Articles | Volume 25, issue 5
Research article 31 May 2021
Research article | 31 May 2021
Machine-learning methods for stream water temperature prediction
Moritz Feigl et al.
No articles found.
Michael Weber, Franziska Koch, Matthias Bernhardt, and Karsten Schulz
Hydrol. Earth Syst. Sci., 25, 2869–2894,Short summary
We compared a suite of globally available meteorological and DEM data with in situ data for physically based snow hydrological modelling in a small high-alpine catchment. Although global meteorological data were less suited to describe the snowpack properly, transferred station data from a similar location in the vicinity and substituting single variables with global products performed well. In addition, using 30 m global DEM products as model input was useful in such complex terrain.
Christoph Klingler, Karsten Schulz, and Mathew Herrnegger
Earth Syst. Sci. Data Discuss.,
Preprint under review for ESSDShort summary
LamaH is a large-sample catchment hydrology dataset for Central Europe. The dataset contains hydrometeorological time series (daily and hourly resolution) and various attributes for 859 observed basins. Sticking closely to the CAMELS datasets, LamaH also includes a basin delineation and attributes for describing a large interconnected river network. Given the scope, LamaH might reveal deeper insights into water transfer and storage with appropriate methods of modelling or regression.
Christian Voigt, Karsten Schulz, Franziska Koch, Karl-Friedrich Wetzel, Ludger Timmen, Till Rehm, Hartmut Pflug, Nico Stolarczuk, Christoph Förste, and Frank Flechtner
Hydrol. Earth Syst. Sci. Discuss.,
Revised manuscript under review for HESSShort summary
A continuously operating gravimeter at the summit of Mt. Zugspitze is introduced to support hydrological studies of the Research Catchment Zugspitze. The observed gravity residuals reflect total water storage variations at the observation site. Hydro-gravimetric sensitivity analysis reveal a gravimetric footprint of up to 4 km enabling important and representative integral insights into this high-alpine as a basis for upcoming long-term hydro-gravimetric research projects.
Josef Fürst, Hans Peter Nachtnebel, Josef Gasch, Reinhard Nolz, Michael Paul Stockinger, Christine Stumpp, and Karsten Schulz
Earth Syst. Sci. Data Discuss.,
Revised manuscript accepted for ESSDShort summary
Rosalia is a 950 ha forested research watershed in eastern Austria to study water, energy and solute transport processes. The paper describes site, monitoring network and instrumentation, as well as the datasets: High resolution (10 min interval) time series, starting 2015, of 4 discharge gauging stations, 7 rain-gauges, together with observation of air and water temperature, relative humidity and conductivity, as well as soil water content and temperature in different depths at 4 profiles.
Christoph Schürz, Bano Mehdi, Jens Kiesel, Karsten Schulz, and Mathew Herrnegger
Hydrol. Earth Syst. Sci., 24, 4463–4489,Short summary
The USLE is a commonly used model to estimate soil erosion by water. It quantifies soil loss as a product of six inputs representing rainfall erosivity, soil erodibility, slope length and steepness, plant cover, and support practices. Many methods exist to derive these inputs, which can, however, lead to substantial differences in the estimated soil loss. Here, we analyze the effect of different input representations on the estimated soil loss in a large-scale study in Kenya and Uganda.
Benjamin Müller, Matthias Bernhardt, and Karsten Schulz
Hydrol. Earth Syst. Sci. Discuss.,
Manuscript not accepted for further reviewShort summary
Time series of thermal remote sensing images include more information than usually used. Land surface related processes are combined into a single image. Activity of these processes change from image to image. Thus, information on land surface characteristics is to be found
somewhere in betweenthe images. We provide an algorithm to test the presence of such characteristics within a set of images. The algorithm can be used for process understanding, model evaluation, data assimilation, etc.
Christoph Schürz, Brigitta Hollosi, Christoph Matulla, Alexander Pressl, Thomas Ertl, Karsten Schulz, and Bano Mehdi
Hydrol. Earth Syst. Sci., 23, 1211–1244,Short summary
For two Austrian catchments we simulated discharge and nitrate-nitrogen (NO3-N) considering future changes of climate, land use, and point source emissions together with the impact of different setups and parametrizations of the implemented eco-hydrological model. In a comprehensive analysis we identified the dominant sources of uncertainty for the simulation of discharge and NO3-N and further examined how specific properties of the model inputs control the future simulation results.
Abolanle E. Odusanya, Bano Mehdi, Christoph Schürz, Adebayo O. Oke, Olufiropo S. Awokola, Julius A. Awomeso, Joseph O. Adejuwon, and Karsten Schulz
Hydrol. Earth Syst. Sci., 23, 1113–1144,Short summary
The main objective was to calibrate and validate the eco-hydrological model Soil and Water Assessment Tool (SWAT) with satellite-based actual evapotranspiration (AET) data for the data-sparse Ogun River Basin (20 292 km2) located in southwestern Nigeria. The SWAT model, composed of the Hargreaves PET equation and calibrated using the GLEAM_v3.0a data (GS1), performed well for the simulation of AET and provided a good level of confidence for using the SWAT model as a decision support tool.
Maik Renner, Claire Brenner, Kaniska Mallick, Hans-Dieter Wizemann, Luigi Conte, Ivonne Trebs, Jianhui Wei, Volker Wulfmeyer, Karsten Schulz, and Axel Kleidon
Hydrol. Earth Syst. Sci., 23, 515–535,Short summary
We estimate the phase lag of surface states and heat fluxes to incoming solar radiation at the sub-daily timescale. While evapotranspiration reveals a minor phase lag, the vapor pressure deficit used as input by Penman–Monteith approaches shows a large phase lag. The surface-to-air temperature gradient used by energy balance residual approaches shows a small phase shift in agreement with the sensible heat flux and thus explains the better correlation of these models at the sub-daily timescale.
Lu Gao, Jianhui Wei, Lingxiao Wang, Matthias Bernhardt, Karsten Schulz, and Xingwei Chen
Earth Syst. Sci. Data, 10, 2097–2114,Short summary
High-resolution temperature data sets are important for the Chinese Tian Shan, which has a complex ecological environment system. This study presents a unique high-resolution (1 km, 6-hourly) air temperature data set for this area from 1979 to 2016 based on a robust statistical downscaling framework. The strongest advantage of this method is its independence of local meteorological stations due to a model internal, vertical lapse rate scheme. This method was validated for other mountains.
