Climate Change Impact Assessment Studies Research Articles

The Effect of Climate Change on the Hydropower Potential in the Kunhar River Watershed, Pakistan

Climate change plays a vital role in the hydrology of any river basin, which may have multidimensional consequences. There is a need to conduct climate change impact assessment studies with updated models and scenarios. This study aimed to assess the impact of climate change on the streamflow and hydropower in Pakistan’s Kunhar River basin. Three general circulation models (GCMs), under two Shared Socioeconomic Pathway scenarios (SSPs 2–45 and 5–85), the Soil and Water Assessment Tool, and the flow duration curve were used to project the change in climatic parameters, streamflow, and hydropower potential, respectively. The findings indicated that in the 2080s, the precipitation, maximum, and minimum temperatures are projected to increase by 10%, 2.0 °C, and 3.0 °C under the SSP 2–45 scenario and are projected to increase by 8%, 3.7 °C, and 4.4 °C under the SSP 5–85 scenario, respectively. The annual streamflow may increase by 15 to 11%, and the seasonal fluctuations are more likely to be dominant compared with the annual fluctuations. The hydropower potential will probably increase by 24 to 16% under the SSP 2–45 and 5–85 scenarios in the 2080s. However, seasonal changes in streamflow and hydropower may impact the hydropower plant operation in the basin. The Kunhar River’s hydrology may change from snow-fed to a rainfall–runoff river.

A novel multi-scale parameter estimation approach to the Hargreaves-Samani equation for estimation of Penman-Monteith reference evapotranspiration

The main focus of this study is to develop a multi-scale surrogate model for the FAO-56 Penman-Monteith (PM) evapotranspiration (ETo) using Hargreaves-Samani (HS) equation, which uses only temperature as a hydrometeorological variable to estimate ET. This feature is particularly useful for scarce data regions and climate change impact assessment studies, where the direct estimation of ETo from the PM equation can be problematic. As the parameters of the HS equation may vary across space, a Bayesian approach was adopted to estimate (or recalibrate) them rather than relying on the fixed values as suggested in the traditional model. The Bayesian approach allows a sound development of our model in a multi-scale temporal framework, where the ETo at daily, monthly and annual scales are jointly used to estimate the HS equation parameters. The proposed and reference models are applied and tested using meteorological data from 17 stations located across the Han river basin in South Korea. The results indicate that the traditional HS equation with fixed parameters and without recalibration tends to overestimate the reference ET for all stations. The locally recalibrated approach to the HS equation at a daily temporal scale can effectively reduce the systematic bias associated with the use of the traditional HS equation but fails to reproduce the underlying distribution of ETo at different temporal scales (e.g., monthly and annual). This leads to a large systematic bias in ETo at these scales. In contrast, the proposed multi-scale surrogate model offers a more precise estimation of the reference ET at a daily timescale as well as at the aggregated monthly and annual temporal scales. This is particularly useful to minimize the systematic bias often observed when using traditional surrogate models for the reference ET in hydrological studies such as rainfall-runoff modeling and assessment of climate change impact on water resources.

Evaluating Winter Precipitation over the Western Himalayas in a High-Resolution Indian Regional Reanalysis Using Multisource Climate Datasets

Abstract Considerable uncertainties are associated with precipitation characteristics over the western Himalayan region (WHR). These are due to typically small-scale but high-intensity storms caused by the complex topography that are under-resolved by a sparse gauge network. Additionally, both satellite and gauge precipitation measurements remain subject to systematic errors, typically resulting in underestimation over mountainous terrains. Reanalysis datasets provide prospective alternative but are limited by their resolution, which has so far been too coarse to properly resolve orographic precipitation. In this study, we evaluate and cross compare Indian Monsoon Data Assimilation and Analysis (IMDAA), the first high-resolution (12 km) regional reanalysis over India, with various precipitation products during winter season over WHR. We demonstrate IMDAA’s efficiency in representing winter precipitation characteristics at seasonal, diurnal, interannual scales, as well as heavy precipitation associated with western disturbances (WDs). IMDAA shows closer agreement to other reanalyses than to gauge-based and satellite products in error and bias analysis. Although depicting higher magnitudes, its fine resolution allows a much closer insight into localized spatial patterns and the diurnal cycle, a key advantage over other datasets. Mean winter precipitation over WHR shows a significant decreasing trend in IMDAA, despite no significant trend in the frequency of WDs tracked in either IMDAA or ERA5. The study also exhibits the potential use of IMDAA for characterizing winter atmospheric dynamics, both for climatological studies and during WD activity such as localized valley winds. Overall, these findings highlight the potential utility for IMDAA in conducting monitoring and climate change impact assessment studies over the fragile western Himalayan ecosystem.

