Simulated Water Temperatures Research Articles

Hazard assessment of rice cold damage based on energy balance in paddy field

Cold stress seriously affects rice yield in Northeast China and, as a result of climate change, there are new trends in the characterization of cold damage. Accurate simulation of water temperature in paddy fields and assessment of cold damage hazard can contribute to improving the accuracy of agrometeorological disaster risk assessment in the context of climate change. Hence, in this study, first, we simulated the water temperature of paddy fields based on the paddy field energy-balance model and evaluated the simulation results. Subsequently, a composite cooling-degree-day (CCDD) indicator for rice fields was constructed based on water and air temperatures to characterize cold stress and identify cold damage events. Finally, a cold damage hazard-assessment model was constructed for rice based on the frequency and intensity of cold damage to assess cold damage hazard in Northeast China from 1979 to 2018. The results showed that the root-mean-square error between simulated and observed water temperatures of the rice field energy-balance model was 0.45, which means that the model accurately simulated the water temperature of the rice field on a large spatial scale. In 2004, cold damage to late-maturity rice accounted for 30% of the total study area, which was one of the years with most severe cold damage in late-maturity rice-growing areas thus far in the 21st century. Water temperature affected 76.4% of the area in 2004, and was related to radiation and saturated vapor pressure. The low-hazard areas accounted for more than 20% in 2009–2018, and the high-hazard areas were still higher than 10%, indicating that the threat of cold damage in the northeastern region of China remained severe as the climate has become increasingly warm.

Priority sites for coral aquaculture in Kume Island based on numerical simulation

Global and local stresses such as increased water temperature and terrestrial sediment runoff significantly affect coral reef ecosystems. Over the past three decades, the increase in the intensity and frequency of coral bleaching has been a significant problem in tropical and subtropical coastal seas, and maintaining healthy coral reef ecosystems has long been a challenge. Coral cover in Shimajiri Bay on Kume Island, located in the Nansei Islands (Okinawa Prefecture, southern Japan), has declined due to exposure to high water temperatures, an outbreak of coral predators, and sediment (red soil) inflow from the land. From such a background, coral aquaculture has been implemented at a few stations in the bay. Therefore, it is essential to conduct aquaculture in suitable locations for coral growth and survival. This study aimed to select suitable locations for coral culture with high spatial resolution in an enclosed bay that was greatly affected by red soil runoff, mainly from a sugarcane field. A freshwater and red soil runoff model for the watershed was merged with a coastal hydrodynamic, red soil dispersal, and accumulation model at a spatial resolution of 75 m for Shimajiri Bay. The simulation results were validated using in situ continuous and routine observations. Simulations were conducted over ten years. Locations not subjected to high water temperature in the bay or river loading were selected based on topography, simulated water temperature, salinity, and red soil content of sea sediment. This coupled model enables us to evaluate the growth and survival conditions of cultured corals with high spatiotemporal resolution and quantitatively consider the reduction target of red soil runoff to realize robust coral growth and survival.

Shedding Light on the Failure Factors of Subsea Critical Fastener Bolts

Offshore facilities, such as oil and gas rigs, wind turbines, and related subsea equipment, typically use flanges fastened using bolts and nuts as the main connectors. In this study, multidisciplinary parameters, namely the preload torque used to tighten bolts, simulated subsea water currents, water temperature, and impressed current cathodic protection, were applied to ASTM A193 B7 bolts. An experimental supervisory control and data acquisition system was designed to obtain measurements every 5 min throughout a 21-day experiment. Finite element analysis was performed to predict the structurally vulnerable areas of the bolts. A strong correlation was found between the reference electrode readings and the measured electrical current, tightening torque, and water temperature. As the water temperature rises during the day, the reference electrode reading becomes less negative and the electrical current decreases. Subsea water currents cause about a four-time increase in the bolt corrosion rate, with unprotected bolts suffering a nine-time-higher corrosion rate than protected bolts. A unique supply–demand interaction is observed; less protection is supplied to areas with lower corrosion rates (lower demand for protection). Finally, scanning electron microscopy examination reveals new insights into the failure mechanisms of subsea bolts.