Frederik Kratzert, Daniel Klotz, Claire Brenner, Karsten Schulz, and Mathew Herrnegger
Hydrol. Earth Syst. Sci., 22, 6005–6022,Short summary
In this paper, we propose a novel data-driven approach for rainfall–runoff modelling, using the long short-term memory (LSTM) network, a special type of recurrent neural network. We show in three different experiments that this network is able to learn to predict the discharge purely from meteorological input parameters (such as precipitation or temperature) as accurately as (or better than) the well-established Sacramento Soil Moisture Accounting model, coupled with the Snow-17 snow model.
Stefan Härer, Matthias Bernhardt, Matthias Siebers, and Karsten Schulz
The Cryosphere, 12, 1629–1642,Short summary
The paper presents an approach which can be used to process satellite-based snow cover maps with a higher-than-today accuracy at the local scale. Many of the current satellite-based snow maps are using the NDSI with a threshold as a tool for deciding if there is snow on the ground or not. The presented study has shown that, firstly, using the standard threshold of 0.4 can result in significant derivations at the local scale and that, secondly, the deviations become smaller for coarser scales.
Karsten Schulz, Reinhard Burgholzer, Daniel Klotz, Johannes Wesemann, and Mathew Herrnegger
Hydrol. Earth Syst. Sci., 22, 2607–2613,Short summary
The unit hydrograph has been one of the most widely employed modelling techniques to predict rainfall-runoff behaviour of hydrological catchments. We developed a lecture theatre experiment including some student involvement to illustrate the principles behind this modelling technique. The experiment only uses very simple and cheap material including a set of plastic balls (representing rainfall), magnetic stripes (tacking the balls to the white board) and sieves (for ball/water gauging).
Benjamin Müller, Matthias Bernhardt, Conrad Jackisch, and Karsten Schulz
Hydrol. Earth Syst. Sci., 20, 3765–3775,Short summary
A technology for the spatial derivation of soil texture classes is presented. Information about soil texture is key for predicting the local and regional hydrological cycle. It is needed for the calculation of soil water movement, the share of surface runoff, the evapotranspiration rate and others. Nevertheless, the derivation of soil texture classes is expensive and time-consuming. The presented technique uses soil samples and remotely sensed data for estimating their spatial distribution.
S. Härer, M. Bernhardt, and K. Schulz
Geosci. Model Dev., 9, 307–321,Short summary
This paper describes a new method to produce spatially and temporally calibrated NDSI-based satellite snow cover maps utilizing simultaneously captured terrestrial photographs as in situ information. First results confirm a high quality of the produced satellite snow cover maps and emphasize the need for calibration of the NDSI threshold value to ensure a high accuracy and reproduciblity. The software "PRACTISE V.2.1" was developed to automatically process the photographs and satellite images.
M. Herrnegger, H. P. Nachtnebel, and K. Schulz
Hydrol. Earth Syst. Sci., 19, 4619–4639,Short summary
Especially in alpine catchments, areal rainfall estimates often exhibit large errors. Runoff measurements are, on the other hand, one of the most robust observations within the hydrological cycle. We therefore calculate mean catchment rainfall by inverting an HBV-type rainfall-runoff model, using runoff observations as input. The inverse model may e.g. be used to analyse rainfall conditions of extreme flood events or estimation of snowmelt contribution.
B. Müller, M. Bernhardt, and K. Schulz
Hydrol. Earth Syst. Sci., 18, 5345–5359,Short summary
We present a method to define hydrological landscape units by a time series of thermal infrared satellite data. Land surface temperature is calculated for 28 images in 12 years for a catchment in Luxembourg. Pattern measures show spatio-temporal persistency; principle component analysis extracts relevant patterns. Functional units represent similar behaving entities based on a representative set of images. Resulting classification and patterns are discussed regarding potential applications.
E. Zehe, U. Ehret, L. Pfister, T. Blume, B. Schröder, M. Westhoff, C. Jackisch, S. J. Schymanski, M. Weiler, K. Schulz, N. Allroggen, J. Tronicke, L. van Schaik, P. Dietrich, U. Scherer, J. Eccard, V. Wulfmeyer, and A. Kleidon
Hydrol. Earth Syst. Sci., 18, 4635–4655,
S. Härer, M. Bernhardt, J. G. Corripio, and K. Schulz
Geosci. Model Dev., 6, 837–848,
Related subject area
Subject: Rivers and Lakes | Techniques and Approaches: Modelling approachesStreamflow drought: implication of drought definitions and its application for drought forecastingQuantifying floodwater impacts on a lake water budget via volume-dependent transient stable isotope mass balanceRiver runoff in Switzerland in a changing climate – changes in moderate extremes and their seasonalityRiver runoff in Switzerland in a changing climate – runoff regime changes and their time of emergenceBathymetry and latitude modify lake warming under iceLake thermal structure drives interannual variability in summer anoxia dynamics in a eutrophic lake over 37 yearsReservoir evaporation in a Mediterranean climate: comparing direct methods in Alqueva Reservoir, PortugalDiverging hydrological drought traits over Europe with global warmingAnthropogenic influence on the Rhine water temperaturesA new form of the Saint-Venant equations for variable topographySimulations of future changes in thermal structure of Lake Erken: proof of concept for ISIMIP2b lake sector local simulation strategyAssessment of the geomorphic effectiveness of controlled floods in a braided river using a reduced-complexity numerical modelWorldwide lake level trends and responses to background climate variationModeling inorganic carbon dynamics in the Seine River continuum in FranceA data-based predictive model for spatiotemporal variability in stream water qualityFlooding in the Mekong Delta: the impact of dyke systems on downstream hydrodynamicsReconstruction of the 1941 GLOF process chain at Lake Palcacocha (Cordillera Blanca, Peru)Historical modelling of changes in Lake Erken thermal conditionsImproving lake mixing process simulations in the Community Land Model by using K profile parameterizationUpgraded global mapping information for earth system modelling: an application to surface water depth at the ECMWFSediment transport modelling in riverine environments: on the importance of grain-size distribution, sediment density, and suspended sediment concentrations at the upstream boundaryReplication of ecologically relevant hydrological indicators following a modified covariance approach to hydrological model parameterizationLidar-based approaches for estimating solar insolation in heavily forested streamsNumerical study on the response of the largest lake in China to climate changeUnraveling the hydrological budget of isolated and seasonally contrasted subtropical lakesFuture projections of temperature and mixing regime of European temperate lakesConservative finite-volume forms of the Saint-Venant equations for hydrology and urban drainageModelling