-class research

Abstract

Change in river water temperature has important consequences for the environment and people. This review provides a new perspective on the topic by evaluating changes in river water temperature for the UK over the 20th century and possible changes over the 21st century. There is limited knowledge of space-time variability in, and controls on, river temperature at the region scale and beyond over the 20th century. There is historical evidence that UK river temperature has increased in the latter part of the 20th century, but low agreement on the attribution of changes to climatic warming because river temperature is a complex, dynamic response to climate and hydrological patterns moderated by basin properties and anthropogenic impacts. Literature is scarce to evaluate changes to UK river temperature in the 21st century, but it appears as likely as not that UK river temperature will increase in the future. However, there are a number of interlinked sources of uncertainty (related to observations, scenarios, process interactions and feedback) that make estimating direction and rate of temperature change for rivers across the UK with confidence very challenging. Priority knowledge gaps are identified that must be addressed to improve understanding of past, contemporary and future river temperature change.

Keywords

adaptation climate change controls and processes review river/stream temperature spatial and temporal heterogeneity thermal dynamics UK water temperature

I Introduction

Water temperature is recognized increasingly by scientists, environment managers and regulators as an important and highly sensitive ‘master’ variable of water quality (Hannah et al., 2008b). Temperature directly influences: distribution (e.g. Boisneau et al., 2008), predator-prey interactions (e.g. Boscarino et al., 2007), survival (e.g. Wehrly et al., 2007), growth rates (e.g. Imholt et al., 2010, 2011; Jensen, 2003), timing of life history events (e.g. Harper and Peckarsky, 2006) and metabolism (e.g. Alvarez and Nicieza, 2005) of aquatic organisms in river systems. Indirectly, temperature controls in-stream processes such as rates of production, nutrient consumption and thus food availability, decomposition (e.g. Ormerod, 2009) and dissolved oxygen concentration (e.g. Sand-Jensen and Pedersen, 2005), which influence ecological processes further. In addition, water temperature is of economic importance for electric power (e.g. Foerster and Lillestam, 2010; van Vliet et al., 2013b), drinking water production (e.g. Ramaker et al., 2005) and fisheries (e.g. Bartholow, 1991; Ficke et al., 2007). Hence, there are clear ecological and socio-economic benefits (Poole and Berman, 2001) to be accrued from: (1) understanding the sensitivity of river temperature to climate and other drivers of change using observations for the 20th century; and (2) assessing possible future river temperature changes in the 21st century to inform management and adaptation strategies (Wilby et al., 2010). As a consequence of the increasingly recognized importance of river temperature, there has been an upsurge in research on this water quality variable (Hannah et al., 2008b).

There are four comprehensive reviews of the river temperature literature. In chronological order, Smith (1972) considers the physical processes driving river temperature variability in near-natural systems and also evaluates human impacts that include thermal pollution. Ward (1985) focuses on the Southern Hemisphere to consider controls on the thermal regime and anthropogenic factors. Caissie (2006) overviews water temperature modelling, natural and human influences on thermal conditions and implications for aquatic ecology. Most recently, Webb et al. (2008) capture renewed interest by evaluating significant advances in river and stream temperature research since 1990. Notably, Webb et al. (2008) identify improving understanding of: (1) thermal heterogeneity at different spatial and temporal scales; and (2) past and future trends as major issues for contemporary river temperature research. Such understanding is required urgently by environment regulators as a first step in assessing how climate changes will alter river systems and interact with other pressures affecting ecological status and societal use of flowing waters.

This article offers a different perspective from other reviews. It aims to evaluate changes in river water temperature for the United Kingdom (UK) over the 20th century and possible changes over the 21st century by synthesizing the peer-reviewed literature. Although the emphasis is on the UK, pertinent research is reviewed to contextualize UK-based information. Throughout, the term ‘river temperature’ refers to the temperature of the water column in the river channel. Riverbed temperature is not considered in detail.

The remainder of this article is structured as follows: state-of the-art understanding of processes, controls and dynamics and drivers of change are reviewed to provide a clear scientific basis for evaluating changes (section II); changes in UK river temperature over the 20th century are evaluated with reference to the wider international literature to assess what has happened (section III); what may happen in the future is speculated based on the very limited number of predictive modelling studies of possible UK temperature over the 21st century with findings contextualized by international research (section IV); and conclusions and emergent priority knowledge gaps are outlined to provide guidance on how to improve understanding of past, contemporary and future projections of river temperature change in the UK (section V).

II Understanding processes, controls, dynamics and drivers of change

1 Processes

River water temperature is controlled by dynamic energy (heat) and hydrological fluxes at the air-water and water-riverbed interfaces (Figure 1; Hannah et al., 2008a). Land and water management impact on these drivers and, thus, modify river thermal characteristics (Webb et al., 2008). Rivers are hierarchical systems (Montgomery, 1999) and therefore for a specific point on a river, water column temperature is determined initially by the mix of water source contributions (surface/shallow subsurface flows, groundwater, snow melt, etc.) and subsequently the energy gained or lost across the water surface and riverbed interfaces as the river flows downstream. Thus, spatial and temporal variability in heat flux and hydrological processes create heterogeneity in river temperature at a range of scales.

Figure 1.

A schematic representation of the energy and hydrological fluxes controlling river water temperature.

Heat transfer within river systems is complex, occurring through a combination of radiation, conduction, convection and advection (Webb and Zhang, 1997). These energy exchanges add and remove heat to and from the river. Inputs may occur by incident short-wave (solar) and longwave (downward atmospheric) radiation, condensation, friction at the channel bed and banks, and chemical and biological processes. Losses may include reflection of solar radiation, emission of longwave (back) radiation and evaporation. Sensible heat and water column-bed energy transfers may cause gains or losses. In addition to these exchanges, energy may be advected by: in/out-flowing channel discharge, hyporheic exchange, groundwater up/down-welling, tributary inflows and precipitation. These heat fluxes may be related together using a heat budget (or energy balance) to give the total energy available to heat or cool river water (Q_n ) at a given point:

Q_{n} = Q_{a} + Q^{*} + Q_{e} + Q_{h} + Q_{b h f} + Q_{f}

where Q_a is advected heat due to groundwater discharge, hyporheic exchange, precipitation, and the longitudinal advective heat flux, Q* is net radiation, Q_e is latent heat, Q_h is sensible heat, Q_bhf is bed heat exchange and Q_f is friction at the bed and banks.

Changes in water temperature (dTw) may be calculated within either an Eularian (e.g. Caissie et al., 2007; Hebert et al., 2011) or a Lagrangian framework (e.g. Gooseff et al., 2005; Leach and Moore, 2011; MacDonald et al., 2014; Rutherford et al., 2004). Within the Eularian framework, dTw is calculated as a function of time (t) (equation 2) using a reference system (i) that is fixed in space and through which water flows:

\frac{d T_{w}}{d t} = \frac{Q_{n (i, t)}}{ρ C D_{(i, t)}}

where $ρ$ is density of water, C is specific heat capacity of water and D is river depth. Within the Lagrangian framework dTw is calculated as a function of space (equation 3) using a reference system that moves with the water:

\frac{d T_{w}}{d x} = \frac{W_{(i)} [Q_{n (i, t)}]}{C * F_{(i, t)}}

where W is width of the stream surface and F is river discharge.