Spatial-temporal change of river thermal environment and anthropogenic impact in China

River water temperature is important and closely related to river ecosystem, concerning fishery industry, human health, and the land-sea exchange of nutrients, especially for great powers with a good deal of heat emission from once-through cooling systems of thermal power plants. However, the changes in river water temperature under the joint action of climate change and human activity such as the heat emission have not been well investigated for rising powers, hampering environmental policy making for sustainable development. Therefore, we have taken advantage of a recently-developed land surface model including river water temperature calculation with anthropogenic thermal discharge and zonal statistics to quantitatively make out the river water temperature variation and the man-made influence over the past thirty years (1981–2010) in China for the first time. Results show that the estimated water temperature in major rivers is generally close to the observation with the r2 of 0.83, though the underestimation exists in some rivers. The river water in the Pearl River Basin was the warmest with the mean temperature of 19.2 °C and the others in order were in the Southeast Basin, the Huaihe River Basin, the Yangtze River Basin, the Haihe River Basin, the Yellow River Basin, the Southwest Basin, the Song-Liao River Basin, and the Continental Basin, ranging from 16.7 °C to 6.3 °C. The Huaihe River Basin had the fastest mean increase rate of ca. 0.27 °C decade−1, while the slowest mean increase rate of ca. 0.13 °C decade−1 existed in the Pearl River Basin. At the subbasin scale, a meridionally-distributed hot spot zone (along the 115°E) of the increasing water temperature has been identified, where the trends ranged from 0.2 °C decade−1 to 0.5 °C decade−1. Air temperature exerted a major control on the spatial pattern of climatological water temperature, while both air temperature and downwelling solar flux played a leading role in the distribution of water temperature change trends. Although anthropogenic thermal emission heated the rivers locally, the impacts in the Song-Liao River, the Haihe River, the Huaihe River, and the middle and lower reaches of Yellow River and Yangtze River had been raised up to ca. 4.5 °C, when comparing with those controlled by climate change only. In general, these results show an acceptable level of river water temperature simulation in the land surface model, and could provide a scientific reference for the assessment on riverine thermal environment under the climate change and social impact in China.

Simulation and prediction of water temperature in a water transfer channel during winter periods using a new approach based on the wavelet noise reduction-deep learning method

Abstract In winter, the water transfer channel of the Middle Route of South-to-North Water Transfer Project (MR-StNWTP) in China always encounters ice problems. The preciously simulation and prediction of water temperature is essential for analyzing the ice condition, which is important for the safety control of the water transfer channel in winter. Due to the difference of specific heat between water and air, when the air temperature rises and falls dramatically, the range of change of water temperature is relatively small and has a lag, which often affects the accuracy of simulation and prediction of water temperature based on air temperature. In the present study, a new approach for simulating and predicting water temperature in water transfer channels in winter has been proposed. By coupling the neural network theory to equations describing water temperature, a model has been developed for predicting water temperature. The temperature data of prototype observations in winter are preprocessed through the wavelet decomposition and noise reduction. Then, the wavelet soft threshold denoising method is used to eliminate the fluctuation of certain temperature data of prototype observations, and the corresponding water temperature is calculated afterward. Compared to calculation results using both general neural network and multiple regression approaches, the calculation results using the proposed model agree well with those of prototype measurements and can effectively improve the accuracy of prediction of water temperature.

One- and Three-Dimensional Hydrodynamic, Water Temperature, and Dissolved Oxygen Modeling Comparison