Assessment of the benefits of climate model weights for ensemble analysis in three urban precipitation frequency studies

AbstractIn hydrology, projected climate change impact assessment studies typically rely on ensembles of downscaled climate model outputs. Due to large modeling uncertainties, the ensembles are often averaged to provide a basis for studying the effects of climate change. A key issue when analyzing averages of a climate model ensemble is whether to weight all models in the ensemble equally, often referred to as the equal‐weights or unweighted approach, or to use a weighted approach, where, in general, each model would have a different weight. Many studies have advocated for the latter, based on the assumption that models that are better at simulating the past, that is, the models with higher hindcast accuracy, will give more accurate forecasts for the future and thus should receive higher weights. To examine this issue, observed and modeled daily precipitation frequency (PF) estimates for three urban areas in the United States, namely Boston, Massachusetts; Houston, Texas; and Chicago, Illinois, were analyzed. The comparison used the raw output of 24 Coupled Model Intercomparison Project Phase 5 (CMIP5) models. The PFs from these models were compared with the observed PFs for a specific historical training period to determine model weights for each area. The unweighted and weighted averaged model PFs from a more recent testing period were then compared with their corresponding observed PFs to determine if weights improved the estimates. These comparisons indeed showed that the weighted averages were closer to the observed values than the unweighted averages in nearly all cases. The study also demonstrated how weights can help reduce model spread in future climate projections by comparing the unweighted and weighted ensemble standard deviations in these projections. In all studied scenarios, the weights actually reduced the standard deviations compared to the equal‐weights approach. Finally, an analysis of the results' sensitivity to the areal reduction factor used to allow comparisons between point station measurements and grid‐box averages is provided.

Impact of Gridded Weather Data Sources and its Temporal Resolution on Crop Evapotranspiration and Effective Rainfall of Major Crops in Eastern Region of India

Synthetic climate data are being increasingly used as key input in number of climate change impact assessment studies, hydrological assessments, and water resource management projects. The present study quantified the impact of two-gridded climate data sources [India Meteorological Department (IMD) and National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Centre (CPC)] on estimation of evapotranspiration (ETc ) and effective rainfall (ER) of paddy (Kharif) and wheat (Rabi) crops across four agroecological sub-regions, and compared the results with observed station data. The effect of daily, decadal, and monthly temporal resolutions of climatic data were evaluated on ETc and ER using climatological model CROPWAT 8.0. Results showed that compared to daily resolution, the monthly temporal resolution estimated significantly higher (1.7-4.0%) ETc for wheat crop. Effect of temporal resolution of climate data on ER of both crops was insignificant, implying that the uncertainties in estimation of ER emanating from use of daily, decadal, or monthly temporal resolutions would be within acceptable limits. Across the locations, the CPC-based estimates of paddy ETc deviated from (-)1.3% to 21.6% of the station ETc , while the deviations for wheat ETc were in the range of (-)7.1% to 9.3 per cent. Significant variations from station etc were also observed for CPC (1.7 - 19.4%) and IMD [(-)34.3 - 31.1%] based ETc estimates of wheat ETc . The study recommends that the technological and policy research outputs based on gridded climate data products should be carefully analysed keeping in view the uncertainty accruing on account of use of gridded data products.