The energy balance is a useful tool for analysis of river temperature processes (e.g. Hannah et al., 2008b; Leach and Moore, 2010; Garner et al., 2014; MacDonald et al., 2014) and for river temperature prediction (e.g. Gooseff et al., 2005; Westhoff et al., 2007; Leach and Moore, 2011; MacDonald et al., 2014). For the UK, river energy balance studies have been conducted in the chalk streams (Webb and Zhang, 1997, 1999, 2004) and regulated rivers (e.g. Evans et al., 1998) of central and southern England, and the Scottish Cairngorm mountains (Hannah et al., 2004, 2008a; Garner et al., 2014).

2 Controls and dynamics

River temperature controls are multivariate and nested at macro-, meso- and micro-scales (Figure 2) (Webb, 1996). Climate drives the thermal regime in rivers and is thus the first-order control on regional patterns in the magnitude and timing of seasonal dynamics (Ward, 1985; Garner et al., 2013). Basin-wide characteristics are second-order controls; water sources, basin aspect and precipitation regime may moderate the influence of climate and thus modify the timing and magnitude of subseasonal water temperature dynamics. Reach-specific controls, which interact with the water column as it moves through the catchment, such as topographic and riparian shading, hyporheic exchanges and localized groundwater contributions, may further moderate the influence of climate on water temperature dynamics. Thus, the cumulative effect of controls at each scale produces river temperature dynamics at a given site and determines its sensitivity to climate (Figure 2).

Figure 2.

Conceptual model of multivariate, nested controls on river temperature that interact to produce site-specific dynamics. Green text indicates controls that moderate dynamics and sensitivity to climate. Red text indicates controls that enhance dynamics and sensitivity to climate.

The large number of potential controls on river temperature means that it is difficult to disentangle their multivariate influence on energy exchange, hydrological processes and, ultimately, water temperature (Figure 2). However, water is typically cooler and less variable at subseasonal scales where runoff is sourced predominantly from groundwater (Figure 3) (Erickson and Stefan, 2000; Kelleher et al., 2012; O’Driscoll and DeWalle, 2006; Tague et al., 2007; Webb and Nobilis, 2007; Webb et al., 2008; Garner et al., 2013), or snow melt (Blaen et al., 2013), or where frequent precipitation maintains thermal capacity (Webb and Nobilis, 1997). At smaller scales, solar (shortwave) radiation input is reduced by channel orientation, and shading effects of topography (e.g. incised channels; Webb and Zhang, 1997) and riparian vegetation (e.g. forest cover, as reviewed by Moore, 2005). Finally, streamflow contributions from hyporheic exchange and/or localized groundwater contributions moderate minimum and maximum temperatures (Malcolm et al., 2005; Garner et al., 2014).

Figure 3.

River water temperature at Trowbridge on Biss (BFI = 0.17) and Amesbury on Avon (BFI = 0.97). (a) Diurnal regimes for hottest seven days in 2007. (b) Diurnal regimes for coldest seven days in 2007. (c) Annual regimes for 2007. (d) Location of river temperature gauging stations within the UK. BFI denotes the Base Flow Index, which is a measure of the proportion of runoff derived from stored sources. Higher BFI values represent the more permeable catchment, i.e. higher water storage in bedrock, superficial deposits and soils.

3 Drivers of change

There are a number of drivers of change that influence controls on the energy balance and water fluxes and so river temperature. The impacts of forestry practice are the best documented and the highly variable findings are synthesized and evaluated by Moore et al. (2005), with a UK focus provided by Hannah et al. (2008a). Land-use change from riparian forest to grassland for agriculture may elevate river temperature in summer (e.g. Isaak and Hubert, 2001; Li et al., 1994). Urbanization is associated with increased river temperature compared with rural environments due to runoff of water across warmed paved surfaces (e.g. Herb et al., 2008; Kaushal et al., 2010; Hester and Bauman, 2013; Xin and Kinouchi, 2013) and channel widening/vegetation removal (i.e. exposure) to increase flow conveyance (Klein, 1979). Direct flow augmentation and abstraction change river thermal capacity, hence river temperature (Poole and Berman, 2001). Heated effluent from power plants and other point sources may have a profound warming impact (e.g. Maderich et al., 2008). The thermal effects of reservoirs are well documented for the UK, the USA and elsewhere (e.g. Webb and Walling, 1997; Olden and Naiman, 2010; and more recently Casado et al., 2013, who report increases in temperature minima and decreases in temperature maxima at a range of spatial scales).

In addition to these drivers of change, it may be anticipated that climate change will have direct and indirect impacts on river temperature. Direct effects may occur due to shifts in the energy exchange and hydrological processes that determine river temperature. For example, longwave radiation from the atmosphere, sensible heat (e.g. Ramanathan, 1981; Wild et al., 1997; Andrews et al., 2009) and groundwater temperature are anticipated to increase (e.g. Taylor and Stefan, 2009; Kurylyk et al., 2013, 2014), while groundwater contributions to streamflow (e.g. Kurylyk et al., 2014) and summer streamflow are anticipated to decrease (e.g. Cappel et al., 2013). Indirectly, river temperature may be affected by climate-induced alteration of riparian land use (Hrachowitz et al., 2010).

III What has happened?

This section reviews changes in UK water temperature over the 20th century and draws on the international literature to set this information in a wider context.

Research on river temperature in the UK has investigated (Table 1):

point-scale heat fluxes that fundamentally control water temperature;

reach-scale variability in the water column and riverbed;

effects of forestry;

spatial and temporal dynamics across river networks.

Table 1.

Research on river water temperature in the UK: scales, controls and processes.

