Climate and basin drivers of seasonal river water temperature dynamics

• Published in: Hydrology and earth system sciences 2017-06, Vol.21 (6), p.3231-3247
• Main Authors: Laizé, Cédric L. R, Bruna Meredith, Cristian, Dunbar, Michael J, Hannah, David M
• Format: Article
• Language: English
• Subjects: Air temperature; Analysis; Atmospheric models; Autumn; Basins; Biogeochemistry; Climate; Climate change; Climate effects; Climate models; Climate prediction; Climate variability; Coefficients; Computational fluid dynamics; Cooling effects; Creeks & streams; Dynamics; Ecology; Electric power generation; Electric power plants; Elevation; Energy; Environmental aspects; Fluid flow; Groundwater; Heat; Heat exchange; Humidity; Hydraulics; Hydrology; Long wave radiation; Modelling; Permeability; Power plants; Precipitation; Properties; Properties (attributes); Radiation; Recreational use; Regression analysis; River water; Rivers; Robustness (mathematics); Seasons; Short wave radiation; Specific humidity; Spring (season); Statistical analysis; Studies; Summer; Temperature effects; Temperature range; Temporal variability; Temporal variations; Time series; Variability; Variables; Variance analysis; Water temperature; Wind speed; Winter
• ISSN: 1607-7938; 1027-5606; 1607-7938
• DOI: 10.5194/hess-21-3231-2017

Stream water temperature is a key control of many river processes (e.g. ecology, biogeochemistry, hydraulics) and services (e.g. power plant cooling, recreational use). Consequently, the effect of climate change and variability on stream temperature is a major scientific and practical concern. This paper aims (1) to improve the understanding of large-scale spatial and temporal variability in climate–water temperature associations, and (2) to assess explicitly the influence of basin properties as modifiers of these relationships. A dataset was assembled including six distinct modelled climatic variables (air temperature, downward short-wave and long-wave radiation, wind speed, specific humidity, and precipitation) and observed stream temperatures for the period 1984–2007 at 35 sites located on 21 rivers within 16 basins (Great Britain geographical extent); the study focuses on broad spatio-temporal patterns, and hence was based on 3-month-averaged data (i.e. seasonal). A wide range of basin properties was derived. Five models were fitted (all seasons, winter, spring, summer, and autumn). Both site and national spatial scales were investigated at once by using multi-level modelling with linear multiple regressions. Model selection used multi-model inference, which provides more robust models, based on sets of good models, rather than a single best model. Broad climate–water temperature associations common to all sites were obtained from the analysis of the fixed coefficients, while site-specific responses, i.e. random coefficients, were assessed against basin properties with analysis of variance (ANOVA). All six climate predictors investigated play a role as a control of water temperature. Air temperature and short-wave radiation are important for all models/seasons, while the other predictors are important for some models/seasons only. The form and strength of the climate–stream temperature association vary depending on season and on water temperature. The dominating climate drivers and physical processes may change across seasons and across the stream temperature range. The role of basin permeability, size, and elevation as modifiers of the climate–water temperature associations was confirmed; permeability has the primary influence, followed by size and elevation. Smaller, upland, and/or impermeable basins are the most influenced by atmospheric heat exchanges, while larger, lowland and permeable basins are the least influenced. The study showed the importa