Lake Titicaca's daily and monthly evaporationPrincipal components of thermal regimes in mountain river networksModelling the water balance of Lake Victoria (East Africa) – Part 1: Observational analysisModelling the water balance of Lake Victoria (East Africa) – Part 2: Future projectionsBreeze effects at a large artificial lake: summer case studyImpacts of changing hydrology on permanent gully growth: experimental resultsExtrapolating regional probability of drying of headwater streams using discrete observations and gauging networksPassive acoustic measurement of bedload grain size distribution using self-generated noiseCan riparian vegetation shade mitigate the expected rise in stream temperatures due to climate change during heat waves in a human-impacted pre-alpine river?Tributaries affect the thermal response of lakes to climate changeParameter sensitivity analysis of a 1-D cold region lake model for land-surface schemesResponse of water temperatures and stratification to changing climate in three lakes with different morphometryImpacts of changes in groundwater recharge on the isotopic composition and geochemistry of seasonally ice-covered lakes: insights for sustainable managementModel simulations of potential contribution of the proposed Huangpu Gate to flood control in the Lake Taihu basin of ChinaConsistent initial conditions for the Saint-Venant equations in river network modelingCan spatial statistical river temperature models be transferred between catchments?Climate and basin drivers of seasonal river water temperature dynamicsExamining the impacts of precipitation isotope input (δ18Oppt) on distributed, tracer-aided hydrological modellingSeasonal thermal regime and climatic trends in lakes of the Tibetan highlandsThe internal seiche field in the changing South Aral Sea (2006–2013)A practical approach to lake water density from electrical conductivity and temperatureComment on "Using groundwater age and hydrochemistry to understand sources and dynamics of nutrient contamination through the catchment into Lake Rotorua, New Zealand" by Morgenstern et al. (2015)Trends and abrupt changes in 104 years of ice cover and water temperature in a dimictic lake in response to air temperature, wind speed, and water clarity drivers
Samuel J. Sutanto and Henny A. J. Van Lanen
Hydrol. Earth Syst. Sci., 25, 3991–4023,Short summary
This paper provides a comprehensive overview of the differences within streamflow droughts derived using different identification approaches, namely the variable threshold, fixed threshold, and the Standardized Streamflow Index, including an analysis of both historical drought and implications for forecasting. Our results clearly show that streamflow droughts derived from different approaches deviate from each other in terms of drought occurrence, timing, duration, and deficit volume.
Janie Masse-Dufresne, Florent Barbecot, Paul Baudron, and John Gibson
Hydrol. Earth Syst. Sci., 25, 3731–3757,Short summary
A volume-dependent transient isotopic mass balance model was developed for an artificial lake in Canada, in a context where direct measurements of surface water fluxes are difficult. It revealed that floodwater inputs affected the dynamics of the lake in spring but also significantly influenced the long-term water balance due to temporary subsurface storage of floodwater. Such models are paramount for understanding the vulnerability of lakes to changes in groundwater quantity and quality.
Regula Muelchi, Ole Rössler, Jan Schwanbeck, Rolf Weingartner, and Olivia Martius
Hydrol. Earth Syst. Sci., 25, 3577–3594,Short summary
This study analyses changes in magnitude, frequency, and seasonality of moderate low and high flows for 93 catchments in Switzerland. In lower-lying catchments (below 1500 m a.s.l.), moderate low-flow magnitude (frequency) will decrease (increase). In Alpine catchments (above 1500 m a.s.l.), moderate low-flow magnitude (frequency) will increase (decrease). Moderate high flows tend to occur more frequent, and their magnitude increases in most catchments except some Alpine catchments.
Regula Muelchi, Ole Rössler, Jan Schwanbeck, Rolf Weingartner, and Olivia Martius
Hydrol. Earth Syst. Sci., 25, 3071–3086,Short summary
Runoff regimes in Switzerland will change significantly under climate change. Projected changes are strongly elevation dependent with earlier time of emergence and stronger changes in high-elevation catchments where snowmelt and glacier melt play an important role. The magnitude of change and the climate model agreement on the sign increase with increasing global mean temperatures and stronger emission scenarios. This amplification highlights the importance of climate change mitigation.
Cintia L. Ramón, Hugo N. Ulloa, Tomy Doda, Kraig B. Winters, and Damien Bouffard
Hydrol. Earth Syst. Sci., 25, 1813–1825,Short summary
When solar radiation penetrates the frozen surface of lakes, shallower zones underneath warm faster than deep interior waters. This numerical study shows that the transport of excess heat to the lake interior depends on the lake circulation, affected by Earth's rotation, and controls the lake warming rates and the spatial distribution of the heat flux across the ice–water interface. This work contributes to the understanding of the circulation and thermal structure patterns of ice-covered lakes.
Robert Ladwig, Paul C. Hanson, Hilary A. Dugan, Cayelan C. Carey, Yu Zhang, Lele Shu, Christopher J. Duffy, and Kelly M. Cobourn
Hydrol. Earth Syst. Sci., 25, 1009–1032,Short summary
Using a modeling framework applied to 37 years of dissolved oxygen time series data from Lake Mendota, we identified the timing and intensity of thermal energy stored in the lake water column, the lake's resilience to mixing, and surface primary production as the most important drivers of interannual dynamics of low oxygen concentrations at the lake bottom. Due to climate change, we expect an increase in the spatial and temporal extent of low oxygen concentrations in Lake Mendota.
Carlos Miranda Rodrigues, Madalena Moreira, Rita Cabral Guimarães, and Miguel Potes
Hydrol. Earth Syst. Sci., 24, 5973–5984,Short summary
In Mediterranean environments, evaporation is a key component of reservoir water budgets. Prediction of surface evaporation becomes crucial for adequate reservoir water management. This study provides an applicable method for calculating evaporation based on pan measurements applied at Alqueva Reservoir (southern Portugal), one of the largest artificial lakes in Europe. Moreover, the methodology presented here could be applied to other Mediterranean reservoirs.
Carmelo Cammalleri, Gustavo Naumann, Lorenzo Mentaschi, Bernard Bisselink, Emiliano Gelati, Ad De Roo, and Luc Feyen
Hydrol. Earth Syst. Sci., 24, 5919–5935,Short summary
Climate change is anticipated to alter the demand and supply of water at the earth's surface. This study shows how hydrological droughts will change across Europe with increasing global warming levels, showing that at 3 K global warming an additional 11 million people and 4.5 ×106 ha of agricultural land will be exposed to droughts every year, on average. These effects are mostly located in the Mediterranean and Atlantic regions of Europe.