Author(s)	Study scale and location	Study length	Key findings	Controls and processes
Point-scale heat fluxes controlling river temperature
Webb and Zhang (1997)	11 sites in River Exe basin, Devon, southwest England	495 days across 18 study windows	Averaged over the entire data set non-advective energy gain contributions = net radiation (56%), friction (22%), sensible heat (13%), condensation (6%) and bed heat flux (3%). Non-advective energy loss contributions = net radiation (49%), evaporation (30%), sensible heat (11%) and bed heat flux (10%). Magnitude and relative importance of heat fluxes varied in time and space.	Channel morphology, valley topography, riparian vegetation, substratum composition hydrological conditions, and river regulation influence variability and magnitude of heat fluxes over time and between sites.
Evans et al. (1998)	1 site in English Midlands	8 days	On average, over 82% of the total energy transfers occurred at the air-water interface with 15% at the channel bed-water interface. Heat exchange at the channel bed varied considerably (max. 24%) in response to varying bed thermal, and periphyton and macrophyte cover.	Temporal heterogeneity in bed heat flux caused by varying riverbed albedo (due to seasonal changes in periphytion, macrophyte and silt cover) and relative contribution of surface- versus groundwater.
Webb and Zhang (1999)	1 site in River Exe basin, southwest England	1 year	Net radiation contributed ˜90% of energy gains during summer months. Sensible heat enhanced during summer. Bed heat flux reduced considerably at 1 site where weed growth extensive.	Sensible heat transfers enhanced in groundwater-fed streams during summer due to their lower water cf. air temperature. Macrophytes (lower thermal conductivity cf. sediment) decrease bed heat fluxes.
Hannah et al. (2004)	1 site, Cairngorms, northeast Scotland	7 months (autumn-spring)	Streambed (atmosphere) dominant energy source (sink) for heating (cooling) channel water.	Groundwater upwelling advects heat into the bed. Friction important heat source in winter. Sensible heat primary atmospheric heat source when radiative transfers limited.
Webb and Zhang (2004)	4 sites in River Exe basin, southwest England	1 year	Sensible heat source (sink) in summer (winter). Bed heat sink (source) in summer (winter). Friction heat source. Evaporation heat sink. Differences between forested and non-forested sites, especially in terms of radiation.	Forest canopy reduces shortwave-radiation. Tree trunks and branches in deciduous forests cause shading effects even during winter.
Hannah et al. (2008b)	1 open moorland and 1 semi-natural forest site, Cairngorms, northeast Scotland	2 years	Net radiation greater summer (less winter) for moorland cf. forest. Magnitude and variability of turbulent fluxes greater for moorland cf. forest.	Forest canopy shades in summer (but offsets net radiation loss in winter due to longwave emission) and sheltering effects on sensible and latent heat.
Reach-scale temperature variability in the water-column and riverbed
Carling et al. (1994)	Channel cross-sections, River Severn	<1 day	Water temperature in dead zones of meanders 2°C warmer cf. flowing main channel.	Shallower, slower flowing water warms more rapidly than deeper, faster flowing thalweg (i.e. atmospheric equilibration time).
Clark et al. (1999)	202 channel cross-sections, River Frome and Bere Stream, southwest England	<1 day	0.2–0.4°C vertical and/ or horizontal temperature contrasts across 78% of channel cross-sections.
Hannah et al. (2009)	Riffle-pool sequence, River Tern, English Midlands	22 months	Hyporheic temperature cooler (warmer) than water column in summer (winter), with convergence in spring and autumn. Temperature varies across and between riffles.	River geomorphology alters groundwater-surface water interactions, hence thermal dynamics.
Krause et al. (2011)	Riffle-pool sequence, River Tern, English Midlands	9 months (summer-winter)	Streambed temperature variability 0.75°C (3°C) over 16 m (0.4 m) longitudinally (vertically).
Forest effects on river temperature
Crisp (1997)	5 sites in 2 catchments, upper-Severn, mid-Wales (2 sites pre-clear-felled; 3 sites post-clearfelled)	15–52 months	Mean annual water temperature reduced by 0.4 °C and daily range lower pre-clearfelling. Greatest effects in summer.
Weatherley and Ormerod (1990)	6 sites across clearfelled and afforested locations, upper River Severn, mid-Wales	1–2 years	Mean daily water temperature lower (higher) in forest cf. moorland in spring and summer (winter).
Stott and Marks (2000)	1 site pre-clearfelling and 1 site post-clearfelling, River Severn, mid-Wales	28 months	Monthly mean temperature increased by 7°C in July and 5.3°C during August post-clearfelling.
Malcolm et al. (2004)	5 sites situated in open moorland and one in mixed (deciduous, coniferous) woodland	3 years	Riparian woodland reduced diurnal variability and extremes of temperature.
Malcolm et al. (2008)	2 sites situated in open moorland sites and three sites in mixed woodland	35 months	Under forest: amplitude of annual temperature regime reduced, daily mean and maximum decreases, in daily minimum increases, and diurnal variability reduced.
Broadmeadow et al. (2011)	14 forested and 4 open-moorland sites in 2 catchments, New Forest, southern England	3 years	Largest differences between forested and open sites during summer. Effects observed at canopy shading of 20–40%.
Imholt et al. (2013)	7 sites with the Tanar and 5 sites within the Girnock catchment, Dee river basin, northeast Scotland	13 months	Presence of semi-natural woodland reduces summer maximum (by up to 4°C) and dial temperature range (by up to 6.6°C).
Garner et al. (2014)	1 open moorland and 1 semi-natural forest site, Cairngorms, northeast Scotland	7 years	Inter-annual variability is reduced under semi-natural forest canopy.	Forest effective at reducing net energy inputs to the water-column and reducing water temperature when solar radiation gains are high and/or wind speeds are low.
Spatial and temporal dynamics across river networks
Webb and Walling (1992)	3 sites on tributaries with contrasting land use, River Exe basin, southwest England	14 years	Significant increase in water temperature 0.05 to 0.092°C y^-1 associated with air temperature increases and removal of riparian vegetation shading.	Climate and land use/land cover influence long-term water temperature trends.
Crisp et al. (1982)	7 streams (two watercress beds) in southern England	1–8 years	Surface water fed streams lower annual mean but greater cycle amplitude cf. groundwater fed streams.	Increased groundwater inputs, longer hydrological residence times and increased thermal capacity buffer influence of climate on water temperature.
Webb and Walling (1986)	17 sites in River Exe basin, southwest England	5 years	Thermal variability buffered for larger basin areas and longer hydrological residence times.	As above
Webb et al. (2003)	Four sites draining basins of contrasting area, River Exe basin, southwest England	5 years	Air-water temperature relationships weaken with increased thermal capacity and longer residence times.
Hrachowitz et al. (2010)	25 sites, River Dee, northeast Scotland	2 years	Small, non-forested, inland, upland sites most sensitive to air temperature. Large, forested, lowland sites least sensitive air temperature.
Johnson et al. (2014)	36 sites, Rivers Dive and Manifold, English Peak District	1 year	Upstream riparian shade and perennial spring inflows reduce sensitivity of water temperature to air temperature.
Garner et al. (2013)	88 sites across England and Wales	18 years	Sensitivity to air temperature reduced where groundwater contributions to runoff smallest or water exposure time to the atmosphere greatest.
Imholt et al. (2013)	26 sites across nested scales, River Dee, northeast Scotland	2 years	Least spatial variability at the subreach scale (0.3°C difference between riffles and pools) with greatest variability between tributaries (8.1°C difference in diurnal range) and subcatchments.	Thermal variability scale-dependent and related to river morphology, land use, altitude and forest cover. Notably, similar thermal patterns may be driven by different sets of physical controls.

This research has yielded significant knowledge about UK river temperature patterns, controlling factors and driving mechanisms. However, most previous UK studies have been: (1) restricted to the sub-basin scale; (2) short-term or seasonally constrained to summer with very few long-term year-round and multi-year studies; and (3) focused primarily on unravelling the role of site-specific factors and conditions. Consequently, there is a lack of research on spatial and temporal variability and controls on river temperature at the inter-basin to region scale and beyond over long time periods during the 20th century (notwithstanding exceptions below).

1 International studies

Internationally, long-term and wide spatial coverage river temperature studies are extremely rare because of a lack of reliable unbroken records. Most river temperature time series span only a few years (Caissie, 2006). Webb et al. (2008) reviewed the few long-term studies and comment that, despite most studies indicating warming, ‘considerable variation in the magnitude of river temperature rise is apparent between and within river systems’ (p. 910). For the USA, Webb (1996) reported trends for 90 out of 364 rivers studied from 1974 to 1981. However, when temperature data are flow-adjusted, only 29 rivers exhibit significant trends with 22 rivers warming as flows declined and 7 rivers cooling with increased flows. Interestingly, clear regional patterns related to hydroclimatological regimes did not emerge due to local effects of forestry, land use, flow regulation and heated effluent discharge (Webb, 1996). More recently, Kaushal et al. (2010) analysed data for 40 major US rivers and showed significant, long-term warming trends for 20 rivers that were significantly correlated with air temperature. Rates of warming were most rapid in urban areas, although no further detailed interpretation is provided of controls.