Understanding and modeling water quality in a lake/reservoir is important to the effective management of aquatic ecosystems. The advantages and disadvantages of different water quality models make it challenging to choose the most suitable model; however, direct comparison of 1-D and 3-D models for lake water quality modeling can reveal their relative performance and enable modelers and lake managers to make informed decisions. In this study, we compared the 1-D model MINLAKE and the 3-D model EFDC+ for water temperature, ice cover, and dissolved oxygen (DO) simulation in three Minnesota lakes (50-m Carlos Lake, 23.5-m Trout Lake, and 5.6-m Pearl Lake). EFDC+ performed well for water temperature and DO simulation in the open water seasons with an average root mean square error (RMSE) of 1.32 °C and 1.48 mg/L, respectively. After analyzing the ice thickness with relevant data, it was found that EFDC+ calculates a shorter ice cover period and smaller ice thickness. EFDC+ does not consider snowfall for ice thickness simulation. The results also revealed that EFDC+ considers spatial variance and allows the user to select inflow/outflow locations precisely. This is important for large lakes with complex bathymetry or lakes having multiple inlets and outlets. MINLAKE is computationally less intensive than EFDC+, allowing rapid simulation of water quality parameters over many years under a variety of climate scenarios.

Exploring the Sensitivity of Water Temperature Models for High Dams and Large Reservoirs: A Case Study of the Nuozhadu Reservoir in China

The parameters governing a water temperature model play a pivotal role in determining the uncertainties associated with the model’s outcome. In this study, a two-dimensional (2D) hydrodynamic and water temperature coupling model is constructed, focusing on the Nuozhadu Reservoir situated along the Lancang River. Employing a single-factor analysis approach, the sensitivity of the thermal balance parameters and hydrodynamic parameters in the model is assessed. This study overcomes the shortcomings of previous sensitivity analyses of hydrodynamic parameters in reservoir water temperature models. The findings reveal that the attenuation parameters of light and Beer’s law parameter exhibit minimal sensitivity to the vertical temperature structure. Conversely, radiation parameter A and radiation parameter B exert tenfold disparate influences on the surface and bottom temperatures of the reservoir. Among the hydrodynamic parameters considered, the horizontal viscosity factor shows no sensitivity to the vertical temperature structure, whereas the vertical viscosity factor serves as a crucial determinant, directly influencing the intensity of vertical temperature stratification. An increased vertical viscosity factor promotes heat exchange between the upper and lower water layers, thereby reducing the vertical temperature gradient and weakening stratification. Conversely, diminishing this factor intensifies stratification. Thus, when conducting water temperature simulations in high dams and large reservoirs, careful attention should be given to calibrating vertical viscosity factor.

Modular Compositional Learning Improves 1D Hydrodynamic Lake Model Performance by Merging Process‐Based Modeling With Deep Learning

AbstractHybrid Knowledge‐Guided Machine Learning (KGML) models, which are deep learning models that utilize scientific theory and process‐based model simulations, have shown improved performance over their process‐based counterparts for the simulation of water temperature and hydrodynamics. We highlight the modular compositional learning (MCL) methodology as a novel design choice for the development of hybrid KGML models in which the model is decomposed into modular sub‐components that can be process‐based models and/or deep learning models. We develop a hybrid MCL model that integrates a deep learning model into a modularized, process‐based model. To achieve this, we first train individual deep learning models with the output of the process‐based models. In a second step, we fine‐tune one deep learning model with observed field data. In this study, we replaced process‐based calculations of vertical diffusive transport with deep learning. Finally, this fine‐tuned deep learning model is integrated into the process‐based model, creating the hybrid MCL model with improved overall projections for water temperature dynamics compared to the original process‐based model. We further compare the performance of the hybrid MCL model with the process‐based model and two alternative deep learning models and highlight how the hybrid MCL model has the best performance for projecting water temperature, Schmidt stability, buoyancy frequency, and depths of different isotherms. Modular compositional learning can be applied to existing modularized, process‐based model structures to make the projections more robust and improve model performance by letting deep learning estimate uncertain process calculations.

Incorporating physically-based water temperature predictions into the National water model framework

While river water temperatures are a strong control on instream processes and aquatic ecosystems, monitoring networks for river water temperatures are often sparse. Despite recent advancements in water temperature modeling strategies, current models struggle to provide real-time and reach-specific predictions across broad spatial domains. We developed a physically-based water temperature model coupled to the National Water Model (NWM) to assess the potential for water temperature prediction to be incorporated into the NWM at the continental scale. Using model forcings and outputs from the NWM v2.1 retrospective, we evaluated the ability of four model configurations of increasing complexity to simulate hourly water temperatures in the forested headwaters of H.J. Andrews Experimental Forest, Oregon, USA during a six-week summer period. Our modeling framework, representing a first effort at pairing water temperature simulation with predictions from the NWM, suggests that the NWM can be leveraged to give insight into other water quality variables.

