High‐Resolution Global Water Temperature Modeling

Abstract

The temperature of river water plays a crucial role in many physical, chemical, and aquatic ecological processes. Despite the importance of having detailed information on this environmental variable at locally relevant scales (≤50 km), high‐resolution simulations of water temperature on a large scale are currently lacking. We have developed the dynamical 1‐D water energy routing model (DynWat), that solves both the energy and water balance, to simulate river temperatures for the period 1960–2014 at a nominal 10‐km and 50‐km resolution. The DynWat model accounts for surface water abstraction, reservoirs, riverine flooding, and formation of ice, enabling a realistic representation of the water temperature. We present a novel 10‐km water temperature data set at the global scale for all major rivers, lakes, and reservoirs. Validated results against 358 stations worldwide indicate a decrease in the simulated root‐mean‐square error (0.2 °C) and bias (0.7 °C), going from 50‐ to 10‐km simulations. We find an average global increase in water temperature of 0.16 °C per decade between 1960 and 2014, with more rapid warming toward 2014. Results show increasing trends for the annual daily maxima in the Northern Hemisphere (0.62 °C per decade) and the annual daily minima in the Southern Hemisphere (0.45 °C per decade) for 1960–2014. The high‐resolution modeling framework not only improves the model performance, it also positively impacts the relevance of the simulations for regional‐scale studies and impact assessments in a region without observations. The resulting global water temperature data set could help to improve the accuracy of decision‐support systems that depend on water temperature estimates.

Threats of global warming to the world’s freshwater fishes

“…This is because water temperatures vary less within species ranges and are projected to rise almost everywhere, while flow conditions are more spatially variable hence projected future flow is less likely to exceed present-day extremes within the species’ ranges. In line with previous studies, we found that climate change will result in reduced flows mainly in drought-prone regions 21 , 24 . In addition, depletion of low flows might be most important at low stream orders, which are not well captured by the 5 arcminutes resolution of the hydrological model PCR-GLOBWB employed in this study 20 .…”

“…The 3.2 °C warming scenario represents the maximum warming predicted to occur by the end of the century (with 66% probability) if all current greenhouse gas emissions reductions targets (unconditional Nationally Determined Contributions) for 2030 are met and no further cuts are performed. We calculate the present and future weekly flow and water temperature values corresponding with each warming level at a spatial resolution of 5 arcminutes (~10 km) using a global hydrological model coupled to a dynamic water temperature model 20 , 21 . We force the hydrological model with meteorological input from five Global Climate Models (GCMs) combined with four Representative Concentration Pathway (RCP) representing future greenhouse gas emissions.…”

Vulnerability of existing and planned coal-fired power plants in Developing Asia to changes in climate and water resources

“…The water temperature module is an improvement of a previous energy balance model. 56,57 We conducted the simulations consecutively for the historical period and the future period under RCP6.0 scenario (2006-2100). The PCR-GLOBWB 2 model considers the impacts of water demands from the irrigation, livestock, industry, and household sectors, and the impact of reservoir management on streamflow under the ''Middle of the Road'' Shared Socioeconomic Pathway (SSP2).…”

A Multilayer Reservoir Thermal Stratification Module for Earth System Models

“…That is, with stratification, downstream temperature is warmer during winter and spring, and colder during summer and fall compared to simulation without stratification. Performance comparison between our validation result and the study of Wanders et al (2019) indicated that in general, our model has a bias indicating overestimation ( Figure S4) contrary to a negative bias values in Wanders et al (2019). However, this comparison is based on only the 28 reservoirs Figure 11.…”

“…For lake models coupled to earth system models, a multilayer stratification parametrization is used to simulate the vertical energy balance of the reservoir (Bonan et al, ; Subin et al, ). Wanders et al () studied lake and reservoir effect by identifying the epilimnion from thermocline depth and solving water and energy balance in the well‐mixed layer. With advancement in computing, we propose a representation of reservoir stratification using multilayer parametrization and realistic reservoir geometry (storage‐area‐depth relationship).…”