Abstract The paper appraises two approaches for the treatment of heteroscedasticity and autocorrelation in residual errors hydrological models. Both use weighted least squares (WLS), with modeled as a linear function predicted flows represented using an AR(1) process. In first approach, parameters are inferred jointly model parameters. second approach is two‐stage “postprocessor” scheme, where Stage 1 infers ignoring 2 conditionally These compared to WLS scheme that ignores autocorrelation....

Residual errors of hydrological models are usually both heteroscedastic and autocorrelated. However, only a few studies have attempted to explicitly include these two statistical properties into the residual error model jointly infer them with parameters. This technical note shows that applying autoregressive raw residuals, as done in some recent studies, can lead unstable poor predictive performance. instability be avoided by process standardized residuals. The theoretical analysis is...

Abstract. Many multi-site stochastic models have been proposed for the generation of daily precipitation, but they generally focus on reproduction low to high precipitation amounts at stations concerned. This paper proposes significant extensions model introduced by Wilks, with aim reproducing statistical features extremely rare events (in terms frequency and magnitude) different temporal spatial scales. In particular, first extended version integrates heavy-tailed distributions, tail...

Abstract. In this study, we analyze how precipitation, antecedent conditions, and their spatial patterns interactions lead to extreme floods in a large catchment. The analysis is based on 10 000 years of continuous simulations from hydro-meteorological modelling chain for catchment, the Aare River basin, Switzerland. To account different flood-generating processes, our work with hourly time resolution. consisted stochastic weather generator (GWEX), bucket-type hydrological model (HBV),...

The quantification of uncertainty sources in ensembles climate projections obtained from combinations different scenarios and impact models is a key issue studies. small size the simulation chains their incomplete sampling scenario model makes analysis difficult. In popular single-time ANOVA approach for instance, precise estimate internal variability requires multiple members each chain (e.g., emission scenario–climate combination), but are typically available few only. most also, partition...

Abstract The development of renewable electricity in Africa could be massive coming decades, as a response to the rapid rising demand while complying with Paris Agreements. This study shows that high-resolution climate experiments CORDEX-AFRICA, annual mean solar potential is expected decrease on average by 4% over most continent end century, reaching up 6% Horn Africa, direct result radiation and increase air surface temperature. These projections are associated large uncertainties,...

Abstract Over the past decades, large variations of precipitation were observed in Africa, which often led to dramatic consequences for local society and economy. To avoid such disasters future, it is crucial better anticipate expected changes, especially current context climate change population growth. this date, however, projections over Africa are still associated with very uncertainties. understand how uncertainty can be reduced, study uses an advanced Bayesian analysis variance (ANOVA)...

Abstract. In a context of climate change, trends in extreme snow loads need to be determined minimize the risk structure collapse. We study 50-year return levels ground load (GSL) using non-stationary value models. These are assessed at mountain massif scale from GSL data, provided for French Alps 1959 2019 by meteorological reanalysis and snowpack model. Our results indicate temporal decrease 900 4200 m, significant northwest up 2100 m. detect most important m with an average −30 % between...

Abstract. Large multiscenario multimodel ensembles (MMEs) of regional climate model (RCM) experiments driven by global models (GCMs) are made available worldwide and aim at providing robust estimates changes associated uncertainties. Due to many missing combinations emission scenarios leading sparse scenario–GCM–RCM matrices, these large ensembles, however, very unbalanced, which makes uncertainty analyses impossible with standard approaches. In this paper, the assessment is carried out...

The Mediterranean region is a climate change hotspot for water resources. However, uncertainty analyses of hydrological projections are rarely quantified. In this study, an in-depth analysis and uncertainties high low flows performed. Climatic derived from recent downscaling method were used, two representative concentration pathway scenarios (RCPs), five general circulation model/regional model (GCM/RCM) couples, three models (HMs), 29 calibration schemes. A quasi-ergodic variance was used...