Combining Landsat TIR‐imagery data and ERA5 reanalysis information with different calibration strategies to improve simulations of streamflow and river temperature in the Canadian Subarctic

AbstractArctic and Subarctic environments are among the most vulnerable regions to climate change. Increases in liquid precipitation and changes in snowmelt onset are cited as the main drivers of change in streamflow and water temperature patterns in some of the largest rivers of the Canadian Arctic. However, in spite of this evidence, there is still a lack of research on water temperature, particularly in the eastern Canadian Arctic. In this paper, we use the CEQUEAU hydrological‐water temperature model to derive consistent long‐term daily flow and stream temperature time series in Aux Mélèzes River, a non‐regulated basin (41 297 km2) in the eastern Canadian subarctic. The model was forced using reanalysis data from the fifth‐generation ECMWF atmospheric reanalyses (ERA5) from 1979 to 2020. We used water temperature derived from thermal infrared (TIR) images as reference data to calibrate CEQUEAU's water temperature model, with calibration performed using single‐site, multi‐site, and upscaling factors approaches. Our results indicate that the CEQUEAU model can simulate streamflow patterns in the river and shows excellent spatiotemporal performance with Kling‐Gupta Efficiency (KGE) metric >0.8. Using the best‐performing flow simulation as one of the inputs allowed us to produce synthetic daily water temperature time series throughout the basin, with the multi‐site calibration approach being the most accurate with root mean square errors (RMSE) <2.0°C. The validation of the water temperature simulations with a three‐year in situ data logger dataset yielded an RMSE = 1.38°C for the summer temperatures, highlighting the robustness of the calibrated parameters and the chosen calibration strategy. This research demonstrates the reliability of TIR imagery and ERA5 as sources of model calibration data in data‐sparse environments and underlines the CEQUEAU model as an assessment tool, opening the door to its use to assess climate change impact on the arctic regions of Canada.

Improvements and Evaluation of the FLake Model in Dagze Co, Central Tibetan Plateau

FLake has been one of the most extensively used lake models in many studies for lake thermal structure simulations. However, due to the scarcity of lake temperature observations, its applicability and performance on lakes over the Tibetan Plateau are still poorly investigated, especially in small- to medium-sized lakes. In this study, based on water profile observations in Dagze Co, a medium-sized lake on the central Tibetan Plateau, the sensitivity of lake thermal features to three key parameters in FLake was investigated. The performance of FLake in reproducing the lake thermal features was evaluated and improved by optimizing these key parameters. The results showed that the FLake model with default parameter settings can generally reproduce the thermal features of Dagze Co, but there are still significant deviations compared to observation. The sensitive experiments demonstrated that the thermal structure of the lake obviously responds to the change in the water extinction coefficient (Kd), friction velocity (u*), and ice albedo (αice). Based on previous studies and sensitive experiments, the three key parameters were set to the optimized value, which substantially improved the performance of FLake. The values of bias and RMSE of simulated lake surface water temperature decreased from 3.08 °C and 3.62 °C to 2.0 °C and 2.48 °C after parameter optimization. The integration of a simple salinity scheme further improved the ability of FLake to reproduce the observed thermal features of Dagze Co. These results will improve our understanding of thermal processes in lakes on the Tibetan Plateau, as well as the applicability of lake models.