Webb and Nobilis (2007) conducted a landmark study of changes in the thermal regime of Austrian rivers between 1901 and 2001. Annual mean river temperatures were found to rise by up to +1.5°C over the 20th century and to track patterns in the North Atlantic Oscillation index (NAOI, i.e. a large-scale climate diagnostic based on the atmospheric pressure differential between the Icelandic Low and Azores High that is associated with heat and moisture advection across Europe; e.g. Jones et al., 1997). Similarly, Hari et al. (2006) analysed data for 25 Swiss alpine rivers to find substantial warming over the last quarter of the 20th century, with a step change in 1987 linked to the shift in the NAOI to a highly positive phase. For the East Creek in British Columbia, Canada, Kiffney et al. (2002) found the highest winter temperature to occur in El Niño years.

Attribution of warming trends to climate patterns and change is difficult. For example, Kinouchi et al. (2007) suggest that warming of spring and winter temperature for central Tokyo from 1978 to 1998 is related to increases in discharge of heated wastewater, not climate. Petersen and Kitchell (2001) demonstrate that July temperature for the Columbia and Snake rivers in the Pacific Northwest, USA, have not responded to climate shifts but rather increasing dam construction since 1930. Impoundment has been shown also to decrease mid open-water season temperature for the Lena River, Siberia (Yang et al., 2005).

Kelleher et al. (2012) investigated controls on the thermal sensitivity of Pennsylvania rivers using linear and non-linear regression and found Strahler stream order and baseflow contribution to be the primary controls. For small streams, baseflow contribution was the primary determinant of sensitivity, with greater contributions from baseflow resulting in decreased sensitivity; whereas, for large rivers, thermal sensitivity increased with stream size, explained by accumulated heat through the stream network. These findings resonate with other studies reporting the importance of groundwater in buffering climatic sensitivity of river temperature (Mohseni et al., 1999; Erickson and Stefan, 2000; O’Driscoll and DeWalle, 2006). However, small headwater streams are often forested and/or drain frozen water stores (snowpacks and glaciers) both of which factors also reduce direct climate sensitivity (Caissie, 2006). In contrast to Kelleher et al. (2012), other studies have demonstrated weaker air-water temperature associations due to increased thermal capacity and response lag times for larger downstream river reaches (e.g. Webb et al., 2003). Hence there is some, but no definitive, consensus on moderation by basin properties of river temperature sensitivity to climate-related change/variability, which suggests scale of investigation and environmental context are important.

Thus, it is evident from the international literature that variability in river temperature over the 20th century is a complex, dynamic response to climate patterns and hydrological change moderated by basin properties and anthropogenic impacts. It should be noted also that, although the majority of studies illustrate a warming trend over recent decades, some studies show river cooling (Langford, unpublished data cited by Webb et al., 2008; Smith, 1995; Arismendi et al., 2012; Isaak et al., 2012).

2 UK studies

For the UK, there are only a handful of long-term water temperature studies that mainly provide very sparse national coverage. Water temperature trends in the River Exe basin, Devon, were analysed by Webb and Walling (1992) over a 14-year study period (1977–1990) and revealed rising water temperature to match patterns in air temperature and land use (i.e. removal of riparian shade). The 1975–1976 drought distorts analyses of water temperature trends at the end of the 20th century for many UK rivers (Webb, 1996) as the low flows (reduced thermal capacity) and high energy inputs cause river temperature to peak. Such inter-annual variability raises questions about the search to detect linear trends without attribution of causes. Langan et al. (2001) analysed a 30-year record for a Cairngorm river and found no significant trend in annual water temperature, but an increase in spring and winter maxima and spring mean water temperature. These increases were attributed to a reduction in spring snow melt contributions to flow. For mid-Wales, Durance and Ormerod (2007) suggest that forest and moorland river temperature have warmed by +1.5°C and +1.7°C from 1981 to 2005, respectively, when variation in the NAOI is factored out.

To date, only two published studies (Garner et al., 2013; Orr et al., 2014) have explored long-term river temperature across the UK. Both use data from the Environment Agency Surface Water Temperature Archive (Orr et al., 2010) that spans England and Wales for the latter part of the 20th century. There is no national-scale river temperature database for Scotland or Northern Ireland. The Archive contains over 28,000 records but <50% are >10 years long (Figure 4). Most data were collected opportunistically during hydrographic or water quality sampling campaigns (Orr et al., 2014). Sensitivity analysis identifies 18 years (1989–2006) as the optimum timespan to maximize spatial network extent (Figure 5). Sampling intervals are fortnightly to monthly predominantly (Figure 6). Figure 7 displays spatial patterns in summary statistics for mean, minimum, maximum and range of water temperature at 2832 sites with >250 temperature samples between 1990 and 2007 (Cooper, 2010).

Figure 4.

Length of records (years) for sites in the Environment Agency of England and Wales’ Freshwater Temperature Archive.

Figure 5.

Spatial distribution of records of varying lengths in the Environment Agency of England and Wales’ Freshwater Temperature Archive.

Figure 6.

Environment Agency of England and Wales’ Freshwater Temperature Archive mean monthly sampling frequency (1989–2006) by Environment Agency region.

Figure 7.

Maps of (A) mean, (B) minimum, (C) maximum and (D) range in river water temperature for 2832 sites with >250 samples between 1990 and 2007 from Environment Agency of England and Wales’ Freshwater Temperature Archive. For each variable, the England and Wales mean (μ) and standard deviation (σ) are provided as text; and each site is colour-coded by z-scores (blue for negative site z-values, below the England and Wales average; red for positive site z-values, above the average).

Orr et al. (2014) assessed temporal trends using 2773 sites from the Archive and reported significant spatial variation. However, mean annual water temperatures increased between 1990 and 2006 at an average rate of 0.03°C year^-1 (Figure 8), comparable to changes in UK air temperature over the same period (as reported by Jenkins et al., 2009). Garner et al. (2013) investigated spatiotemporal variability in the shape (timing) and magnitude (size) of long-term annual river temperature regimes at 88 sites across England and Wales, and also quantified sensitivity of inter-annual water to air temperature regimes at 39 sites to infer the influence of basin properties. Four river temperature magnitude regimes were identified (Figure 9) with cooler regimes in the north and warmer regimes in the south. Central southern river temperature regimes, overlying Chalk aquifers, had reduced seasonality. River basin properties appeared to modify inter-annual air-water temperature links with climatic sensitivity the most where (1) groundwater contributions to runoff (characterized by basin permeability) were smallest and (2) water exposure times to the atmosphere (estimated by basin area) were greatest (Figure 10).

Figure 8.

Estimated mean annual river temperature change (°C/decade) between 1 January 1990 and 31 December 2006. Environment Agency regions: NW = North West; NE = North East; MD = Midlands; WA = Wales; TH = Thames; SO = Southern; SW = South West; AN = Anglian.

Figure 9.

Map of England and Wales showing long-term average river water temperature magnitude regime classes.

Figure 10.

Box plots of (a) air-water temperature sensitivity index (SI) strength, (b) mean basin permeability, (c) basin area and (d) mean basin elevation for river temperature stations in each region. See Figure 7 for spatial distribution of stations by regions. SI values close to 0 indicated greater river temperature sensitivity to air temperature than those closer to 1.

IV What may happen in the future?

Climate change is expected to alter future river thermal regimes (Huguet et al., 2008); however, this is a much under-researched topic (Webb et al., 2008). Modellers of river temperature under climate change use mainly statistical associations between air and water temperature (e.g. Morrill et al., 2005; Mantua et al., 2010; studies reviewed below), although deterministic heat transport studies are emerging (e.g. Gooseff et al., 2005; Meisner, 1990; Sinokrot et al., 1995; Stefan and Sinokrot, 1993; van Vliet et al., 2013a; MacDonald et al., 2014). However, the most comprehensive work on climate-change impacts on river temperature has been conducted in North America and at coarse spatial scales.