Superensembles of raw and bias‐adjusted regional climate models for Mediterranean region, Turkey

AbstractFor regional‐scale studies on climate change and relevant impact assessment, the projections of regional climate models (RCMs) are used due to their advantage of high resolution and better representation of the local climate relative to the global climate models. However, direct use of RCM outputs is prone to uncertainties and biases that may significantly diminish the accuracy of results. EURO‐COordinated Regional Downscaling EXperiment (CORDEX) initiative that is a part of the global Coordinated Regional Downscaling Experiment Project provides high‐resolution RCM projections for the European domain and bias‐adjusted regional projections under the “CORDEX‐Adjust” Project for climate change impact assessment studies. This study aims to perform a multi‐model analysis of precipitation data using bias‐adjusted and raw/non‐bias adjusted CORDEX RCMs to obtain an evaluation of their representativeness for local climate conditions and adequacy to be used for climate change impact assessment. For this purpose, the analysis focuses on four CORDEX RCMs and their 12 bias‐adjusted versions generated with cumulative distribution function transformation, quantile mapping, and distribution‐based scaling methodologies. For the analysis in total, 16 hindcast results of raw and bias‐adjusted RCMs, and three superensembles (SEs) generated through multiple linear regression are compared for their performance regarding their goodness of fit to the ground‐based precipitation monitoring data from eight meteorological stations in the Mediterranean region in Turkey. The analysis verified that the skill of individual simulations including the bias‐adjusted outputs is significantly variable in spatial and temporal means. On the other hand, SE formed by using all 16 hindcast outputs has the highest skill for the representation of variability in precipitation in time as well as for the reproduction of annual climatology at all stations, although potential drawbacks concerning seasonality and the range of anomaly may still exist which might be significant depending on the specific aim of impact assessment.

Development of Integrated Land Simulator

Accurate simulations of land processes are crucial for many purposes, such as climate simulation, weather, flood, and drought prediction, and climate change impact assessment studies. In this paper, we present a new land simulator called the Integrated Land Simulator (ILS). The ILS consists of multiple models that represent processes related to land (hereafter, referred to as “land models”). They are coupled by a general-purpose coupler, Jcup, and executed using the Multiple Program Multiple Data approach. Currently, ILS includes a physical land surface model, the Minimal Advanced Treatments of Surface Interaction and Runoff model, and a hydrodynamic model, the Catchment-based Macro-scale Floodplain model, and the inclusion of additional land models is planned. We conducted several test simulations to evaluate the computational speed and scalability and the basic physical performance of the ILS. The results will become a benchmark for further development.

Correcting lateral boundary biases in regional climate modelling: the effect of the relaxation zone

Regional climate models (RCM) are an important tool for simulating atmospheric information at finer resolutions often of greater relevance to local scale climate change impact assessment studies. The lateral and lower boundary conditions, which form the inputs to the RCM downscaling application, are outputs from the global climate model (GCM). These boundary variables are known to be biased in GCMs, providing the potential to use a statistical approach that corrects these biases before use in downscaling. An array of bias correction techniques have been developed to remove these biases before being used to drive the RCM, but questions remain on their efficacy in terms of the final downscaled output. This study assesses the impact of these bias correction strategies by focussing on how these corrections are translated as one proceeds from the lateral boundaries into the model interior. Of specific interest is the change in the correction from generation of the lateral boundary conditions as well as how correction information moves through the relaxation zone and into the interior of the model. Here we show that bias correction information passing into the regional climate model is limited by interpolations required to generate lateral boundary conditions and dominant outflow wind conditions in the boundaries. This work suggests that these limitations should be addressed in order for bias correction of lateral boundary conditions to robustly influence RCM simulations of climate in the interior of the model domain.