Alex Zavarsky and Lars Duester
Hydrol. Earth Syst. Sci., 24, 5027–5041,Short summary
River water temperature is an important parameter for water quality and an important variable for physical, chemical and biological processes. River water is also used as a cooling agent by power plants and production facilities. We study long-term trends in river water temperature and correlate them to meteorological influences and power production or economic indices.
Cheng-Wei Yu, Ben R. Hodges, and Frank Liu
Hydrol. Earth Syst. Sci., 24, 4001–4024,Short summary
This study investigates the effects of bottom slope discontinuity on the stability of numerical solutions for the Saint-Venant equations. A new reference slope concept is proposed to ensure smooth source terms and eliminate potential numerical oscillations. It is shown that a simple algebraic transformation of channel geometry provides a smooth reference slope while preserving the correct cross-sectional flow area and the piezometric pressure gradient that drives the flow.
Ana I. Ayala, Simone Moras, and Donald C. Pierson
Hydrol. Earth Syst. Sci., 24, 3311–3330,Short summary
The impacts of different levels of global warming on the thermal structure of Lake Erken are assessed. We used the General Ocean Turbulence Model (GOTM) to simulate water temperature driven by meteorological scenarios supplied by the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) and tested its ability at different frequencies. Then, daily ISIMIP meteorological scenarios were disaggregated and assessed for the effects of climate change on lake thermal structure.
Luca Ziliani, Nicola Surian, Gianluca Botter, and Luca Mao
Hydrol. Earth Syst. Sci., 24, 3229–3250,Short summary
Although geomorphic recovery is a key issue in many rivers worldwide, controlled floods have been rarely designed using geomorphological criteria. An integrated approach is used to assess the effects of different controlled-flood scenarios in a strongly regulated river. None of the controlled-flood strategies provide significant morphological benefits. Nevertheless, this study represents a significant contribution for the management and restoration of highly disturbed rivers.
Benjamin M. Kraemer, Anton Seimon, Rita Adrian, and Peter B. McIntyre
Hydrol. Earth Syst. Sci., 24, 2593–2608,Short summary
Lake levels go up and down due to natural variability in the climate. But the effects of natural variability on lake levels can sometimes be confused for the influence of humans. Here we used long-term data from 200 globally distributed lakes and an advanced statistical approach to show that the effects of natural variability on lake levels can be disentangled from other effects leading to better estimates of long-term changes that may be partially caused by humans.
Audrey Marescaux, Vincent Thieu, Nathalie Gypens, Marie Silvestre, and Josette Garnier
Hydrol. Earth Syst. Sci., 24, 2379–2398,Short summary
Rivers have been recognized as an active part of the carbon cycle where transformations are associated with CO2 outgassing. To understand it, we propose a modeling approach with the biogeochemical model, pyNuts-Riverstrahler. We implemented it on the human-impacted Seine River. Sources of carbon to the river were characterized by field measurements in groundwater and in wastewater. Outgassing was the most important in streams, and peaks were simulated downstream of wastewater treatment effluent.
Danlu Guo, Anna Lintern, J. Angus Webb, Dongryeol Ryu, Ulrike Bende-Michl, Shuci Liu, and Andrew William Western
Hydrol. Earth Syst. Sci., 24, 827–847,Short summary
This study developed predictive models to represent the spatial and temporal variation of stream water quality across Victoria, Australia. The model structures were informed by a data-driven approach, which identified the key controls of water quality variations from long-term records. These models are helpful to identify likely future changes in water quality and, in turn, provide critical information for developing management strategies to improve stream water quality.
Vo Quoc Thanh, Dano Roelvink, Mick van der Wegen, Johan Reyns, Herman Kernkamp, Giap Van Vinh, and Vo Thi Phuong Linh
Hydrol. Earth Syst. Sci., 24, 189–212,Short summary
The Vietnamese Mekong Delta (VMD) is a rice bowl of not only Vietnam, but also the world; agriculture is the main source of livelihood in the delta. The VMD is facing threats related to water management and hydraulic structures. Dykes are built to protect agricultural crops in the floodplains and may influence water regimes downstream in the VMD. If the VMD floodplains are completely protected by dykes, yearly mean water levels could increase by 3 cm (at Can Tho) and 1.5 cm (at My Thuan).
Martin Mergili, Shiva P. Pudasaini, Adam Emmer, Jan-Thomas Fischer, Alejo Cochachin, and Holger Frey
Hydrol. Earth Syst. Sci., 24, 93–114,Short summary
In 1941, the glacial lagoon Lake Palcacocha in the Cordillera Blanca (Peru) drained suddenly. The resulting outburst flood/debris flow consumed another lake and had a disastrous impact on the town of Huaraz 23 km downstream. We reconstuct this event through a numerical model to learn about the possibility of prediction of similar processes in the future. Remaining challenges consist of the complex process interactions and the lack of experience due to the rare occurrence of such process chains.
Simone Moras, Ana I. Ayala, and Don C. Pierson
Hydrol. Earth Syst. Sci., 23, 5001–5016,Short summary
We used a hydrodynamic model to reconstruct daily historical water temperature of Lake Erken (Sweden) between 1961 and 2017 to demonstrate the ongoing effect of climate change on lake thermal conditions. The results show that the lake has warmed most rapidly in the last 30 years and that it is now subject to a longer and more stable stratification. The methods used here to reconstruct historical water temperature records can be easily extended to other lakes.
Qunhui Zhang, Jiming Jin, Xiaochun Wang, Phaedra Budy, Nick Barrett, and Sarah E. Null
Hydrol. Earth Syst. Sci., 23, 4969–4982,Short summary
We improved lake mixing process simulations by applying a vertical mixing scheme, K profile parameterization (KPP), in the Community Land Model (CLM) version 4.5, developed by the National Center for Atmospheric Research. The current vertical mixing scheme in CLM requires an arbitrarily enlarged eddy diffusivity to enhance water mixing. The coupled CLM-KPP considers a boundary layer for eddy development. The improved lake model provides an important tool for lake hydrology and ecosystem studies.