1 International studies of future change

a Statistical models

Analyses of historical records indicate that the strength of air-water temperature relationships increase from subdaily to monthly resolution, but weaken for annual data (Webb et al., 2008) because water temperature shows less year-to-year variability than air temperature (Pilgrim et al., 1998; Erickson and Stefan, 2000; Webb et al., 2003). Thus, most predictive modelling studies concern weekly or monthly averages. For example, Mohseni et al. (2003) projected changes in river temperature for 803 locations across the contiguous USA using a non-linear air-water temperature relationship. Projections were based on air temperature from the Canadian Centre of Climate GCM for a doubling of atmospheric CO₂ concentration. Mean annual stream temperature increased significantly by +3.1 ± 1.7°C on average for 764 stations, maximum weekly temperature increased significantly by +2.2 ± 1.9°C, on average, at 399 stations, and minimum weekly stream temperature increased significantly by +2.3 ± 1.1°C, on average, at 455 stations. In terms of geographical patterns, greater maximum weekly river temperature increases were observed in the southern than the northern USA. In a continental European study (Danube at Linz, Austria), Webb and Nobilis (1994) predicted an increase in monthly mean river temperature of +1.2–2.1°C for an air temperature increase of +2.3–3.4°C and a 15% reduction in flow between June and September by 2030.

At the global scale, van Vliet et al. (2011) tested the performance of a non-linear water temperature model (modified from Mohseni et al., 1998, by using daily air temperature and river flow as predictors) for 157 river stations from 1980 to 1999. Using river discharge as an input variable improved predictive power at >87% of stations (cf. air temperature alone), particularly during heat wave and low flow conditions, at stations with high winter discharges and low summer discharges, and for large rivers. Inclusion of discharge yielded modest increase in model performance for stations influenced by reservoirs or snow melt. Sensitivity analysis indicated increases in annual mean temperature of +1.3°C, +2.8°C and +3.6°C under air temperature increases of +2.0°C, +4.0°C and +6.0°C, respectively. Discharge decreases of –20% and –40% increased annual mean water temperatures by +0.3°C and +0.8°C on average, respectively; while a discharge increase of +20% reduced annual mean water temperatures by –0.2°C on average. Importantly, an increase in air temperature of +4°C combined with a decrease in discharge of –40% resulted in a greater increase in river temperature than an increase in air temperature of +6°C alone, which emphasizes the importance of considering hydrological as well as climate change.

Climate change will alter hydrology (surface water and groundwater flows) and basin/site characteristics (e.g. riparian vegetation), and so moderate river temperature (Morrison et al., 2002; Kurylyk et al., 2014). Simple air-water temperature associations, especially when air temperature is taken from sites distant from the river basin of interest, may lead to poor future predictions due to failure to capture local controls. As well as spatial differences, river temperature responses may vary seasonally with research in the north-central USA suggesting spring-summer warming may be less marked than autumn cooling (Stefan and Sinokrot, 1993). These seasonal shifts may reflect future change in influential climatic variables (see section II) and/or seasonal contrasts in the sensitivity of rivers to the effects of climate change (Webb, 1996).

b Deterministic models

Deterministic models offer great potential to incorporate local climatological, hydrological and land-use controls, and their anticipated future conditions, and model at finer spatial and temporal resolution. Gooseff et al. (2005) assessed the effects of anticipated future air temperature and solar radiation magnitude, and observed increases in the duration that water temperatures exceeded ecologically important thresholds in the lower Madison River, Montana. For a headwater tributary of the Canadian Rockies, MacDonald et al. (2014) used a coupled process-based hydrometeorological and stream temperature model forced by projected air and groundwater temperatures to investigate seasonal response of summer water temperature for the 2050s and 2080s. Projected changes in water temperature were driven by changes in riverflow as opposed to air temperature; winter and early-spring water temperature decreased by up to –1.5°C whereas late-spring and summer temperature increased by up to +1.3°C. At 0.5×0.5° resolution globally, van Vliet et al. (2013a) coupled a physically based hydrological model with a one-dimensional water temperature model to predict changes to 2071–2100 water temperature relative to 1971–2000. Mean and maximum (defined as 95th percentile) water temperatures were projected to increase by up to 1.2 and 2.2°C, respectively.

2 UK studies of future change

For the UK, there are very few predictive modelling studies of river temperature under climate change. Hence, there is scant evidence to evaluate what may happen to UK river temperature in the 21st century. At 36 UK river sites, Webb and Walling (1992) modelled a rise of +1–3.6°C in monthly mean water temperature by 2050, given a scenario of monthly mean air temperature increase of +2–3°C. The magnitude of increase was dependent on geographical location and river basin characteristics, with greater groundwater contributions and forest cover modulating responsiveness to climate change. Subsequently, Webb and Walsh (2004) modelled 27 UK river monitoring sites by incrementing present conditions (defined by applying air-water temperature relationships to a baseline annual cycle of weekly mean air temperature averaged over >5 years of record) for air temperature increases predicted by low and high warming UKCIP02 scenarios for 2020, 2050 and 2080. Results indicate that warming will be modest to very significant depending on the scenario applied; and warming will be moderated by site and basin characteristics. A comparison of four UK rivers revealed the importance of groundwater in moderating river temperature response to climate change, with surface water-fed rivers on non-calcareous lithology more responsive (+1.8–3.6C) than spring-fed, chalk rivers (+1.1–2.2 C) to an increase of +2–4 in monthly mean air temperature (Mackey and Berrie, 1991).

Van Vliet et al. (2013a) (see section IV,1,b) performed the only process-based assessment of future water temperature that includes the UK to date. Maximum warming of between 1.2 and 1.4°C was predicted for mean water temperature between 2071 and 2100. However, increases in maximum temperature, spatial and seasonal variation throughout the UK were not reported. Furthermore, the 0.5×0.5° resolution of the model did not represent the effects of finer-scale hydrological and land-use controls; consequently, the spatial patterns in future river temperature may differ from those projected for driving climate variables (e.g. Webb and Walsh, 2004).

V Conclusions and knowledge gaps

In addition to existing drivers of change, it may be anticipated that climate change will have direct (i.e. shifts in the energy exchange and hydrological processes; see section II) and indirect impacts on river temperature (e.g. riparian land use, human response to reduced water security). Therefore, improved understanding of (1) sensitivity of river temperature to climate and other drivers of change (based on 20th-century records), and (2) possible future river temperature change (based on projections for the 21st century) is critical to inform management and adaptation strategies in the UK and elsewhere.

Most UK river temperature investigations have been restricted to the (sub-)basin scale and by relatively short observational records. Consequently, there is limited knowledge of spatial and temporal variability of, and the controls on, river temperature at the inter-basin to regional scale and beyond over long time periods during the 20th century. There is a medium level of agreement (and as likely as not) that UK river temperature increased in the latter part of the 20th century, but low agreement on the attribution of these river temperature changes to climatic warming because river temperature is a complex, dynamic response to climate and hydrological patterns moderated by basin properties and anthropogenic impacts (see section III). Uncertainties exist over past changes related to historical river temperature data (e.g. irregularity of sampling, differences in instrumentation, variable measurement bias), different timespans of different authors’ analyses, and patchy spatial coverage of the UK.

