scholarly journals A general differential split-sample test to select sub-periods of discontinuous years gathering similar to different climate conditions

MethodsX ◽  
2020 ◽  
Vol 7 ◽  
pp. 101008
Author(s):  
Hamouda Dakhlaoui ◽  
Denis Ruelland ◽  
Yves Tramblay
Keyword(s):  
Climate Conditions ◽  
Split Sample ◽  
Sample Test

scholarly journals Is bias correction of Regional Climate Model (RCM) simulations possible for non-stationary conditions?

2012 ◽  
Vol 9 (11) ◽  
pp. 12765-12795 ◽  
Author(s):  
C. Teutschbein ◽  
J. Seibert
Keyword(s):  
Climate Change ◽  
Bias Correction ◽  
Climate Change Impact ◽  
Climate Model ◽  
Regional Climate ◽  
Climate Conditions ◽  
Impact Studies ◽  
Split Sample ◽  
Sample Test ◽  
Change Impact

Abstract. In hydrological climate-change impact studies, Regional Climate Models (RCMs) are commonly used to transfer large-scale Global Climate Model (GCM) data to smaller scales and to provide more detailed regional information. However, there are often considerable biases in RCM simulations, which have led to the development of a number of bias correction approaches to provide more realistic climate simulations for impact studies. Bias correction procedures rely on the assumption that RCM biases do not change over time, because correction algorithms and their parameterizations are derived for current climate conditions and assumed to apply also for future climate conditions. This underlying assumption of bias stationarity is the main concern when using bias correction procedures. It is in principle not possible to test whether this assumption is actually fulfilled for future climate conditions. In this study, however, we demonstrate that it is possible to evaluate how well bias correction methods perform for conditions different from those used for calibration. For five Swedish catchments, several time series of RCM simulated precipitation and temperature were obtained from the ENSEMBLES data base and different commonly-used bias correction methods were applied. We then performed a differential split-sample test by dividing the data series into cold and warm respective dry and wet years. This enabled us to evaluate the performance of different bias correction procedures under systematically varying climate conditions. The differential split-sample test resulted in a large spread and a clear bias for some of the correction methods during validation years. More advanced correction methods such as distribution mapping performed relatively well even in the validation period, whereas simpler approaches resulted in the largest deviations and least reliable corrections for changed conditions. Therefore, we question the use of simple bias correction methods such as the widely used delta-change approach and linear scaling for RCM-based climate-change impact studies and recommend using higher-skill bias correction methods.

scholarly journals Is bias correction of regional climate model (RCM) simulations possible for non-stationary conditions?

2013 ◽  
Vol 17 (12) ◽  
pp. 5061-5077 ◽  
Author(s):  
C. Teutschbein ◽  
J. Seibert
Keyword(s):  
Climate Change ◽  
Bias Correction ◽  
Climate Change Impact ◽  
Climate Model ◽  
Regional Climate ◽  
Model Errors ◽  
Climate Conditions ◽  
Split Sample ◽  
Sample Test ◽  
Change Impact

Abstract. In hydrological climate-change impact studies, regional climate models (RCMs) are commonly used to transfer large-scale global climate model (GCM) data to smaller scales and to provide more detailed regional information. Due to systematic and random model errors, however, RCM simulations often show considerable deviations from observations. This has led to the development of a number of correction approaches that rely on the assumption that RCM errors do not change over time. It is in principle not possible to test whether this underlying assumption of error stationarity is actually fulfilled for future climate conditions. In this study, however, we demonstrate that it is possible to evaluate how well correction methods perform for conditions different from those used for calibration with the relatively simple differential split-sample test. For five Swedish catchments, precipitation and temperature simulations from 15 different RCMs driven by ERA40 (the 40 yr reanalysis product of the European Centre for Medium-Range Weather Forecasts (ECMWF)) were corrected with different commonly used bias correction methods. We then performed differential split-sample tests by dividing the data series into cold and warm respective dry and wet years. This enabled us to cross-evaluate the performance of different correction procedures under systematically varying climate conditions. The differential split-sample test identified major differences in the ability of the applied correction methods to reduce model errors and to cope with non-stationary biases. More advanced correction methods performed better, whereas large deviations remained for climate model simulations corrected with simpler approaches. Therefore, we question the use of simple correction methods such as the widely used delta-change approach and linear transformation for RCM-based climate-change impact studies. Instead, we recommend using higher-skill correction methods such as distribution mapping.

scholarly journals Multimodel evaluation of twenty lumped hydrological models under contrasted climate conditions

2011 ◽  
Vol 8 (6) ◽  
pp. 10895-10933
Author(s):  
G. Seiller ◽  
F. Anctil ◽  
C. Perrin
Keyword(s):  
Test Procedure ◽  
Model Performance ◽  
Structural Diversity ◽  
Model Ensemble ◽  
Climate Conditions ◽  
Split Sample ◽  
Sample Test ◽  
Model Efficiency ◽  
Improved Performance ◽  
The Individual