Margarita Choulga, Ekaterina Kourzeneva, Gianpaolo Balsamo, Souhail Boussetta, and Nils Wedi
Hydrol. Earth Syst. Sci., 23, 4051–4076,Short summary
Lakes influence weather and climate of regions, especially if several of them are located close by. Just by using upgraded lake depths, based on new or more recent measurements and geological methods of depth estimation, errors of lake surface water forecasts produced by the European Centre for Medium-Range Weather Forecasts became 12–20 % lower compared with observations for 27 lakes collected by the Finnish Environment Institute. For ice-off date forecasts errors changed insignificantly.
Jérémy Lepesqueur, Renaud Hostache, Núria Martínez-Carreras, Emmanuelle Montargès-Pelletier, and Christophe Hissler
Hydrol. Earth Syst. Sci., 23, 3901–3915,Short summary
This article evaluates the influence of sediment representation in a sediment transport model. A short-term simulation is used to assess how far changing the sediment characteristics in the modelling experiment changes riverbed evolution and sediment redistribution during a small flood event. The study shows in particular that representing sediment with extended grain-size and grain-density distributions allows for improving model accuracy and performances.
Annie Visser-Quinn, Lindsay Beevers, and Sandhya Patidar
Hydrol. Earth Syst. Sci., 23, 3279–3303,Short summary
The ecological impact of changes in river flow may be explored through the simulation of ecologically relevant flow indicators. Traditional approaches to model parameterization are not well-suited for this. To this end, this paper considers the ability of a
modified covariance approach, applied to five hydrologically diverse catchments. An overall improvement in consistency is observed, whilst timing and rate of change represent the best and worst replicated indicators respectively.
Jeffrey J. Richardson, Christian E. Torgersen, and L. Monika Moskal
Hydrol. Earth Syst. Sci., 23, 2813–2822,Short summary
High stream temperatures can be detrimental to the survival of aquatic species such as endangered salmon. Stream temperatures can be reduced by shade provided by trees in riparian areas. Two lidar-based methods were effective at assessing stream shading. These methods can be used in place of expensive field measurements.
Dongsheng Su, Xiuqing Hu, Lijuan Wen, Shihua Lyu, Xiaoqing Gao, Lin Zhao, Zhaoguo Li, Juan Du, and Georgiy Kirillin
Hydrol. Earth Syst. Sci., 23, 2093–2109,Short summary
In this study, freshwater lake model simulation results, verified by satellite and buoy observation data, were used to quantify recent climate change effects on the thermal regime of the largest lake in China. Results indicate that the FLake model can reproduce the lake thermal pattern nicely. The lake surface is warming, while the lake bottom has no significant trend. Climate change also caused an earlier ice-off and later ice-on, leading to an obvious change in the energy balance of the lake.
Chloé Poulin, Bruno Hamelin, Christine Vallet-Coulomb, Guinbe Amngar, Bichara Loukman, Jean-François Cretaux, Jean-Claude Doumnang, Abdallah Mahamat Nour, Guillemette Menot, Florence Sylvestre, and Pierre Deschamps
Hydrol. Earth Syst. Sci., 23, 1705–1724,Short summary
This study investigates the water budget of two intertropical lake systems in the absence of long-term hydrological monitoring. By coupling dry season isotopic data with satellite imagery, we were able to provide quantitative constrains on the hydrological balance and show that these two lake systems can be considered miniature analogs of Lake Chad, making them important targets in the future setup of any large-scale program on the hydro-climatic evolution in the Sahel region.
Tom Shatwell, Wim Thiery, and Georgiy Kirillin
Hydrol. Earth Syst. Sci., 23, 1533–1551,Short summary
We used models to project future temperature and mixing in temperate lakes. Lakes will probably warm faster in winter than in summer, making ice less frequent and altering mixing. We found that the layers that form seasonally in lakes (ice, stratification) and water clarity affect how lakes accumulate heat. Seasonal changes in climate were thus important. This helps us better understand how different lake types respond to warming and which physical changes to expect in the future.
Ben R. Hodges
Hydrol. Earth Syst. Sci., 23, 1281–1304,Short summary
A new derivation of the equations for one-dimensional open-channel flow in rivers and storm drainage systems has been developed. The new approach solves some long-standing problems for obtaining well-behaved solutions with conservation forms of the equations. This research was motivated by the need for highly accurate models of large-scale river networks and the storm drainage systems in megacities. Such models are difficult to create with existing equation forms.
Ramiro Pillco Zolá, Lars Bengtsson, Ronny Berndtsson, Belen Martí-Cardona, Frederic Satgé, Franck Timouk, Marie-Paule Bonnet, Luis Mollericon, Cesar Gamarra, and José Pasapera
Hydrol. Earth Syst. Sci., 23, 657–668,Short summary
The evaporation was computed at a daily time step and compared with the estimated evaporation using mean monthly meteorological observations. We found that the most reliable method of determining the annual lake evaporation is using the heat balance approach.
Daniel J. Isaak, Charles H. Luce, Gwynne L. Chandler, Dona L. Horan, and Sherry P. Wollrab
Hydrol. Earth Syst. Sci., 22, 6225–6240,Short summary
Description of thermal regimes in flowing waters is key to understanding physical processes and improving bioassessments, but has been limited by sparse data sets. Using a large annual temperature data set from a mountainous area of the western U.S., we explored thermal regimes using principle component analysis. A small number of summary metrics adequately represented most of the variation in this data set given strong temporal coherence among sites.
Inne Vanderkelen, Nicole P. M. van Lipzig, and Wim Thiery
Hydrol. Earth Syst. Sci., 22, 5509–5525,Short summary
Lake Victoria is the largest lake in Africa and one of the two major sources of the Nile river. The water level of Lake Victoria is determined by its water balance, consisting of lake precipitation and evaporation, inflow from rivers and lake outflow, controlled by two hydropower dams. Here, we present a water balance model for Lake Victoria, which closely represents the observed lake levels. The model results highlight the sensitivity of the lake level to human operations at the dam.
Inne Vanderkelen, Nicole P. M. van Lipzig, and Wim Thiery
Hydrol. Earth Syst. Sci., 22, 5527–5549,Short summary
Lake Victoria is the second largest freshwater lake in the world and one of the major sources of the Nile River, which is controlled by two hydropower dams. In this paper we estimate the potential consequences of climate change for future water level fluctuations of Lake Victoria. Our results reveal that the operating strategies at the dam are the main controlling factors of future lake levels and that regional climate simulations used in the projections encompass large uncertainties.
