Reply on RC2

Hydrology and Earth System Sciences

Climate change impact studies require a reference climatological dataset providing a baseline period to assess future changes and post-process climate model biases. High-resolution gridded precipitation and temperature datasets interpolated from weather stations are available in regions of high-density networks of weather stations, as is the case in most parts of Europe and the United States. In many of the world's regions, however, the low density of observational networks renders gauge-based datasets highly uncertain. Satellite, reanalysis and merged product datasets have been used to overcome this deficiency. However, it is not known how much uncertainty the choice of a reference dataset may bring to impact studies. To tackle this issue, this study compares nine precipitation and two temperature datasets over 1145 African catchments to evaluate the dataset uncertainty contribution to the results of climate change studies. These deterministic datasets all cover a common 30-year period needed to define the reference period climate. The precipitation datasets include two gauge-only products (GPCC and CPC Unified), two satellite products (CHIRPS and PERSIANN-CDR) corrected using groundbased observations, four reanalysis products (JRA55, NCEP-CFSR, ERA-I and ERA5) and one merged gauged, satellite and reanalysis product (MSWEP). The temperature datasets include one gauged-only (CPC Unified) product and one reanalysis (ERA5) product. All combinations of these precipitation and temperature datasets were used to assess changes in future streamflows. To assess dataset uncertainty against that of other sources of uncertainty, the climate change impact study used a topdown hydroclimatic modeling chain using 10 CMIP5 (fifth Coupled Model Intercomparison Project) general circulation models (GCMs) under RCP8.5 and two lumped hydrological models (HMETS and GR4J) to generate future streamflows over the 2071-2100 period. Variance decomposition was performed to compare how much the different uncertainty sources contribute to actual uncertainty. Results show that all precipitation and temperature datasets provide good streamflow simulations over the reference period, but four precipitation datasets outperformed the others for most catchments. They are, in order, MSWEP, CHIRPS, PERSIANN and ERA5. For the present study, the two-member ensemble of temperature datasets provided negligible levels of uncertainty. However, the ensemble of nine precipitation datasets provided uncertainty that was equal to or larger than that related to GCMs for most of the streamflow metrics and over most of the catchments. A selection of the four best-performing reference datasets (credibility ensemble) significantly reduced the uncertainty attributed to precipitation for most metrics but still remained the main source of uncertainty for some streamflow metrics. The choice of a reference dataset can therefore be critical to climate change impact studies as apparently small differences between datasets over a common reference period can propagate to generate large amounts of uncertainty in future climate streamflows.

International Journal of Climatology, 2016

Gridded precipitation data sets are frequently used to evaluate climate models or to remove model output biases. Although precipitation data are error prone due to the high spatio-temporal variability of precipitation and due to considerable measurement errors, relatively few attempts have been made to account for observational uncertainty in model evaluation or in bias correction studies. In this study, we compare three types of European daily data sets featuring two Pan-European data sets and a set that combines eight very high-resolution station-based regional data sets. Furthermore, we investigate seven widely used, larger scale global data sets. Our results demonstrate that the differences between these data sets have the same magnitude as precipitation errors found in regional climate models. Therefore, including observational uncertainties is essential for climate studies, climate model evaluation, and statistical post-processing. Following our results, we suggest the following guidelines for regional precipitation assessments. (1) Include multiple observational data sets from different sources (e.g. station, satellite, reanalysis based) to estimate observational uncertainties. (2) Use data sets with high station densities to minimize the effect of precipitation undersampling (may induce about 60% error in data sparse regions). The information content of a gridded data set is mainly related to its underlying station density and not to its grid spacing. (3) Consider undercatch errors of up to 80% in high latitudes and mountainous regions. (4) Analyses of small-scale features and extremes are especially uncertain in gridded data sets. For higher confidence, use climate-mean and larger scale statistics. In conclusion, neglecting observational uncertainties potentially misguides climate model development and can severely affect the results of climate change impact assessments.