Lake Ice Simulation and Evaluation for a Typical Lake on the Tibetan Plateau

This study aims to simulate the lake ice conditions in the Nam Co lake using a lake ice model, which is a one-dimensional physics-based model that utilizes enthalpy as the predictor variable. We modified the air density schemes within the model to improve the accuracy of the lake ice simulation. Additionally, the process of lake ice sublimation was included, and the effect of lake water salinity on the freezing point was considered. Using the improved lake ice model, we simulated lake surface water temperature, lake ice thickness, and interannual variations in lake ice phenology, and we compared these results with observations at Nam Co. The results demonstrate that the improved model better reproduces the lake surface water temperature, lake ice thickness, and lake ice phenology at Nam Co. Additionally, the thin air density affects lake processes by weakening sensible heat and latent heat, which ultimately leads to a delayed ice-on date and a slightly earlier ice-free date in Nam Co. This study contributes to an enhanced understanding of the freeze–thaw processes in Nam Co and reduces the biases in lake ice simulation on the Tibetan Plateau through the lake model improvement.

Simulated Clear-Sky Water Vapor and Temperature Retrievals from PREFIRE Measurements

Abstract The Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) mission will measure Earth’s emission at wavelengths ranging from 3 to 54 μm. The prelaunch clear-sky retrieval algorithm, evaluated with simulated test data, indicates that PREFIRE measurements will be valuable for retrieving atmospheric water vapor and temperature profiles. Far infrared measurements provide unique retrieval information, indicated by the high ranking of select FIR channels as primary contributors to the total degrees of freedom for signal (DFS). In utilizing all the PREFIRE channels, the average total DFS of 4 test regions ranges from 1.90 to 4.71. The information content increases with higher column water vapor and in the presence of near-surface temperature inversions. Using the DFS profiles for guidance, the retrieval concentrates information into 7 distinct layers to reduce the retrieval uncertainty per layer. Sensitivity tests indicate forward model error due to surface emissivity uncertainty results in about a 9% increase in column water vapor uncertainty. The clear-sky retrieval is sensitive to the presence of undetected ice clouds, especially those with optical depths larger than 0.2. Hence, in addition to a separate PREFIRE cloud mask, optimal estimation retrieval metrics are explored as possible indicators of cloudy scenes.

The Respondence of Wave on Sea Surface Temperature in the Context of Global Change

Several aspects of global climate change, e.g., the rise of sea level and water temperature anomalies, suggest the advantages of studying wave distributions. In this study, WAVEWATCH-III (WW3) (version 6.07), which is a well-known numerical wave model, was employed for simulating waves over global seas from 1993–2020. The European Centre for Medium-Range Weather Forecasts (ECMWF), Copernicus Marine Environment Monitoring Service (CMEMS), current and sea level were used as the forcing fields in the WW3 model. The validation of modelling simulations against the measurements from the National Data Buoy Center (NDBC) buoys and Haiyang-2B (HY-2B) altimeter yielded a root mean square error (RMSE) of 0.49 m and 0.63 m, with a correlation (COR) of 0.89 and 0.90, respectively. The terms calculated by WW3-simulated waves, i.e., breaking waves, nonbreaking waves, radiation stress, and Stokes drift, were included in the water temperature simulation by a numerical circulation model named the Stony Brook Parallel Ocean Model (sbPOM). The water temperature was simulated in 2005–2015 using the high-quality Simple Ocean Data Assimilation (SODA) data. The validation of sbPOM-simulated results against the measurements obtained from the Array for Real-time Geostrophic Oceanography (Argo) buoys yielded a RMSE of 1.12 °C and a COR of 0.99. By the seasonal variation, the interrelation of the currents, sea level anomaly, and significant wave heights (SWHs) were strong in the Indian Ocean. In the strong current areas, the distribution of the sea level was consistent with the SWHs. The monthly variation of SWHs, currents, sea surface elevation, and sea level anomalies revealed that the upward trends of SWHs and sea level anomalies were consistent from 1993–2015 over the global ocean. In the Indian Ocean, the SWHs were obviously influenced by the SST and sea surface wind stress. The rise of wind stress intensity and sea level enlarges the growth of waves, and the wave-induced terms strengthen the heat exchange at the air–sea layer. It was assumed that the SST oscillation had a negative response to the SWHs in the global ocean from 2005–2015. This feedback indicates that the growth of waves could slow down the amplitude of water warming.