Streamflow response to projected climate changes in the Northwestern Upper Indus Basin based on regional climate model (RegCM4.3) simulation

In the present research, we used bias corrected output of regional climate model (RegCM4.3) to generate streamflow for future climate change projections in the Northwestern Upper Indus Basin (NUIB) under RCP4.5 and RCP8.5 scenarios. The average annual runoff was found to be continuously increasing by the end of 21st century. The increase in precipitation as well as the streamflow for summer and winter in the NUIB indicated that RegCM4.3 simulated climate change projections were realistically transferred to HEC-HMS hydrological model. Maximum streamflow peaks, both in frequency and magnitude, were observed in July and August over three streamflow gauges (i.e. Besham, Nowshera and Khairabad) posing possible threats of flood risks during the monsoon season in Pakistan. Significant streamflow changes were projected for winter and spring seasons. The analysis of flow duration curve for three hydrometric stations indicates that the flow occurrence probability could be greater under both scenarios relative to the reference period. The increase in streamflow was projected under high (Q5) and medium (Q50) flow, which could possibly be related with the increase in monsoon precipitation events in the NUIB. Overall it is to be expected that the NUIB will confront more floods as a result of projected increase in medium and high flow, which demands better policies and management towards the water resources of the NUIB. This study is expected to contribute considerably to climate change impact assessment studies over the NUIB, Pakistan.

Climate model bias correction for nonstationary conditions

In most of the climate change impact assessment studies, climate model bias is considered to be stationary between the control and scenario periods. Few methods are found in the literature that addresses the issue of nonstationarity in correcting the bias. To overcome the shortcomings reported in these approaches, three new methods of bias correction (NBC_μ, NBC_σ, and NBC_bs) are presented. The methods are improvised versions of previous techniques relying on distribution mapping. The methods are tested using split sample approach over 50-year historical period for nine climate stations in Ontario, using six regional climate models. The average bias reduction improvement by new methods, in mean daily and monthly precipitation, was found to be 73.9%, 74.3%, and 77.4%, respectively, higher than that obtained by the previous methods (eQM 67.7% and CNCDFm_NP 64.1%). Thus, the methods are found to be more effective in accounting for nonstationarity in the model bias.

A novel approach for selecting extreme climate change scenarios for climate change impact studies

One of the main challenges in climate change impact assessment studies is selecting climate change scenarios. By focusing on selecting projected extremes in a high dimensional space, one is confronted with the shrinkage of ensemble size while preserving the projection spread. This study proposes a novel integrated computational geometry algorithm to select extreme climate change scenarios in a high dimensional space. A set of 12 prominent climate extremes indices were used (as input to the algorithm) out of the 27 core indices recommended by the World Meteorological Organization's Expert Team on Climate Change Detection and Indices (ETCCDI). The ETCCDI indices were projected by Coupled Model Intercomparison Project Phase 5 (CMIP5) for the period of 2081–2100 relative to the baseline period 1986–2005. The approach enables the user to shrink the initial selected ensemble into smaller sub-ensembles while still capturing a wide range of simulated changes for selected climatological variables. The conservation of the projection spread was evaluated using a robust validation method when the spread error was calculated for each simulation run. The developed algorithm was applied to three different regions where the geographical domain was narrowed-down from sub-continental (western North America) to its regional (Alberta, Canada), and local (Athabasca River basin, Alberta, Canada) subdomains. Results revealed that selected extreme scenarios can vary from one region to another within the same geographical domain in response to the spatial variation in climatic regime.

Factorial inferential grid grouping and representativeness analysis for a systematic selection of representative grids