Abstract. This paper proposes a methodology to interpret hydrological projections in a climate change context and to quantify model suitability as well as their potential transposability in time. This is achieved by applying the Differential Split Sample Test procedure on twenty lumped conceptual models, for two different catchments, in the Province of Québec (Canada) and in the State of Bavaria (Germany). First, a calibration/validation procedure was applied on four historical non-continuous periods with contrasted climate conditions. Then, model efficiency was quantified individually (for each model) and collectively (for the model ensemble). The individual analysis evaluated model performance and robustness. The ensemble investigation, based on the average of simulated discharges, focused on the twenty-member ensemble and all possible model subsets. Results showed that using a single model without performing a Differential Split Sample Test may provide hazardous results in terms of climate transposability. Overall, some models turned out as a good compromise in terms of performance and robustness, but never as much as the twenty-model ensemble. Model subsets offered yet improved performance and structural diversity, but at the expanse of spatial transposability.

RAT (Robustness Assessment Test): a straightforward evaluation of hydrological model robustness to a changing climate

2021 ◽  
Author(s):  
Vazken Andréassian ◽  
Léonard Santos ◽  
Torben Sonnenborg ◽  
Alban de Lavenne ◽  
Göran Lindström ◽  
...  
Keyword(s):  
Climate Change ◽  
Hydrological Model ◽  
Hydrological Models ◽  
Climate Variables ◽  
Data Set ◽  
Hydrological Characteristics ◽  
Changing Climate ◽  
Split Sample ◽  
Sample Test ◽  
Model Robustness

<p>Hydrological models are increasingly used under evolving climatic conditions. They should thus be evaluated regarding their temporal transferability (application in different time periods) and extrapolation capacity (application beyond the range of known past conditions). In theory, parameters of hydrological models are independent of climate. In practice, however, many published studies based on the Split-Sample Test (Klemeš, 1986), have shown that model performances decrease systematically when it is used out of its calibration period. The RAT test proposed here aims at evaluating model robustness to a changing climate by assessing potential undesirable dependencies of hydrological model performances to climate variables. The test compares, over a long data period, the annual value of several climate variables (temperature, precipitation and aridity index) and the bias of the model over each year. If a significant relation exists between the climatic variable and the bias, the model is not considered to be robust to climate change on the catchment. The test has been compared to the Generalized Split-Sample Test (Coron et al., 2012) and showed similar results.</p><p>Here, we report on a large scale application of the test for three hydrological models with different level of complexity (GR6J, HYPE, MIKE-SHE) on a data set of 352 catchments in Denmark, France and Sweden. The results show that the test behaves differently given the evaluated variable (be temperature, precipitation or aridity) and the hydrological characteristics of each catchment. They also show that, although of different level of complexity, the robustness of the three models is similar on the overall data set. However, they are not robust on the same catchments and, then, are not sensitive to the same hydrological characteristics. This example highlights the applicability of the RAT test regardless of the model set-up and calibration procedure and its ability to provide a first evaluation of the model robustness to climate change.</p><p> </p><p><strong>References</strong></p><p>Coron, L., V. Andréassian, C. Perrin, J. Lerat, J. Vaze, M. Bourqui, and F. Hendrickx, 2012. Crash testing hydrological models in contrasted climate conditions: An experiment on 216 Australian catchments, Water Resour. Res., 48, W05552, doi:10.1029/2011WR011721</p><p>Klemeš, V., 1986. Operational testing of hydrological simulation models, Hydrol. Sci. J., 31, 13–24, doi:10.1080/02626668609491024</p><p> </p>

scholarly journals Process-based modelling to evaluate simulated groundwater levels and frequencies in a Chalk catchment in south-western England

2018 ◽  
Vol 18 (2) ◽  
pp. 445-461 ◽  
Author(s):  
Simon Brenner ◽  
Gemma Coxon ◽  
Nicholas J. K. Howden ◽  
Jim Freer ◽  
Andreas Hartmann
Keyword(s):  
Groundwater Level ◽  
Climate Changes ◽  
Data Availability ◽  
Model Complexity ◽  
High Porosity ◽  
Groundwater Levels ◽  
Subsurface Heterogeneity ◽  
Split Sample ◽  
Sample Test ◽  
Simulated Groundwater

Abstract. Chalk aquifers are an important source of drinking water in the UK. Due to their properties, they are particularly vulnerable to groundwater-related hazards like floods and droughts. Understanding and predicting groundwater levels is therefore important for effective and safe water management. Chalk is known for its high porosity and, due to its dissolvability, exposed to karstification and strong subsurface heterogeneity. To cope with the karstic heterogeneity and limited data availability, specialised modelling approaches are required that balance model complexity and data availability. In this study, we present a novel approach to evaluate simulated groundwater level frequencies derived from a semi-distributed karst model that represents subsurface heterogeneity by distribution functions. Simulated groundwater storages are transferred into groundwater levels using evidence from different observations wells. Using a percentile approach we can assess the number of days exceeding or falling below selected groundwater level percentiles. Firstly, we evaluate the performance of the model when simulating groundwater level time series using a spilt sample test and parameter identifiability analysis. Secondly, we apply a split sample test to the simulated groundwater level percentiles to explore the performance in predicting groundwater level exceedances. We show that the model provides robust simulations of discharge and groundwater levels at three observation wells at a test site in a chalk-dominated catchment in south-western England. The second split sample test also indicates that the percentile approach is able to reliably predict groundwater level exceedances across all considered timescales up to their 75th percentile. However, when looking at the 90th percentile, it only provides acceptable predictions for long time periods and it fails when the 95th percentile of groundwater exceedance levels is considered. By modifying the historic forcings of our model according to expected future climate changes, we create simple climate scenarios and we show that the projected climate changes may lead to generally lower groundwater levels and a reduction of exceedances of high groundwater level percentiles.