Modeling River Plume Dynamics in a Large Wind-Forced Embayment

Water quality degradation, in the form of undesirable algae, occurs near the Nottawasaga River mouth and along Wasaga Beach, southeastern Georgian Bay. The ability to manage this water resource is compromised due to lack of monitoring and science. In the present study, a 3D hydrodynamic model was applied to gain an understanding of how advection and dilution of the river plume can trap nutrients along Wasaga Beach. Simulated water temperatures and currents had a root-mean-square error between 1.5°C and 2.5°C and ∼0.06 m s−1, respectively. Maximum nearshore river concentrations occurred near the beach during high river discharge >20 m3 s−1 and westerly winds. Generally, the river plume was advected by winds >∼4 m s−1, traveling along the northern shoreline during southerly winds and along the southwest shoreline during northeasterly winds. This process was analytically modeled when the wind strength indicator or Froude number (ratio of the characteristic wind-velocity scale to the buoyancy–velocity scale) was greater than one.

Observation and Simulation of Mosquito Breeding Site Water Temperature for Malaria Transmission at Kaita Local Governmet Township of Katsina State, Nigeria

Studies showed that transmission of malaria is influenced by environmental factors such as temperature. This work is aimed at finding the impact of mosquito's breeding site water temperature on mosquito's larva development time. An artificial mosquito's breeding habitat was created. The water temperature of the habitat was measured at an hourly interval, then it is averaged into daily time scale. Weather variables of the experimental site were inpu into the the energy balance model to simulate the breeding habit water temperature. The mosquito's larva development time was then predicted by inputting both the observed water and simulated water temperature into the vector borne disease community model (VECTRI) .The daily maximum, and minimum observed water temperatures were 27.9°C, 32.6°C and 21.7°C, respectively. The daily mean, maximum, and minimum simulated water temperatures were 29.8°C, 35.6°C, and 23.5°C respectively. These temperatures are within the temperature range that supports mosquito’s larva development. Mosquito's larva development was predicted using the VECTRI model. According to this study larva development reached completion in 7.1 days using the observed water temperature, 6.03 days using the simulated water temperature and 8.01days using the observed air temperature. This energy balance model is an improved water temperature scheme over the assumption that air temperature is equal to air temperature. This work shows the importance of water temperature and the value of degree day required for emergence of an adult mosquito in the simulation of aquatic stage development. Both the observed water and simulated water temperatures are higher than the on observed air temperature, thus air temperature cannot be used as the water temperature in the simulation of the mosquito’s larva development time. The finding of the work can be used as source toward mosquito's larval control through water temperature. It is however clear from the finding that could be as result of temperature due to shorter time predicted for mosquito's larval development.

Assessing the uncertainties in modeling water temperatures during river cooling and freeze-up periods

Accurate simulation of water temperatures is crucial to the simulation of ice processes in river ice models. This study presents a quantitative assessment of uncertainties in simulating the water temperatures during river cooling and freeze-up periods. In particular, the selections of the heat transfer model and the proximity of the weather station to the study reach are investigated using the University of Alberta's River1D Ice Process model. Results show that using local versus remote weather data only had a significant impact on the performance of the full energy budget model during the river cooling period. This was mainly attributed to the large difference in the locally and remotely measured wind speeds. The full energy budget model during the freeze-up period and the linear heat transfer model were relatively insensitive to the source of weather data. The full energy budget model using local weather data was the most accurate in simulating water temperatures. When using remote weather data, the full energy budget model was generally more accurate than the linear heat transfer model in the calibration year and the two validation years, except during the cooling period of the calibration year. Neither model was able to consistently and accurately simulate the timing and magnitude of observed supercooling events. It is recommended that the full energy budget model be used whenever comprehensive weather data are available due to its overall better accuracy and less calibrated parameters than the linear heat transfer model.