Maksim Iakunin, Rui Salgado, and Miguel Potes
Hydrol. Earth Syst. Sci., 22, 5191–5210,Short summary
Lakes and reservoirs can affect local weather regimes but usually it is difficult to trace and assess it. In this work we used the Meso-NH atmospheric model to study the impact of the Alqueva reservoir, one of the largest artificial lakes in western Europe, located in the southeast of Portugal, on meteorological parameters and the formation of a lake breeze system. The magnitude of this impact as well as the intensity of the breeze are shown in the paper.
Stephanie S. Day, Karen B. Gran, and Chris Paola
Hydrol. Earth Syst. Sci., 22, 3261–3273,Short summary
Permanent gullies are deep steep-sided channels that erode as water falls over the upstream end. Erosion of these features is a concern where people and climate change have altered how water moves over the land. This paper analyzes a set of experiments that were used to determine how changing gully flows impact erosion. We found that while increasing the volume of water will increase erosion, changing the flow rate into gullies will not impact the total erosion, but will alter gully shape.
Aurélien Beaufort, Nicolas Lamouroux, Hervé Pella, Thibault Datry, and Eric Sauquet
Hydrol. Earth Syst. Sci., 22, 3033–3051,Short summary
Streams which may stop flowing are poorly gauged. To improve their characterisation, we use an extended network providing monthly visual observations stating whether streams are flowing or not across France. These observations are combined with discharge and groundwater level in models to predict daily regional probability of drying. This approach allows identification of the most impacted regions by flow intermittence and estimation of the probability of drying dynamics over the last 27 years.
Teodor Petrut, Thomas Geay, Cédric Gervaise, Philippe Belleudy, and Sebastien Zanker
Hydrol. Earth Syst. Sci., 22, 767–787,Short summary
Our interest is focused on developing the hydrophone technique to estimate the size of particles transported in rivers. The analytic spectral solution of the impact between rigid particles is used to model the power spectrum of a sediment mixture, or the sediment-generated noise. Estimations of grain size distributions in the Isère River using real measured spectra are successfully validated by the physical sampling techniques. Moreover, the grain size sorting process is revealed by acoustics.
Heidelinde Trimmel, Philipp Weihs, David Leidinger, Herbert Formayer, Gerda Kalny, and Andreas Melcher
Hydrol. Earth Syst. Sci., 22, 437–461,Short summary
In eastern Austria, where air temperature rise is double that recorded globally, stream temperatures of a human-impacted river were simulated during heat waves, as calculated by regional climate models until 2100. An increase of up to 3 °C was predicted – thus exceeding thresholds of resident cold-adapted species. Vegetation management scenarios showed that adding vegetation can reduce both absolute temperatures and its rate of increase but is not able to fully mitigate the expected rise.
Love Råman Vinnå, Alfred Wüest, Massimiliano Zappa, Gabriel Fink, and Damien Bouffard
Hydrol. Earth Syst. Sci., 22, 31–51,Short summary
Responses of inland waters to climate change vary on global and regional scales. Shifts in river discharge regimes act as positive and negative feedbacks in influencing water temperature. The extent of this effect on warming is controlled by the change in river discharge and lake hydraulic residence time. A shift of deep penetrating river intrusions from summer towards winter can potentially counteract the otherwise negative climate effects on deep-water oxygen content.
José-Luis Guerrero, Patricia Pernica, Howard Wheater, Murray Mackay, and Chris Spence
Hydrol. Earth Syst. Sci., 21, 6345–6362,Short summary
Lakes are sentinels of climate change, and an adequate characterization of their feedbacks to the atmosphere could improve climate modeling. These feedbacks, as heat fluxes, can be simulated but are seldom measured, casting doubt on modeling results. Measurements from a small lake in Canada established that the model parameter modulating how much light penetrates the lake dominates model response. This parameter is measurable: improved monitoring could lead to more robust modeling.
Madeline R. Magee and Chin H. Wu
Hydrol. Earth Syst. Sci., 21, 6253–6274,Short summary
Water temperatures and stratification in three morphometrically different lakes over the century are examined. Simulations showed that epilimnetic temperatures increased, hypolimnetic temperatures decreased, the length of the stratified season increased due to earlier stratification onset and later fall overturn, and stability increased. Results showed that wind speed has a large effect on temperature and stratification variables, sometimes greater than changes in air temperature.
Marie Arnoux, Florent Barbecot, Elisabeth Gibert-Brunet, John Gibson, and Aurélie Noret
Hydrol. Earth Syst. Sci., 21, 5875–5889,Short summary
Future climate and recharge change scenarios are tested on lake hydrological balances to simulate changes in lake isotopic signature and water quality. Results highlight that lake water evolution depends mainly on the location and the intensity of recharge changes. Lake water isotopic composition appears to be particularly sensitive to future changes in recharge for lakes with between 50 % and 80 % of groundwater in their total inflows.
Hanghui Zhang, Shuguang Liu, Jianchun Ye, and Pat J.-F. Yeh
Hydrol. Earth Syst. Sci., 21, 5339–5355,Short summary
The Huangpu River, an important river of the basin connecting Lake Taihu upstream and the Yangtze River estuary downstream, drains two-fifths of the entire basin. Constructing an estuary gate is considered an effective solution for flood mitigation. The main objective of this paper is to assess the potential contributions of the proposed Huangpu Gate to the flood control capacity of the basin. Results of quantitative analyses show that the Huangpu Gate is effective at evacuating floodwaters.
Cheng-Wei Yu, Frank Liu, and Ben R. Hodges
Hydrol. Earth Syst. Sci., 21, 4959–4972,Short summary
This research proposes a new method for determining efficient and effective synthetic initial conditions for an unsteady model using the Saint-Venant equations for a large river network. The proposed method increases the efficiency of initial condition convergence up to 3700 times. This can not only help the hydrology research move from empirically based to mechanistic-based models for large river networks but also provide a computationally effective approach for initial conditions for modelers.
Faye L. Jackson, Robert J. Fryer, David M. Hannah, and Iain A. Malcolm
Hydrol. Earth Syst. Sci., 21, 4727–4745,Short summary
River temperature (Tw) is important to fish populations, but one cannot monitor everywhere. Thus, models are used to predict Tw, sometimes in rivers with no data. To date, the accuracy of these predictions has not been determined. We found that models including landscape predictors (e.g. altitude, tree cover) could describe spatial patterns in Tw in other rivers better than those including air temperature. Such findings are critical for developing Tw models that have management application.