scholarly journals A percentile approach to evaluate simulated groundwater levels and frequencies in a Chalk catchment in Southwest England

2016 ◽  
Author(s):  
Simon Brenner ◽  
Gemma Coxon ◽  
Nicholas J. K. Howden ◽  
Jim Freer ◽  
Andreas Hartmann
Keyword(s):  
Groundwater Level ◽  
Climate Changes ◽  
Data Availability ◽  
Model Complexity ◽  
High Porosity ◽  
Groundwater Levels ◽  
Subsurface Heterogeneity ◽  
Split Sample ◽  
Sample Test ◽  
Simulated Groundwater

Abstract. Chalk aquifers are an important source of drinking water in the UK. Understanding and predicting groundwater levels is therefore important for effective water management of this resource. Chalk is known for its high porosity and, due to its dissolvability, exposed to karstification and strong subsurface heterogeneity. To cope with the karstic heterogeneity and limited data availability, specialised modelling approaches are required that balance model complexity and data availability. In this study we present a novel approach to simulate groundwater level frequency distributions with a semi-distributed karst model that represents subsurface heterogeneity by distribution functions. Simulated groundwater storages are transferred into groundwater levels using evidence from different observations wells. Using a newly developed percentile approach we can simulate the number of days exceeding or falling below selected groundwater level percentiles. Firstly, we evaluate the performance of the model to simulate three groundwater time series by a spilt sample test and parameter identifiability analysis. Secondly, we apply a split sample test on the simulated groundwater level percentiles to explore the performance in predicting groundwater level exceedances. We show that the model provides robust simulations of discharge and groundwater levels at 3 observation wells at a test site in chalk dominated catchment in Southwest England. The second split sample test also indicates that percentile approach is able to reliably predict groundwater level exceedances across all considered time scales up to their 75th percentile. However, when looking at the 90th percentile, it only provides acceptable predictions for the longest available time scale and it fails when the 95th percentile of groundwater exceedance levels is considered. Modifying the historic forcings of our model according to expected future climate changes, we create simple climate scenarios and we show that the projected climate changes may lead to generally lower groundwater levels and a reduction of exceedances of high groundwater level percentiles.

Integration of Hidden Markov States in a hydrological model calibration/validation protocol

2021 ◽  
Author(s):  
Étienne Guilpart ◽  
Vahid Espanmanesh ◽  
Amaury Tilmant ◽  
François Anctil
Keyword(s):  
Time Series ◽  
Hidden Markov Models ◽  
Model Calibration ◽  
Hydrological Model ◽  
Markov Models ◽  
Hidden Markov ◽  
Change Points ◽  
Climate Trend ◽  
Split Sample ◽  
Sample Test

<p>Due to climate changes, the stationary assumption in hydrology has become obsolete. Moreover, the uncertainty regarding the future evolution of the Earth's climate and its impact on flow regimes is still large. Over the last decade, new risk management approaches have been proposed to support water resources planning under deep uncertainty. Those approaches rely at some point on a hydrological model to derive time series of streamflows for various hydro-climatic scenarios. One of the key issue is to make sure that the hydrological model is robust, i.e. that it performs well over contrasted hydro-climatic conditions. The differential split-sample test principle proposed by Klemes in 1986 recommends partitioning the time series into numerous and independent subperiods with different stationary climate features. Then, the hydrological model calibration is achieved on a specific climate period, and the validation on other(s). Classical detection methods commonly used to partition the times series, such as Mann-Kendall test or Pettitt test, can only detect a single change point, and thus are unable to handle complex climate sequences with multiple change points. We propose a calibration/validation protocol of hydrological models which rely on both the differential split-sample test and on an Hidden Markov Model to identify a succession of subsequences in a time series based on the state of the underlying process. We applied the proposed protocol on the Senegal River (West Africa). The hydrological model used is the conceptual GR2M model. Results show that (i) when the river discharges time series does not display a clear climate trend, and have multiple change points, classical rupture tests are not suitable. Hidden Markov Models are a good alternative as long as the climate sub-sequences are long enough (typically around 30 years or more); (ii) including a Hidden Markov Models in such protocol open up the range of possibilities for calibrate/validate, which can lead to an enhancement of the criterion function (but not necessarily).</p><p>Klemes, V.: Operational testing of hydrological simulation models, Hydrological Sciences Journal, 31, 13-24, 415 https://doi.org/10.1080/02626668609491024, 1986.</p>

Sign in / Sign up

Export Citation Format

Share Document