Flood risk models capture a variety of processes and are associated with large uncertainties. In this paper, the uncertainties due to alternative model assumptions are analysed for various components of a probabilistic flood risk model in the study area of Vorarlberg (Austria). The effect of different model assumptions for five aspects is compared to a reference simulation. This includes: (I, II) the selection of two model thresholds controlling the generation of large sets of possible flood events; (III) the selection of a distribution function for the flood frequency analysis; (IV) the building representation and water level derivation for the exposure analysis and (V) the selection of an appropriate damage function. The analysis shows that each of the tested aspects has the potential to alter the modelling results considerably. The results range from a factor of 1.2 to 3, from the lowest to highest value, whereby the selection of the damage function has the largest effect on the overall modelling results.

Design flood estimation is an essential part of flood risk assessment. Commonly applied are flood frequency analyses and design storm approaches, while the derived flood frequency using continuous simulation has been getting more attention recently. In this study, a continuous hydrological modelling approach on an hourly time scale, driven by a multi-site weather generator in combination with a k-nearest neighbour resampling procedure, based on the method of fragments, is applied. The derived 100-year flood estimates in 16 catchments in Vorarlberg (Austria) are compared to (a) the flood frequency analysis based on observed discharges, and (b) a design storm approach. Besides the peak flows, the corresponding runoff volumes are analysed. The spatial dependence structure of the synthetically generated flood peaks is validated against observations. It can be demonstrated that the continuous modelling approach can achieve plausible results and shows a large variability in runoff volume across the flood events.

This article presents a flood risk analysis model that considers the spatially heterogeneous nature of flood events. The basic concept of this approach is to generate a large sample of flood events that can be regarded as temporal extrapolation of flood events. These are combined with cumulative flood impact indicators, such as building damages, to finally derive time series of damages for risk estimation. Therefore, a multivariate modeling procedure that is able to take into account the spatial characteristics of flooding, the regionalization method top-kriging, and three different impact indicators are combined in a model chain. Eventually, the expected annual flood impact (e.g., expected annual damages) and the flood impact associated with a low probability of occurrence are determined for a study area. The risk model has the potential to augment the understanding of flood risk in a region and thereby contribute to enhanced risk management of, for example, risk analysts and policymakers or insurance companies. The modeling framework was successfully applied in a proof-of-concept exercise in Vorarlberg (Austria). The results of the case study show that risk analysis has to be based on spatially heterogeneous flood events in order to estimate flood risk adequately.

Abstract. Meteorological time series with 1 h time steps are required in many applications in geoscientific modelling. These hourly time series generally cover shorter periods of time compared to daily meteorological time series. We present an open-source MEteoroLOgical observation time series DISaggregation Tool (MELODIST). This software package is written in Python and comprises simple methods to temporally downscale (disaggregate) daily meteorological time series to hourly data. MELODIST is capable of disaggregating the most commonly used meteorological variables for geoscientific modelling including temperature, precipitation, humidity, wind speed, and shortwave radiation. In this way, disaggregation is performed independently for each variable considering a single site without spatial dependencies. The algorithms are validated against observed meteorological time series for five sites in different climates. Results indicate a good reconstruction of diurnal features at those sites. This makes the methodology interesting to users of models operating at hourly time steps, who want to apply their models for longer periods of time not covered by hourly observations.

Alkyl nitrate (AN) and secondary organic aerosol (SOA) from the reaction of nitrate radicals (NO 3 ) with isoprene were observed in the Simulation of Atmospheric PHotochemistry In a large Reaction (SAPHIR) chamber during the NO 3 Isop campaign in August 2018. Based on 15 day-long experiments under various reaction conditions, we conclude that the reaction has a nominally unity molar AN yield (observed range 90 ± 40%) and an SOA mass yield of OA + organic nitrate aerosol of 13–15% (with ∼50 μg m –3 inorganic seed aerosol and 2–5 μg m –3 total organic aerosol). Isoprene (5–25 ppb) and oxidant (typically ∼100 ppb O 3 and 5–25 ppb NO 2 ) concentrations and aerosol composition (inorganic and organic coating) were varied while remaining close to ambient conditions, producing similar AN and SOA yields under all regimes. We observe the formation of dinitrates upon oxidation of the second double bond only once the isoprene precursor is fully consumed. We determine the bulk partitioning coefficient for ANs ( K p ∼ 10 –3 m 3 μg –1 ), indicating an average volatility corresponding to a C 5 hydroxy hydroperoxy nitrate.