Abstract. Hydrological modelling of small mountainous catchments is particularly challenging because the high spatio-temporal resolution required for meteorological forcings. In situ measurements precipitation are typically scarce in these remote areas, at elevations. Precipitation reanalyses propose different alternative forcings simulation streamflow using hydrological models. this paper, we evaluate performances two models representing some key processes (< 300 km2), products with a...

Robust adaptation to hydrological climate change requires an assessment of possible and futures at different scales. Explore2 is a recent multimodel ensemble projections developed accompany local plans for metropolitan France Corsica, 550,000 km2 wide territory with large diversity regimes (Sauquet et al. 2024). The provides transient daily river flow more than 4’000 locations on French rivers. Projections have been obtained three RCP emission scenarios 4 9 models (HMs) driven by...

The concept of hydroclimate services is predominantly recognised as web portals dedicated to the dissemination data potential users. However, scope climate extends beyond sole provision data. This communication presents a comprehensive ecosystem tools and resources associated with development an updated national hydrological projection dataset in France. was brought life through close collaboration between scientists water managers two joint projects: Explore2 LIFE Eau&Climat. Tools...

Whether mixtures of distributions are employed as a flexible modeling device to estimate densities or used model data thought arise from several populations, they provide an efficient tool approximate distribution. Indeed, can multiple modes, different types skewness, etc., but also be classify observations heterogeneous sets. In this paper, we study with normal, gamma, and Gumbel components. Moving away the standard normal setting, gamma developed in order strictly positive hydrological for...

Abstract. Climate change projections indicate that extreme snowfall is expected to increase in cold areas, i.e., at high latitudes and/or elevation, and decrease warmer mid-latitudes low elevation. However, the magnitude of these contrasting patterns their precise relations elevation scale a given mountain range remain poorly known. This study analyzes annual maxima daily based on SAFRAN reanalysis spanning time period 1959–2019 provided within 23 massifs French Alps every 300 m We estimate...

A classical way to model rainfall is use a Poisson process. Authors generally employed cluster of rectangular pulses reproduce the hierarchical structure storms. Although independence between cell intensity and duration turned out be nonrealistic assumption, only few models link these variables. In this paper, Neyman‐Scott process considering dependence depth developed. We introduce with cubic copula. Copulas are multivariate distributions modeling variables, preserving marginal...

Abstract. Very large wildfires have high human, economic, and ecological impacts so that robust evaluation of their return period is crucial. Preventing such events a major objective the new fire policy set up in France 1994, which oriented towards fast massive suppression. Whereas this probably efficient for reducing mean burned area (BA), its effect on largest fires still unknown. In study, we make use statistical extreme value theory (EVT) to compute periods very BAs southern France, two...

Abstract IPCC reports and climate change impact studies generally exploit ensembles of projections based on different socio-economic pathways models, which provide the temporal evolution plausible future climates. However, The Paris Agreement many national international commitments consider adaptation mitigation plans targeting global warming levels. Model uncertainty scenario typically affect both crossing-time levels features at a given level. In this study, we assess uncertainties in...

Abstract. Seasonal precipitation estimation in ungauged mountainous areas is essential for understanding and modeling a physical variable of interest many environmental applications (hydrology, ecology, cryospheric studies). Precipitation lapse rates (PLRs), defined as the increasing or decreasing rate amounts with elevation, play decisive role high-altitude estimation. However, documentation PLR regions remains weak even though their utilization frequent. This article intends to assess...

Abstract. Forecasting the height of new snow (HN) is crucial for avalanche hazard forecasting, road viability, ski resort management and tourism attractiveness. Météo-France operates PEARP-S2M probabilistic forecasting system, including 35 members PEARP Numerical Weather Prediction where SAFRAN downscaling tool refines elevation resolution Crocus snowpack model represents main physical processes in snowpack. It provides better HN forecasts than direct NWP diagnostics but exhibits significant...