Literature is scarce to evaluate what may happen to UK river temperature in the 21st century. Based on a meagre UK evidence base and research conducted elsewhere, there appears a good level of agreement (and as likely as not) that UK river temperature will increase in the 21st century (see section IV). That being said, there are a number of interlinked sources of uncertainty that mean estimating rates of temperature change for rivers across the UK with confidence is beyond current knowledge. As well as scenario uncertainties, there are uncertainties about future interactions and feedbacks between climate, hydrology and basin/site properties (e.g. riparian land use, management human response) that may cause river temperature patterns to diverge from those projected for climate drivers.

From this review, a number of priority knowledge gaps are emergent that include:

establishing a UK-wide monitoring network to collect long-term records with nested spatial coverage and including benchmark near-natural river basins;

gathering metadata to provide information on site-, reach- and catchment-scale specific conditions that influence river temperature;

robust analysis of existing UK data (including collation of data from sources not currently in the Environment Agency Surface Water Temperature Archive) to detect river temperature change and account for confounding effects and inconsistencies in sampling resolution, length of time series, and spatial bias in site locations;

better understanding of space and time scales of river thermal heterogeneity;

systematic attribution of river temperature patterns to drivers of change;

quantification of river temperature sensitivity to drivers of change, including comprehensive assessment of climatic sensitivity;

improved future projection of river temperature change accounting of sources of uncertainty.

Addressing these research needs will improve understanding of past, contemporary and future river temperature change in the UK and internationally.

Footnotes

Acknowledgements

We are grateful for the comments of an anonymous reviewer, which improved the manuscript.

Funding

This work was funded by the UK Department for Environment, Food and Rural Affairs (DEFRA) (project WC1052) and supported by the Natural Environment Research Council and the Environment Agency.

References

Alvarez

Nicieza

(2005) Compensatory response ‘defends’ energy levels but not growth trajectories in brown trout, Salmo trutta L. Proceedings of the Royal Society B: Biological Sciences 272: 601–607.

Andrews

Forster

Gregory

(2009) A surface energy perspective on climate change. Journal of Climate 22: 2557–2570.

Arismendi

Johnson

Dunham

. (2012) The paradox of cooling streams in a warming world: Regional climate trends do not parallel variable local trends in stream temperature in the Pacific continental United States. Geophysical Research Letters 39: L10401.

Bartholow

(1991) A modelling assessment of the thermal regime for an urban sport fishery. Environmental Management 15: 833–845.

Blaen

Hannah

Brown

. (2013) Water temperature dynamics in High Arctic river basins. Hydrological Processes 27: 2958–2972.

Boisneau

Moatar

Bodin

. (2008) Does global warming impact on migration patterns and recruitment of Allis shad (Alosa alosa L.) young of the year in the Loire River, France? Hydrobiologia 602: 179–186.

Boscarino

Rudstam

Mata

. (2007) The effects of temperature and predator-prey interactions on the migration behavior and vertical distribution of Mysis relicta . Limnology and Oceanography 52: 1599–1613.

Broadmeadow

Jones

Langford

TEL

. (2011) The influence of riparian shade on lowland stream water temperatures in southern England and their viability for brown trout. River Research and Applications 22: 226–237.

Caissie

(2006) The thermal regime of rivers: A review. Freshwater Biology 51: 1389–1406.

10.

Caissie

Satish

El-Jabi

(2007) Predicting water temperatures using a deterministic model: Application on Miramichi River catchments (New Brunswick, Canada). Journal of Hydrology 336 (3), 303–315.

11.

Cappel

Tetzlaff

Soulsby

(2013) Will catchment characteristics moderate the projected effects of climate change on flow regimes in the Scottish Highlands? Hydrological Processes 27: 687–699.

12.

Carling

Orr

Glaister

(1994) Preliminary observations and significance of dead zone low structure for solute and fine particle dynamics. In: Beven

Chatwin

Millbank

(eds) Mixing and Transport in the Environment. Chichester: Wiley, 139–157.

13.

Casado

Hannah

Peiry

J-L

. (2013) Influence of dam-induced hydrological regulation on summer water temperature: Sauce Grande River, Argentina. Ecohydrology 6: 523–535.

14.

Clark

Webb

Ladle

(1999) Microthermal gradients and ecological implications in Dorset rivers. Hydrological Processes 13: 423–438.

15.

Cooper

(2010) Stream and river temperature response to vegetation across England and Wales: A remote sensing approach. Unpublished M.Phil. dissertation, University of Cambridge, 68 pp.

16.

Crisp

(1997) Water temperature of Plynlimon streams. Hydrology and Earth System Sciences 1: 535–540.

17.

Crisp

Matthews

Westlake

(1982) The temperatures of nine flowing waters in Southern England. Hydrobiologia 89: 193–204.

18.

Durance

Ormerod

(2007) Climate change effects on upland stream macroinvertebrates over a 25-year period. Global Change Biology 13: 942–957.

19.

Erickson

Stefan

(2000) Linear air/water temperature correlations for streams during open water periods. American Society of Civil Engineers, Journal of Hydrologic Engineering 5: 317–321.

20.

Evans

McGregor

Petts

(1998) River energy budgets with special reference to river bed processes. Hydrological Processes 12: 575–595.

21.

Ficke

Myrick

Hansen

(2007) Potential impacts of climate change on freshwater fisheries. Reviews in Fish Biology and Fisheries 17: 581–613.

22.

Foerster

Lillestam

(2010) Modelling thermoelectric power generation in view of climate change. Regional Environmental Change 10 (4): 327–338.

23.

Garner

Hannah

Sadler

. (2013) River temperature regimes of England and Wales: Spatial patterns, inter-annual variability and climatic sensitivity. Hydrological Processes. doi: 10.1002/hyp.9992.

24.

Garner

Malcolm

Sadler

. (2014) Inter-annual variability in the effects of riparian microclimate, energy exchanges and water temperature of an upland Scottish stream. Hydrological Processes. doi: 10.1002/hyp.10223.

25.

Gooseff

Strzepek

Chapra

(2005) Modelling the potential effects of climate change on water temperature downstream of a shallow reservoir, Lower Madison River, MT. Climatic Change 68: 331–353.

26.

Hannah

Malcolm

Bradley

(2009) Seasonal hyporheic temperature dynamics over riffle bedforms. Hydrological Processes 23: 2178–2194.

27.

Hannah

Malcolm

Soulsby

. (2004) Heat exchanges and temperatures within a salmon spawning stream in the cairngorms, Scotland: Seasonal and sub-seasonal dynamics. River Research and Applications 20: 635–652.

28.

Hannah

Malcolm

Soulsby

. (2008a) A comparison of forest and moorland stream microclimate, heat exchanges and thermal dynamics. Hydrological Processes 22: 919–940.

29.

Hannah

Webb

Nobilis

(2008b) River and stream temperature: Dynamics, processes, model and implications – preface. Hydrological Processes 22: 899–901.

30.

Hari

Livingstone

Siber

. (2006) Consequences of climatic change for water temperature and brown trout populations in Alpine rivers and streams. Global Change Biology 12: 10–26.

31.

Harper

Peckarsky

(2006) Emergence cues of a mayfly in a high-altitude stream ecosystem: Potential response to climate change. Ecological Applications 16: 612–621.

32.

Hebert

Caissie

Satish

. (2011) Study of stream temperature dynamics and corresponding heat fluxes within Miramichi River catchments (New Brunswick, Canada). Hydrological Processes 25: 2439–2455.