AbstractA factorial inferential grid grouping and representativeness analysis (FIGGRA) approach is developed to achieve a systematic selection of representative grids in large‐scale climate change impact assessment and adaptation (LSCCIAA) studies and other fields of Earth and space sciences. FIGGRA is applied to representative‐grid selection for temperature (Tas) and precipitation (Pr) over the Loess Plateau (LP) to verify methodological effectiveness. FIGGRA is effective at and outperforms existing grid‐selection approaches (e.g., self‐organizing maps) in multiple aspects such as clustering similar grids, differentiating dissimilar grids, and identifying representative grids for both Tas and Pr over LP. In comparison with Pr, the lower spatial heterogeneity and higher spatial discontinuity of Tas over LP lead to higher within‐group similarity, lower between‐group dissimilarity, lower grid grouping effectiveness, and higher grid representativeness; the lower interannual variability of the spatial distributions of Tas results in lower impacts of the interannual variability on the effectiveness of FIGGRA. For LP, the spatial climatic heterogeneity is the highest in January for Pr and in October for Tas; it decreases from spring, autumn, summer to winter for Tas and from summer, spring, autumn to winter for Pr. Two parameters, i.e., the statistical significance level (α) and the minimum number of grids in every climate zone (Nmin), and their joint effects are significant for the effectiveness of FIGGRA; normalization of a nonnormal climate‐variable distribution is helpful for the effectiveness only for Pr. For FIGGRA‐based LSCCIAA studies, a low value of Nmin is recommended for both Pr and Tas, and a high and medium value of α for Pr and Tas, respectively.

Challenges to link climate change data provision and user needs: Perspective from the COST‐action VALUE

The application of climate change impact assessment (CCIA) studies in general and especially the linkages between different actor groups typically involved is often not trivial and subject to many limitations and uncertainties. Disciplinary issues like competing downscaling approaches, imperfect climate and impact model data and uncertainty propagation as well as the selection of appropriate data sets are only one part of the story. Interdisciplinary and transdisciplinary challenges add to these, as climate data and impact model data provision and their usage require at least a minimum of common work and shared understanding among actors. Here, we provide the VALUE perspective on current disciplinary challenges and limitations at the downscaling interface and elaborate transdisciplinary issues that hamper a proper working downscaling interface. The perspective is partly based on a survey on user needs of downscaled data that was distributed among 62 participants across Europe involving 22 sectors. Partly, it is based on the exchanges and experiences gained during the lifetime of VALUE that brought together different actor groups of different disciplines: climate modellers, impact modellers, statisticians and stakeholders. We outline a sketch that summarizes the linkages between the main identified actor groups: climate model data providers, impact modellers and societal users. We summarize review and structure current actors groups, needs and issues. We argue that this structuring enables involved actors to tackle these issues in a more organized and hence effective way. A key solution to several difficulties at the downscaling interface is to our understanding the development of guidelines based on benchmark tests like the VALUE framework. In addition, fostering communication between actor groups—and financing this communication—is essential to obtain the best possible CCIA as a prerequisite for robust adaptation.

Evaluation of ERA-Interim, MERRA, NCEP-DOE R2 and CFSR Reanalysis precipitation Data using Gauge Observation over Ethiopia for a period of 33 years

The vital demand of reliable climatic and hydrologic data of fine spatial and temporal resolution triggered the employment of reanalysis datasets as a surrogate in most of the hydrological modelling exercises. This study examines the performance of four widely used reanalysis datasets: ERA-Interim, NCEP-DOE R2, MERRA and CFSR, in reproducing the spatio-temporal characteristics of observed daily precipitation of different stations spread across Ethiopia, East Africa. The appropriateness of relying on reanalysis datasets for hydrologic modelling, climate change impact assessment and regional modelling studies is assessed using various statistical and non-parametric techniques. ERA-Interim is found to exhibit higher correlation and least root mean square error values with observed daily rainfall, which is followed by CFSR and MERRA in most of the stations. The variability of daily precipitation is better captured by ERA, CFSR and MERRA, while NCEP-DOE R2 overestimated the spread of the precipitation data. While ERA overestimates the probability of moderate rainfall, it is seemingly better in capturing the probability of low rainfall. CFSR captures the overall distribution reasonable well. NCEP-DOE R2 appears to be outperforming others in capturing the probabilities of higher magnitude rainfall. Climatological seasonal cycle and the characteristics of wet and dry spells are compared further, where ERA seemingly replicates the pattern more effectively. However, observed rainfall exhibits higher frequency of short wet spells when compared to that of any reanalysis datasets. MERRA relatively underperforms in simulating the wet spell characteristics of observed daily rainfall. CFSR overestimates the mean wet spell length and mean dry spell length. Spatial trend analysis indicates that the northern and central western Ethiopia show increasing trends, whereas the Central and Eastern Ethiopia as well as the Southern Ethiopia stations show either no trend or decreasing trend. Overall, ERA-Interim and CFSR are better in depicting various characteristics of daily rainfall in Ethiopian region.