Water quality modeling of a eutrophic drinking water source: Impact of future climate on Cyanobacterial blooms

Cyanobacterial blooms are becoming more frequent in freshwater sources, causing concern throughout the world. Cyanobacterial blooms affect human health and the entire environment. Numerical modeling is an effective tool for investigating aquatic systems. In this study, a 3D hydrodynamic and water quality (ecological) model was used to simulate eutrophication of a drinking water source, Lake Vomb, in Sweden under present and future scenarios. The hydrodynamic model was set-up in MIKE 3 FM software based on meteorological, hydrological, and water quality data. The hydrodynamic model performance was satisfactory in terms of the water temperature simulation, with root-mean-square-error (RMSE) ranging from 0.38 to 1.2 °C. In the ecological model, Chlorophyll-a (Chl-a) was used as a proxy for Cyanobacteria, and the model proved acceptable in simulating the Chl-a concentrations, with a Nash-Sutcliffe efficiency (NSE) of 0.93 and 0.87 for calibration and validation respectively. The findings revealed that external nitrogen loading and internal phosphorus loading had significant impact on the nutrient concentrations in Lake Vomb. The findings also showed a correlation between Chl-a levels and total phosphorus levels in the lake. To simulate future water quality in the lake, two Representative Concentration Pathways (RCP) for the year 2050 were used to make projections for changes in air temperature and precipitation. Under the projected future climate, the simulations showed a considerable rise in Cyanobacteria biomass independent of the changes in external nutrient loading. The model findings can assist water managers in planning mitigation strategies by identifying major nutrient sources.

Recursive training based physics-inspired neural network for electric water heater modeling

Aggregating flexibility from residential electric water heaters (EWHs) is fast gaining commercial interest. Flexibility modeling of an EWH involves highly precise and quick simulation of EWH water temperature using the EWH thermal dynamics model for various flexibility control actions. Since EWH tank water temperature data is usually unavailable or costly to obtain, developing an accurate and computationally inexpensive EWH thermal dynamics model with limited sensor data is essential for devising advanced control strategies for EWH flexibility aggregation. In this paper, we present a novel recursive training-based unsupervised physics-informed neural network (PINN) model for predicting tank water temperature which requires only historical EWH power consumption data to train the model. PINN models enable the integration of domain knowledge from traditional physical processes and methods into neural network (NN) models. Single-zone thermal grey-box differential equation model (DEM) is used as the basis to develop and demonstrate proof-of-concept of the proposed approach. Physics from the single-zone model is encoded into the PINN loss function to incorporate domain knowledge and the PINN architecture is structured to mimic the single-zone DEM. The recursive training approach enables the use of previous-step water temperature as an input to the simulation model. Two separate models for EWH ON- and OFF-states are developed and trained with real-world EWH power consumption data. Water temperature prediction results indicate that the proposed approach has similar performance as the traditional single-zone DEM model, thereby demonstrating the ability of the proposed model to learn the underlying physics behind the single-zone model without water temperature data. The proposed model has high accuracy and performs well outside the control set point temperatures indicating its suitability for simulating load shifting and other DR events. Additionally, EWH simulation results for two different scenarios with different water demand compositions are presented to study the effects of propagation errors on temperature prediction. The proposed approach paves the way for developing advanced EWH flexibility modeling tools for the aggregator to precisely control a large portfolio of EWHs considering user comfort and rebound effects.

A simulation–optimization framework for reducing thermal pollution downstream of reservoirs

Abstract Thermal pollution is an environmental impact of large dams altering the natural temperature regime of downstream river ecosystems. The present study proposes a simulation–optimization framework to reduce thermal pollution downstream from reservoirs and tests it on a real-world case study. This framework attempts to simultaneously minimize the environmental impacts as well as losses to reservoir objectives for water supply. A hybrid machine-learning model is applied to simulate water temperature downstream of the reservoir under various operation scenarios. This model is shown to be robust and achieves acceptable predictive accuracy. The results of simulation–optimization indicate that the reservoir could be operated in such a way that the natural temperature regime is reasonably preserved to protect downstream habitats. Doing so, however, would result in significant trade-offs for reservoir storage and water supply objectives. Such trade-offs can undermine the benefits of reservoirs and need to be carefully considered in reservoir design and operation.