33.

Herb

Janke

Mohseni

. (2008) Thermal pollution of streams by runoff from paved surfaces. Hydrological Processes 22: 987–999.

34.

Hester

Bauman

(2013) Stream and retention pond thermal response to heated summer runoff from urban impervious surfaces. Journal of the American Water Resources Association 49: 328–342.

35.

Hrachowitz

Soulsby

Imholt

. (2010) Thermal regimes in a large upland salmon river: A simple model to identify the influence of landscape controls and climate change on maximum temperatures. Hydrological Processes 24: 3374–3391.

36.

Huguet

Parey

Dachuna-Castelle

. (2008) Is there a trend in extremely high river temperature for the next decades? A case study for France. Natural Hazards and Earth Systems Sciences 8: 67–79.

37.

Imholt

Gibbins

Malcolm

. (2010) Influence of riparian cover on stream temperatures and the growth of the mayfly Baetis rhodani in an upland stream. Aquatic Ecology 44: 669–678.

38.

Imholt

Malcolm

Bacon

. (2011) Does diurnal temperature variability affect growth in juvenile Atlantic salmon Salmo salar? Journal of Fish Biology 78: 436–448.

39.

Imholt

Soulsby

Malcolm

. (2013) Influence of scale on thermal characteristics in a large montane river basin. River Research and Applications 29 (4): 403–419.

40.

Isaak

Hubert

(2001) A hypothesis about factors that affect maximum summer stream temperatures across montane landscapes. Journal of the American Water Resources Association 37: 351–366.

41.

Isaak

Wollrab

Horan

. (2012) Climate change effects on stream and river temperature across the northwest US from 1980–2009 and implications for salmonids fishes. Climatic Change 113: 499–524.

42.

Jenkins

Perry

Prior

(2009) The Climate of the UK and Recent Trends. Exeter: Met Office Hadley Centre.

43.

Jensen

(2003) Atlantic salmon (Salmo salar) in the regulated River Alta: Effects of altered water temperature on parr growth. River Research and Applications 19: 733–747.

44.

Johnson

Wilby

Toone

(2014) Inferring air-water temperature relationships from river and catchment properties. Hydrological Processes 28: 2912–2928.

45.

Jones

Jónsson

Wheeler

(1997) Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and south-west Iceland. International Journal of Climatology 17: 1433–1450.

46.

Kaushal

Likens

Jaworski

. (2010) Rising stream temperatures in the United States. Frontiers in Ecology and the Environment 8: 461–466.

47.

Kelleher

Wagener

Gooseff

. (2012) Investigating controls on the thermal sensitivity of Pennsylvania streams. Hydrological Processes 27: 771–785.

48.

Kiffney

Bull

Feller

(2002) Climatic and hydrologic variability in a coastal watershed of southwestern British Columbia. Journal of the American Water Resources Association 38: 1437–1451.

49.

Kinouchi

Yagi

Miyamato

(2007) Increase in stream temperature related to anthropogenic heat input from urban wastewater. Journal of Hydrology 335: 78–88.

50.

Klein

(1979) Urbanisation and stream quality impairment. Water Resources Bulletin 15: 948–963.

51.

Krause

Hannah

Blume

(2011) Interstitial pore-water temperature dynamics across a pool-riffle-pool sequence. Ecohydrology 4: 549–563.

52.

Kurylyk

Bourque

CP-A

MacQuarrie

KTB

(2013) Potential surface temperature and shallow groundwater temperature response to climate change: An example from a small forested catchment in east-central New Brunswick (Canada). Hydrology and Earth System Sciences 17: 2701–2716.

53.

Kurylyk

MacQuarrie

KTB

Voss

(2014) Climate change impacts on the temperature and magnitude of groundwater discharge from shallow, unconfined aquifers. Water Resources Research 50: 3253–3274.

54.

Langan

Johnston

Donaghy

. (2001) Variation in river water temperatures in an upland stream over a 30-year period. Science of the Total Environment 265: 195–207.

55.

Leach

Moore

(2010) Above-stream microclimate and stream surface energy exchanges in a wildfire-disturbed riparian zone. Hydrological Processes 24: 2369–2381.

56.

Leach

Moore

(2011) Stream temperature dynamics in two hydrogeomorphically distinct reaches. Hydrological Processes 25: 679–690.

57.

Lamberti

Pearsons

. (1994) Cumulative effects of riparian disturbances along high desert trout streams of the John Day Basin, Oregon. Transactions of the American Fisheries Society 123: 627–640.

58.

MacDonald

Boon

Byrne

. (2014) Potential future climate effects on mountain hydrology, stream temperature, and native salmonid life history. Canadian Journal of Fisheries and Aquatic Sciences 71: 189–202.

59.

Mackey

Berrire

(1991) The prediction of water temperatures in chalk streams from air temperatures. Hydrobiologia 210: 183–189.

60.

Maderich

Helling

Bezhenar

. (2008) Development and application of 3D numerical model THREETOX to the prediction of cooling water transport and mixing in the inland and coastal waters. Hydrological Processes 22: 1000–1013.

61.

Malcolm

Soulsby

Hannah

. (2008) The influence of riparian woodland on stream temperatures: Implications for the performance of juvenile salmonids. Hydrological Processes 22: 968–979.

62.

Malcolm

Soulsby

Youngson

. (2004) Hydrological influences on hyporheic water quality: Implications for salmon survival. Hydrological Processes 18: 1543–1560.

63.

Malcolm

Soulsby

Youngson

. (2005) Catchment-scale controls on groundwater-surface water interactions in the hyporheic zone: Implications for salmon embryo survival. Hydrological Processes 21: 977–989.

64.

Mantua

Tohver

Hamlet

(2010) Climate change impacts on streamflow extremes and summertime stream temperature and their possible consequences for freshwater salmon habitat in Washington State. Climatic Change 102: 187–223.

65.

Meisner

(1990) Effect of climatic warming on the southern margins of the native range of brook trout, Salvelinus fontinalis. Canadian Journal of Fisheries and Aquatic Sciences 47: 1065–1070.

66.

Mohseni

Erickson

Stefan

(1999) Sensitivity of stream temperatures in the United States to air temperatures projected under a global warming scenario. Water Resources Research 35: 3723–3733.

67.

Mohseni

Stefan

Eaton

(2003) Global warming and potential changes in fish habitat in US streams. Climatic Change 59: 380–409.

68.

Mohseni

Stefan

Erickson

(1998) A nonlinear regression model for weekly stream temperatures. Water Resources Research 34: 2685–2692.

69.

Montgomery

(1999) Process domains and the river continuum. Journal of the American Water Resources Association 35: 397–410.

70.

Moore

Spittlehouse

Story

(2005) Riparian microclimate and stream temperature response to forest harvesting: A review. Journal of the American Water Resources Association 41: 813–834.

71.

Morrill

Bales

Conklin

(2005) Estimating stream temperature from air temperature: Implications for future water quality. Journal of Environmental Engineering 131: 139–146.

72.

Morrison

Quick

Foreman

MGG

(2002) Climate change in the Fraser River watershed: Flow and temperature projections. Journal of Hydrology 263: 230–244.

73.

O’Driscoll

DeWalle (2006) Stream-air temperature relations to classify stream-ground water interactions. Journal of Hydrology 329 (2): 140–153.

74.