Climate and management interaction cause diverse crop phenology trends

Growing evidence suggests that the warming trend observed in many parts of the world has considerably modified crop phenology during the last decades but little is known about the impact of changes in crop management on crop phenology and possible interactions with temperature increase, and whether responses can be generalized across crop types. Here we evaluate the effects of climate and management on crop phenology by using observations for winter rapeseed and winter rye obtained in Germany for the period 1960–2013 by using piecewise linear regressions of temperature and phenology data on year. We show that long-term trends in crop phenology are crop-specific. The length of the vegetative phase of winter rapeseed declined by 4.8days per decade in the period 1979–2013. However, the corresponding decline for winter rye was only 1.3days per decade in the period 1978–2013 with the difference caused by change in management practices such as the introduction of early flowering cultivars of winter rapeseed or changes in sowing date of winter rapeseed and winter rye during the last decades in Germany. The length of the reproductive phase of winter rye declined by 0.9days per decade between 1976 and 2013 in response to the warming trend in that period. In contrast, the extended use of late maturing cultivars with a longer grain filling period and changed planting densities over-compensated for the effect of increasing temperature on the length of the reproductive phase of winter rapeseed and caused an increasing trend of 2.0days per decade between 1992 and 2013. The sowing date of winter rye advanced by 1.3days per decade in the period 1972–2013. The length of the phase between maturity and harvest increased considerably for both crops and compensated partly for the effect of increasing temperature to shorten the preceding phenological phases. We conclude that it is essential to account for interactions between climate and crop management in climate change impact analysis and assessment studies and that differences among crops need to be considered.

Multisite bias correction of precipitation data from regional climate models

ABSTRACTThe characteristics of precipitation in regional climate model simulations deviate considerably from those of the observed data; therefore, bias correction is a standard part of most climate change impact assessment studies. The standard approach is that the corrections are calibrated and applied separately for individual spatial points and meteorological variables. For this reason, the correlation and covariance structures of the observed and corrected data differ, although the individual observed and corrected data sets correspond well in their statistical indicators. This inconsistency may affect impact studies using corrected simulations.This study presents a new approach to the bias correction utilizing principal components in combination with quantile mapping, which allows for the correction of multivariate data sets. The proposed procedure significantly reduces the bias in covariance and correlation structures, as well as that in the distribution of individual variables. This is in contrast to standard quantile mapping, which only corrects the individual distributions, and leaves the dependence structure biased.

Projection of climate change impacts on precipitation using soft-computing techniques: A case study in Zayandeh-rud Basin, Iran

Due to the complexity of climate-related processes, accurate projection of the future behavior of hydro-climate variables is one of the main challenges in climate change impact assessment studies. In regression-based statistical downscaling processes, there are different sources of uncertainty arising from high-dimensionality of atmospheric predictors, nonlinearity of empirical and quantitative models, and the biases exist in climate model simulations. To reduce the influence of these sources of uncertainty, the current study presents a comprehensive methodology to improve projection of precipitation in the Zayandeh-Rud basin in Iran as an illustrative study. To reduce dimensionality of atmospheric predictors and capture nonlinearity between the target variable and predictors in each station, a supervised-PCA method is combined with two soft-computing machine-learning methods, Support Vector Regression (SVR) and Relevance Vector Machine (RVM). Three statistical transformation methods are also employed to correct biases in atmospheric large-scale predictors. The developed models are then employed on outputs of the Coupled Model Intercomparison Project Phase 5 (CMIP5) multimodal dataset to project future behavior of precipitation under three climate changes scenarios. The results indicate reduction of precipitation in the majority of the sites in this basin threatening the availability of surface water resources in future decades.