Regional climate model (RCM) based scenarios are fundamental tools for the estimation of climate change effects on the environment, ecosystems and society. Our aim was to create RCM-based climate scenarios that contain daily meteorological data (precipitation, maximum and minimum temperature) to support climate change related impact studies. We utilized the results of ten RCM experiments that were produced and made accessible within the framework of the ENSEMBLES FP6 project. We performed bias correction using the cumulative distribution functions fitting technique, which allowed for correcting the systematic errors in the RCM results. In case of precipitation both the intensity and the frequency of precipitation was corrected. The resulting database -the so called FORESEE database -contains daily meteorological data based on the ten RCM results for 2010-2100, and observation based data for the period 1951-2009 interpolated to 1/6×1/6 degree spatial resolution grid. Central Europe is the target area of the FORESEE database.

Geoscience Data Journal, 2015

Europe: the region where knowledge of possible climate change effects is inadequate. A questionnaire-based survey was used to specify database structure and content. FORESEE contains the seamless combination of gridded daily observation-based data (1951-2013) built on the E-OBS and CRU TS datasets, and a collection of climate projections (2014-2100). The future climate is represented by bias-corrected meteorological data from 10 regional climate models (RCMs), driven by the A1B emission scenario. These latter data were developed within the frame of the ENSEMBLES FP6 project. Although FORESEE only covers a limited area of Central and Eastern Europe, the methodology of database development, the applied bias correction techniques, and the data dissemination method, can serve as a blueprint for similar initiatives.

Water Resources Management, 2020

Uncertainty in climate change impact studies can be identified in general circulation models (GCM), emission scenarios, downscaling techniques, hydrological models and data. Despite being aleatory or epistemic-sourced uncertainties, their effect size on impact model output can be variable in different climate zones. To test this hypothesis, the present study explored the uncertainty level of each source in the projected streamflow scenarios of the Sarbaz river basin (SRB) with arid climate properties and the tropical Hulu Langat river basin (HLB). An ensemble of five GCMs bias-corrected using EquiDistant-CDF-matching method, two representative concentration pathways (RCPs), and hydrological model parameters sets sourced the overall uncertainty in each case study. Investigations were performed at mean monthly scale of three 30-year periods of baseline, 2016-2045 (2030s), and 2046-2075 (2060s). In both climates, on average, GCM uncertainty was the largest contributor at monthly scale analysis and its effect size escalated following the monsoon months. In the tropical HLB, GCM uncertainty increased across the periods, but it did not show a similar pattern in the arid region of SRB. The RCP uncertainty showed the least effect size during the baseline period and it peaked in 2030s in the HLB climate. However, the only pattern recognizable in the arid SRB was the intensified effect size of all uncertainty sources along with the tremendous impacts projected in monsoon months; at least twice the size of uncertainty sources effect and impacts projected for the tropical HLB. The uncertainty sources effect size altered dramatically as the climate-of the study area-changed. Thus, this research emboldens the need for seasonal-based analysis of uncertainty sources in climate change impact studies at dry climate zones.

International Journal of Climatology, 2017

Precipitation projections are typically obtained from general circulation model (GCM) outputs under different future scenarios, then downscaled for hydrological applications to a watershed or site-specific scale. However, uncertainties in projections are known to be present and need to be quantified. Although GCMs are commonly considered the major contributor of uncertainty for hydrological impact assessment of climate change, other uncertainty sources must be taken into account for a thorough understanding of the hydrological impact. This study investigates uncertainties related to GCMs, GCM initial conditions and representative concentration pathways (RCPs) and their sensitivity to the selection of GCM runs in order to quantify the impact of climate change on extreme precipitation and intensity/duration/frequency statistics. The results from a large ensemble of 140 CMIP5 GCM runs including 15 GCMs, 3-10 GCM initial conditions and 4 RCPs are analysed. Albeit the choice of GCM is the major contributor (up to 65% for some cases) to intense precipitation change uncertainty for all return periods (1 year, 10 years) and aggregation levels (1-, 5-, 10-, 15-and 30-day), uncertainties related to the GCM initial conditions and RCPs of up to 38 and 23%, respectively, are found in some cases. The sensitivity analysis reveals that the GCM, RCP and GCM initial condition uncertainties are greatly influenced by the set of climate model runs considered, especially for more extreme precipitation at finer time scales.