Cédric L. R. Laizé, Cristian Bruna Meredith, Michael J. Dunbar, and David M. Hannah
Hydrol. Earth Syst. Sci., 21, 3231–3247,Short summary
Stream temperature controls many river processes, making it vital to know how climate affects it. Climate and stream temperatures at 35 British sites and associated basin properties were used to model climate–water temperature associations and to assess how they are influenced by basins. Associations vary with season and water temperature range. Basin permeability, size, and elevation have the main influence; smaller upland or impermeable basins are the most sensitive to climate.
Carly J. Delavau, Tricia Stadnyk, and Tegan Holmes
Hydrol. Earth Syst. Sci., 21, 2595–2614,Short summary
Hydrological models have large amounts of uncertainty in streamflow predictions. Using extra data (e.g. isotope tracers) helps evaluate whether the model is getting the right answers for the right reasons. In a Canadian basin, three types of isotope in precipitation input are used to drive a tracer-aided model and assess the resulting model uncertainty. This study shows how a tracer-aided model can be used at the larger scale, and that the model can be of value in such regions.
Georgiy Kirillin, Lijuan Wen, and Tom Shatwell
Hydrol. Earth Syst. Sci., 21, 1895–1909,Short summary
We report a first description of the seasonal temperature, mixing, and ice regime in the two largest freshwater lakes of the Tibetan Plateau. We perform a validation of lake model FLake for the parameterization of the Tibetan lake system in regional climate models and present evidence of the absent warming trend in the Tibetan lakes despite significant atmospheric warming. The reason for this unexpected behavior is the significant decrease in solar radiation at the surface.
Elena Roget, Elizaveta Khimchenko, Francesc Forcat, and Peter Zavialov
Hydrol. Earth Syst. Sci., 21, 1093–1105,Short summary
This is the first work on the internal seiches of the South Aral Sea, which is also of general interest as the lake is considered a paradigm for the lakes whose surface levels are decreasing. It is based on field data and numerical simulations. For autumn 2006 and 2013, the fundamental mode is discussed together with other modes which are found to be higher vertical modes. The importance of sea level variation on the development of higher horizontal modes is also discussed.
Santiago Moreira, Martin Schultze, Karsten Rahn, and Bertram Boehrer
Hydrol. Earth Syst. Sci., 20, 2975–2986,Short summary
Density calculations in lakes are essential in the study of stability, stratification and numerical modelling. However, density equations which use fixed constant coefficients are still common in use, despite their limitations. We present a practical approach which can specifically include the effect of dissolved substances by fitting two coefficients. The method has been tested against density equations and density measurements in lakes.
Jonathan M. Abell, David P. Hamilton, and Christopher G. McBride
Hydrol. Earth Syst. Sci., 20, 2395–2401,Short summary
We comment on "Using groundwater age and hydrochemistry to understand sources and dynamics of nutrient contamination through the catchment into Lake Rotorua, New Zealand" by Morgenstern et al. (2015). They propose that "the only effective way to limit algae blooms and improve lake water quality in such environments is by limiting the nitrate load". We outline four reasons why it is important to instead limit both phosphorus and nitrogen loads to this iconic lake, consistent with current policy.
Madeline R. Magee, Chin H. Wu, Dale M. Robertson, Richard C. Lathrop, and David P. Hamilton
Hydrol. Earth Syst. Sci., 20, 1681–1702,Short summary
This paper employs a one-dimensional hydrodynamic ice model to simulate ice cover and thermal structure of dimictic Lake Mendota, WI, USA, over a continuous 104-year period (1911–2014) with the purpose of better understanding how the changing climate will affect lakes. It is shown that air temperature and wind speed changes have occurred in stages and ice cover and lake thermal structure have responded in a nonlinear way to these changes.
Abba, S. I., Hadi, S. J., and Abdullahi, J.: River water modelling prediction using multi-linear regression, artificial neural network, and adaptive neuro-fuzzy inference system techniques, in: Procedia Computer Science, Elsevier B.V., Budapest, Hungary, 75–82, https://doi.org/10.1016/j.procs.2017.11.212, 2017. a, b
Akaike, H.: Information theory as an extension of the likelihood principle., in: Second Akademiai International Symposium on Information Theory, edited by: Petrov, B. N. and Csaki, F., Kiado, Budapest, 267–281, 1973. a
Allaire, J. J. and Tang, Y.: tensorflow: R Interface to “TensorFlow”, available at: https://github.com/rstudio/tensorflow (last access: 13 Jauary 2021), 2020. a
BMLFUW: Hydrological Atlas of Austria, 3rd Edn., Bundesmin-isterium für Land- und Forstwirtschaft, Umwelt und Wasser-wirtschaft, Vienna, Austria, ISBN 3-85437-250-7, 2007. a
Boisneau, C., Moatar, F., Bodin, M., and Boisneau, P.: Does global warming impact on migration patterns and recruitment of Allis shad (Alosa alosa L.) young of the year in the Loire River, France?, in: Fish and Diadromy in Europe (ecology, management, conservation), Springer, Dordrecht, the Netherlands, 179–186, https://doi.org/10.1007/978-1-4020-8548-2_14, 2008. a
Central Hydrographical Bureau: eHYD, available at: https://www.ehyd.gv.at/, last access: 26 May 2021. a
Chen, T., He, T., Benesty, M., Khotilovich, V., Tang, Y., Cho, H., Chen, K., Mitchell, R., Cano, I., Zhou, T., Li, M., Xie, J., Lin, M., Geng, Y., and Li, Y.: xgboost: Extreme Gradient Boosting, available at: https://cran.r-project.org/package=xgboost (last access: 13 January 2021), 2020. a
Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., and Bengio, Y.: Learning phrase representations using RNN encoder-decoder for statistical machine translation, in: Proceedings of the EMNLP 2014 – 2014 Conference on Empirical Methods in Natural Language Processing, 25–29 October 2014, Doha, Qatar, 1724–1734, https://doi.org/10.3115/v1/d14-1179, 2014. a, b