Correcting for systematic biases in GCM simulations in the frequency domain

Summary Bias correction is considered as a critical post-processing step to remove systematic errors and improve the quality of General Circulation Model (GCM) simulations before their use in climate change impact assessment applications. A majority of the bias correction approaches correct for biases either at a single time scale or at multiple pre-specified time scales. An inappropriate or insufficient selection of time scales may lead to improper or sub-optimal bias corrected outputs, especially when persistence attributes across a range of scales are of interest. In this paper, we present a new bias correction approach that works in the frequency space and is independent of specific time scales. The approach is named as frequency-based bias correction (FBC). The usefulness of the approach is demonstrated by applying it to the monthly rainfall simulations of MIROC5 GCM over Australia and comparing the results with two other approaches, namely, empirical quantile mapping and recursive nesting bias correction, in cross validation. The comparison is based on the reproduction of various observed distribution and persistence attributes. Cross-validation results indicate that the proposed approach shows similar performance in terms of reproducing the first- and second-order moments of observed precipitation time series, however, outperforms with regard to persistence attributes. The approach shows high potential for use in downscaling and other climate change impact assessment studies, especially for the planning and design of hydrologic systems that are sensitive to the characterisation of persistence in the hydrologic time series.

Water resources climate change projections using supervised nonlinear and multivariate soft computing techniques

Summary Accurate projection of global warming on the probabilistic behavior of hydro-climate variables is one of the main challenges in climate change impact assessment studies. Due to the complexity of climate-associated processes, different sources of uncertainty influence the projected behavior of hydro-climate variables in regression-based statistical downscaling procedures. The current study presents a comprehensive methodology to improve the predictive power of the procedure to provide improved projections. It does this by minimizing the uncertainty sources arising from the high-dimensionality of atmospheric predictors, the complex and nonlinear relationships between hydro-climate predictands and atmospheric predictors, as well as the biases that exist in climate model simulations. To address the impact of the high dimensional feature spaces, a supervised nonlinear dimensionality reduction algorithm is presented that is able to capture the nonlinear variability among projectors through extracting a sequence of principal components that have maximal dependency with the target hydro-climate variables. Two soft-computing nonlinear machine-learning methods, Support Vector Regression (SVR) and Relevance Vector Machine (RVM), are engaged to capture the nonlinear relationships between predictand and atmospheric predictors. To correct the spatial and temporal biases over multiple time scales in the GCM predictands, the Multivariate Recursive Nesting Bias Correction (MRNBC) approach is used. The results demonstrate that this combined approach significantly improves the downscaling procedure in terms of precipitation projection.

Is robustness really robust? How different definitions of robustness impact decision-making under climate change

Robust decision-making is being increasingly used to support environmental resources decisions and policy analysis under changing climate and society. In this context, a robust decision is a decision that is as much as possible insensitive to a large degree of uncertainty and ensures certain performance across multiple plausible futures. Yet, the concept of robustness is neither unique nor static. Multiple robustness metrics, such as maximin, optimism-pessimism, max regret, have been proposed in the literature, reflecting diverse optimistic/pessimistic attitudes by the decision maker. Further, these attitudes can evolve in time as a response to sequences of favorable (or adverse) events, inducing possible dynamic changes in the robustness metrics. In this paper, we explore the impact of alternative definitions of robustness and their evolution in time for a case of water resources system management under changing climate. We study the decisions of the Lake Como operator, who is called to regulate the lake by balancing irrigation supply and flood control, under an ensemble of climate change scenarios. Results show a considerable variability in the system performance across multiple robustness metrics. In fact, the mis-definition of the actual decision maker’s attitude biases the simulation of its future decisions and produces a general underestimation of the system performance. The analysis of the dynamic evolution of the decision maker’s preferences further confirms the potentially strong impact of changing robustness definition on the decision-making outcomes. Climate change impact assessment studies should therefore include the definition of robustness among the uncertain parameters of the problem in order to analyze future human decisions under uncertainty.