Climate Dynamics, 2009

An ensemble of regional climate modelling simulations from the European framework project PRU-DENCE are compared across European sub-regions with observed daily precipitation from the European Climate Assessment dataset by characterising precipitation in terms of probability density functions (PDFs). Models that robustly describe the observations for the control period in given regions as well as across regions are identified, based on the overlap of normalised PDFs, and then validated, using a method based on bootstrapping with replacement. We also compare the difference between the scenario period (2071-2100) and the control period precipitation using all available models. By using a metric quantifying the deviation over the entire PDF, we find a clearly marked increase in the contribution to the total precipitation from the more intensive events and a clearly marked decrease for days with light precipitation in the scenario period. This change is tested to be robust and found in all models and in all sub-regions. We find a detectable increase that scales with increased warming, making the increase in the PDF difference a relative indicator of climate change level. Furthermore, the crossover point separating decreasing from increasing contributions to the normalised precipitation spectrum when climate changes does not show any significant change which is in accordance with expectations assuming a simple analytical fit to the precipitation spectrum.

Tellus A: Dynamic Meteorology and Oceanography

Climate model projections are used to investigate the potential impacts of climate change on future weather, agriculture, water resources, human health, the global economy, etc. However, climate projections have a broad range of associated uncertainties, and it is a challenge to take account of these uncertainties in impact studies and risk assessments. Knowing which uncertainties matter and which may be reduced via scientific research or political decisions can help policy-makers in making informed decisions, scientists in focusing their resources, and businesses in building resilience to uncertainties that cannot be avoided. On the global scale, the present political resistance or ability to move from agreements to significant action provides the largest uncertainty in climate projections, followed by the uncertainty associated with climate modelling itself. Here, we show that climate sensitivity is a very important source of model uncertainty over large parts of the globe not only for temperature, but also for precipitation and wind projections. Because 'climate sensitivity' is a collective term that encompasses a wide range of feedback mechanisms in the climate system, we may not know for a long time whether models with high or low climate sensitivities are more relevant for the twenty-first century projections. Nevertheless, investigations of climate impacts cannot wait. Here we argue that it is physically and statistically unsound to mix climate model with high and low climate sensitivities, and that the subset chosen for any impact study should depend on the question one is trying to answer.

International Journal of Climatology, 2018

Climate model response (M) and greenhouse gas emissions (S) uncertainties are consistently estimated as spreads of multi-model and multi-scenario climate change projections. In comparison, there has been less agreement in estimating internal climate variability (V). Recently, an initial condition ensemble (ICE) of a climate model has been developed to study V. This ICE is simulated by running a climate model using an identical climate forcing but different initial conditions. Inter-member differences of an initial condition ensemble manifestly represent V. However, ICE has been barely used to investigate relative importance of climate change uncertainties. Accordingly, this study proposes a method of using ICEs, without assuming V as constant, for investigating the relative importance of climate change uncertainties and its temporalspatial variation. Prior to investigating temporal-spatial variation in China, V estimated using ICE was compared to that using multi-model individual time series at national scale. Results show that V using ICE is qualitatively similar to that using multi-model individual time series for temperature. However, V is not constant for average and extreme precipitations. V and M dominate before 2050s especially for precipitation, while S is dominant in the late 21st century especially for temperature. Mean temperature change is projected to be 30%-70% greater than its uncertainty until 2050s, while uncertainty becomes 10%-40% greater than the change in the late 21st century. Precipitation change uncertainty overwhelms its change by 70%-150% throughout 21st century. Cold regions (e.g. northern China, Qinghai-Tibetan Plateau) tend to have greater projected temperature change uncertainties. In dry regions (e.g. northwest China), all three uncertainties tend to be great for changes in average and extreme precipitations. Overall, this study emphasizes the importance of considering climate change uncertainty in impact studies, especially taking into account that V is irreducible in the future. Using ICEs without assumption of constant V is an appropriate approach to study climate change uncertainty.