Crisp, D. and Howson, G.: Effect of air temperature upon mean water temperature in streams in the north Pennines and English Lake District, Freshwater Biol., 12, 359–367, https://doi.org/10.1111/j.1365-2427.1982.tb00629.x, 1982. a, b
Fernández-Delgado, M., Cernadas, E., Barro, S., and Amorim, D.: Do we need hundreds of classifiers to solve real world classification problems?, J. Mach. Learn. Res., 15, 3133–3181, 2014. a
Freund, Y. and Schapire, R. E.: A decision-theoretic generalization of on-line learning and an application to boosting, in: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer Verlag, 23–37, https://doi.org/10.1007/3-540-59119-2_166, 1995. a
Gauch, M., Tang, R., Mai, J., Tolson, B., Gharari, S., and Lin, J.: Machine Learning for Streamflow Prediction: Current Status and Future Prospects, 9–13 December 2019, San Francisco, USA, AGU Fall Meeting Abstracts, 2019, H33L–2127, 2019. a
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J. N.: The ERA5 global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020. a
Hsu, C.-W., Chang, C.-C., and Lin, C.-J.: A Practical Guide to Support Vector Classification, Tech. Rep., Taipei, 2003. a
Jackson, F. L., Fryer, R. J., Hannah, D. M., Millar, C. P., and Malcolm, I. A.: A spatio-temporal statistical model of maximum daily river temperatures to inform the management of Scotland's Atlantic salmon rivers under climate change, Sci. Total Environ., 612, 1543–1558, https://doi.org/10.1016/j.scitotenv.2017.09.010, 2018. a, b, c
Joslyn, K.: Water quality factor prediction using supervised machine learning REU Final Reports, 6, available at: https://archives.pdx.edu/ds/psu/26231 (last access: 26 May 2021), 2018. a
Kratzert, F., Klotz, D., Shalev, G., Klambauer, G., Hochreiter, S., and Nearing, G.: Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets, Hydrol. Earth Syst. Sci., 23, 5089–5110, https://doi.org/10.5194/hess-23-5089-2019, 2019. a, b
Kuhn, M.: caret: Classification and Regression Training, R package version 6.0-86, available at: https://CRAN.R-project.org/package=caret (last access: 13 Jaunary 2021), 2020. a
Neumann, D. W., Rajagopalan, B., and Zagona, E. A.: Regression model for daily maximum stream temperature, J. Environ. Eng., 129, 667–674, https://doi.org/10.1061/(ASCE)0733-9372(2003)129:7(667), 2003. a
Nielsen, D.: Tree Boosting With XGBoost: Why does XGBoost win every machine learning competition?, Master's Thesis, Norwegian University of Science and Technology, Norway, 98 pp., 2016. a
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R. Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., and Duchesnay, E.: Scikit-learn: Machine Learning in Python, Tech. Rep., available at: https://hal.inria.fr/hal-00650905v2 (last access: 4 December 2020), 2011. a
R Core Team: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria, available at: https://www.r-project.org/ (last access: 13 January 2021), 2020. a
Razafimaharo, C., Krähenmann, S., Höpp, S., Rauthe, M., and Deutschländer, T.: New high-resolution gridded dataset of daily mean, minimum, and maximum temperature and relative humidity for Central Europe (HYRAS), Theor. Appl. Climatol., 142, 1531–1553, https://doi.org/10.1007/s00704-020-03388-w, 2020. a
Risley, J. C., Roehl Jr., E. A., and Conrads, P. A.: Estimating Water Temperatures in Small Streams in Estimating Water Temperatures in Small Streams in Western Oregon Using Neural Network Models, Tech. Rep., USGS Water- Resources Investigation Report 02-4218, https://doi.org/10.3133/wri024218, 2003. a
Srivastava, N., Hinton, G., Krizhevsky, A., and Salakhutdinov, R.: Dropout: A Simple Way to Prevent Neural Networks from Overfitting, Tech. Rep., 15, 1929-−1958, available at: http://jmlr.org/papers/v15/srivastava14a.html (last access: 7 August 2020), 2014. a
Stefan, H. G. and Preud'homme, E. B.: Stream temperature estimation from air temperature, J. Am. Water Resour. As., 29, 27–45, https://doi.org/10.1111/j.1752-1688.1993.tb01502.x, 1993. a, b
Stevens, H., Ficke, J., and Smoot, G.: Techniques of water-resources investigations of the US Geological Survey, US Government Printing Office, Washington, 65 pp., 1975. a
Tavares, M. H., Cunha, A. H. F., Motta-Marques, D., Ruhoff, A. L., Fragoso, C. R., Munar, A. M., and Bonnet, M. P.: Derivation of consistent, continuous daily river temperature data series by combining remote sensing and water temperature models, Remote Sens. Environ., 241, 111721, https://doi.org/10.1016/j.rse.2020.111721, 2020. a, b
Webb, B. W. and Zhang, Y.: Spatial and seasonal variability in the components of the river heat budget, Hydrol. Process., 11, 79–101, https://doi.org/10.1002/(sici)1099-1085(199701)11:1<79::aid-hyp404>3.0.co;2-n, 1997. a
Wenger, S. J., Isaak, D. J., Dunham, J. B., Fausch, K. D., Luce, C. H., Neville, H. M., Rieman, B. E., Young, M. K., Nagel, D. E., Horan, D. L., and Chandler, G. L.: Role of climate and invasive species in structuring trout distributions in the interior Columbia River Basin, USA, Can. J. Fish. Aquat. Sci., 68, 988–1008, https://doi.org/10.1139/f2011-034, 2011. a
Wickham, H.: ggplot2: Elegant Graphics for Data Analysis, Springer-Verlag, New York, USA, available at: https://ggplot2.tidyverse.org (last access: 7 August 2020), 2016. a
Zentralanstalt für Meteorologie und Geodynamik: ZAMG homepage, available at: https://www.zamg.ac.at, last access: 26 May 2021. a
Zhu, S., Heddam, S., Nyarko, E. K., Hadzima-Nyarko, M., Piccolroaz, S., and Wu, S.: Modeling daily water temperature for rivers: comparison between adaptive neuro-fuzzy inference systems and artificial neural networks models, Environ. Sci. Pollut. R., 26, 402–420, https://doi.org/10.1007/s11356-018-3650-2, 2019b. a, b
In this study we developed machine learning approaches for daily river water temperature prediction, using different data preprocessing methods, six model types, a range of different data inputs and 10 study catchments. By comparing to current state-of-the-art models, we could show a significant improvement of prediction performance of the tested approaches. Furthermore, we could gain insight into the relationships between model types, input data and predicted stream water temperature.
In this study we developed machine learning approaches for daily river water temperature...