Olden

Naiman

(2010) Incroporating thermal regimes into environmental flows assessments: Modifying dam operations to restore freshwater ecosystem integrity. Freshwater Biology 55: 86–107.

75.

Ormerod

(2009) Climate change, river conservation and the adaptation challenge. Aquatic Conservation: Marine and Freshwater Ecosystems 19: 609–613.

76.

Orr

des Clers

Simpson

. (2010) Changing water temperatures: A surface water archive for England and Wales. In: Kirby

(ed.) Role of Hydrology in Managing Consequences of a Changing Global Environment. Wallingford: British Hydrological Society, 8 pp.

77.

Orr

Simpson

des Clers

. (2014) Detecting changing river temperatures in England and Wales. Hydrological Processes. doi: 10.1002/hyp.10181.

78.

Petersen

Kitchell

(2001) Climate regimes and water temperature changes in the Columbia River: Bioenergetic implications for predators of juvenile salmon. Canadian Journal of Fisheries and Aquatic Sciences 58: 1831–1841.

79.

Pilgrim

Fang

Stefan

(1998) Stream temperature correlations with air temperatures in Minnesota: Implications for climate warming. Journal of the American Water Resources Association 34: 1109–1121.

80.

Poole

Berman

(2001) An ecological perspective on in-stream temperature: Natural heat dynamics and mechanisms human-caused thermal degradation. Environmental Management 27: 787–802.

81.

Ramaker

TAB

Meuleman

AFM

Bernhardi

. (2005) Climate change and drinking water production in the Netherlands: A flexible approach. Water Science and Technology 51 (5): 37–44.

82.

Ramanathan

(1981) The role of ocean-atmosphere interactions in the CO₂ climate problem. Journal of the Atmospheric Sciences 38: 918–930.

83.

Rutherford

Macaskill

Williams

(1993) Natural water temperature variations in the lower Waikato River, New Zealand. New Zealand Journal of Marine and Freshwater Research 27: 71–85.

84.

Sand-Jensen

Pedersen

(2005) Differences in temperature, organic carbon and oxygen consumption among lowland streams. Freshwater Biology 50: 1927–1937.

85.

Sinokrot

Stefan

McCormick

. (1995) Modelling of climate change effects on stream temperatures and fish habitats below dams and near groundwater inputs. Climatic Change 30: 181–200.

86.

Smith

(1995) Temperatures drop in New Zealand’s rivers. Water and Atmosphere 3: 27.

87.

Smith

(1972) River water temperatures – an environmental review. Scottish Geographical Magazine 88: 211–220.

88.

Stefan

Sinokrot

(1993) Projected global climate change on water temperatures in five north central US streams. Climatic Change 24: 353–381.

89.

Stott

Marks

(2000) Effects of plantation forest clearfelling on stream temperatures in the Plynlimon experimental catchments, mid-Wales. Hydrology and Earth System Sciences 4: 95–104.

90.

Tague

Farrell

Grant

. (2007) Hydrogeologic controls on summer stream temperatures in the McKenzie River basin, Oregon. Hydrological Processes 21: 3288–3300.

91.

Taylor

Stefan

(2009) Shallow groundwater temperature response to climate change and urbanization. Journal of Hydrology 375: 601–612.

92.

van Vliet

MTH

Franssen

WHP

Yearsley

. (2013a) Global river discharge and water temperature under climate change. Global Environmental Change 23: 450–464.

93.

van Vliet

MTH

Ludwig

Zwolsman

JJG

. (2011) Global river temperatures and sensitivity to atmospheric warming and changes in river flow. Water Resources Research 47: W02544.

94.

van Vliet

MTH

Vogele

Rubbelke

(2013b) Water constraints on European power supply under climate change: Impacts on electricity prices. Environmental Research Letters 8: 035010.

95.

Ward

(1985) Thermal characteristics of running waters. Hydrobiologia 125: 31–46.

96.

Weatherley

Ormerod

(1990) Forests and the temperature of upland streams in Wales: A modelling exploration of the biological effects. Freshwater Biology 24: 109–122.

97.

Webb

(1996) Trends in stream and river temperature. Hydrological Processes 10: 205–226.

98.

Webb

Nobilis

(1994) Water temperature behaviour in the River Danube during the twentieth century. Hydrobiologia 291: 105–113.

99.

Webb

Nobilis

(1997) A long-term perspective on the nature of the air–water relationship: A case study. Hydrological Processes 11: 137–147.

100.

Webb

Nobilis

(2007) Long-term changes in river temperature and the influence of climatic and hydrological factors. Hydrological Sciences Journal 52: 74–85.

101.

Webb

Walling

(1986) Spatial variation of water temperature characteristics and behaviour in a Devon river system. Freshwater Biology 16: 585–608.

102.

Webb

Walling

(1992) Long term water temperature behaviour and trends in a Devon, UK, river system. Hydrological Sciences Journal 37: 567–580.

103.

Webb

Walling

(1997) Complex summer water temperature below a UK regulating reservoir. Regulated Rivers: Research and Management 13: 463–477.

104.

Webb

Walsh

(2004) Changing UK river temperatures and their impact on fish populations. In: Webb

Acreman

Maksimovic

. (eds) Hydrology: Science and Practice for the 21st Century, Volume II. Wallingford: British Hydrological Society, 177–191.

105.

Webb

Zhang

(1997) Spatial and seasonal variability in the components of the river heat budget. Hydrological Processes 11: 79–101.

106.

Webb

Zhang

(1999) Water temperatures and heat budgets in Dorset chalk water courses. Hydrological Processes 13: 309–321.

107.

Webb

Zhang

(2004) Intra-annual variability in the non-advective heat energy budget of Devon streams and rivers. Hydrological Processes 18: 2117–2146.

108.

Webb

Clack

Walling

(2003) Water-air temperature relationships in a Devon river system and the role of flow. Hydrological Processes 17: 3069–3084.

109.

Webb

Hannah

Moore

. (2008) Recent advances in river and stream temperature. Hydrological Processes 22: 902–918.

110.

Wehrly

Wang

Mitro

(2007) Field-based estimates of thermal tolerance limits for trout: Incorporating exposure time and temperature fluctuation. Transactions of the American Fisheries Society 136: 365–374.

111.

Westhoff

Savenije

HHG

Luxemburg

WMJ

. (2007) A distributed stream temperature model using high resolution temperature observations. Hydrology and Earth System Sciences 11: 1469–1480.

112.

Wilby

Orr

Watts

. (2010) Evidence needed to manage freshwater ecosystems in a changing climate: Turning adaptation principles into practice. Science of the Total Environment 408: 4150–4164.

113.

Wild

Ohmura

Cubasch

(1997) GCM-simulated energy fluxes in climate change experiments. Journal of Climate 10: 3093–3110.

114.

Xin

Kinouchi

(2013) Analysis of stream temperature and heat budget in an urban river under strong anthropogenic influences. Journal of Hydrology 489: 16–25.

115.

Yang

Liu

(2005) Stream temperature changes over Lena River basin in Siberia. Geophysical Research Letters 32: L05401.

River water temperature in the United Kingdom

Abstract

Keywords

I Introduction

II Understanding processes, controls, dynamics and drivers of change

1 Processes

2 Controls and dynamics

3 Drivers of change

III What has happened?

1 International studies

2 UK studies

IV What may happen in the future?

1 International studies of future change

a Statistical models

b Deterministic models

2 UK studies of future change

V Conclusions and knowledge gaps

Footnotes

Acknowledgements

Funding

References