Search

This paper presents the methodology for the construction of the KNMI'23 national climate scenarios for the Netherlands. We have developed six scenarios, that cover a substantial part of the uncertainty in CMIP6 projections of future climate change in the region. Different sources of uncertainty are disentangled as much as possible, partly by means of a storyline approach. Uncertainty in future emissions is covered by making scenarios conditional on different SSP scenarios (SSP1‐2.6, SSP2‐4.5, and SSP5‐8.5). For each SSP scenario and time horizon (2050, 2100, 2150), we determine a global warming level based on the median of the constrained estimates of climate sensitivity from IPCC AR6. The remaining climate model uncertainty of the regional climate response at these warming levels is covered by two storylines, which are designed with a focus on the annual and seasonal mean precipitation response (a dry‐trending and wet‐trending variant for each SSP). This choice was motivated by the importance of future water management to society. For users with specific interests we provide means how to account for the impact of the uncertainty in climate sensitivity. Since CMIP6 GCM data do not provide the required spatial detail for impact modeling, we reconstruct the CMIP6 responses by resampling internal variability in a GCM‐RCM initial‐condition ensemble. The resulting climate scenarios form a detailed storyline of plausible future climates in the Netherlands. The data can be used for impact calculations and assessments by stakeholders, and will be used to inform policy making in different sectors of Dutch society. Plain Language Summary: To prepare society for the effects of future climate change, we need to know what the future climate will be like. In this paper we explain the method that is used to construct six different scenarios that describe possible future climates of the Netherlands. The scenarios make assumptions about future greenhouse gas emissions, and are based on the outcomes of climate models that simulate the response of the climate to these emissions. The KNMI'23 climate scenarios show that strongly reducing global emissions strongly reduces the expected changes in the climate of the Netherlands. In the scenario in which global emissions continue to rise until 2080, Dutch society will have to adapt to a much stronger increases in heat and precipitation extremes, increased risks of droughts with low river discharge in summer, and increased risk of flooding due to high river discharges in winter. In the coming years the climate scenario data will be used to evaluate what needs to be done to keep the country a safe place for people to live in and to thrive in, under changing climate conditions. Key Points: We present a methodology for the construction of regional climate scenarios using a storyline approach to partition uncertaintyResults from CMIP6 are reconstructed with a GCM‐RCM initial condition ensemble to produce high‐resolution scenario data for end‐usersSix scenario variants cover emission uncertainty (high, moderate, low) and uncertainty in the regional response (dry‐trending, wet‐trending) [ABSTRACT FROM AUTHOR]

A dataset of 9 years in duration (2009–17) of clouds and radiation was obtained at the Cabauw Experimental Site for Atmospheric Research (CESAR) in the Netherlands. Cloud radiative forcings (CRF) were derived from the dataset and related to cloud cover and temperature. Also, the data were compared with RCM output. Results indicate that there is a seasonal cycle (i.e., winter, spring, summer, and autumn) in longwave (CRF-LW: 48.3, 34.4, 30.8, and 38.7 W m−2) and shortwave (CRF-SW: −23.6, −60.9, −67.8, and −32.9 W m−2) forcings at CESAR. Total CRF is positive in winter and negative in summer. The RCM has a cold bias with respect to the observations, but the model CRF-LW corresponds well to the observed CRF-LW as a result of compensating errors in the difference function that makes up the CRF-LW. The absolute value of model CRF-SW is smaller than the observed CRF-SW in summer, mostly because of albedo differences. The majority of clouds from above 2 km are present at the same time as low clouds, so the higher clouds have only a small impact on CRF whereas low clouds dominate their values. CRF-LW is a function of fractional cloudiness. CRF-SW is also a function of fractional cloudiness, if the values are normalized by the cosine of solar zenith angle. Expressions for CRF-LW and CRF-SW were derived as functions of temperature, fractional cloudiness, and solar zenith angle, indicating that CRF is the largest when fractional cloudiness is the highest but is also large for low temperature and high sun angle. [ABSTRACT FROM AUTHOR]

In this contribution four case studies in the Netherlands are discussed, investigating the role of atmospheric drivers, the probability of occurrence of compound flood events and their impact on regional water systems, and the potential changes therein in a warming climate. Observations of extreme compound events are rare by definition and observational datasets are often too short to look far in the tail of the probability distribution of the relevant variables. Here we use a 16-member initial-condition ensemble generated with the global climate model EC-EARTH and further downscaled with the regional climate model KNMI-RACMO2 over Western Europe at 12 km spatial resolution for the period 1950-2100, forced with historical emissions until 2005, and the RCP8.5 emission scenario thereafter. Model time series are used to drive a North Sea surge model. The periods 1951-2000 and 2026-2075 were selected to study 16 x 50 years of current and future climate, respectively. Two of the case studies concern water systems in the north of the Netherlands for which discharge to sea under gravity is blocked when sea levels are too high. Historical data suggest a correlation between storm surges and precipitation. The RACMO EC-EARTH simulations are used to investigate this relation for the present-day and future climate. The relation found in the current climate simulations is similar to the relation in the observations, with the exception of the far tails of the probability distribution. We currently investigate whether the climate model is realistically capturing the synoptic-scale patterns associated to the most extreme observed events. Hydrological and hydraulic models are used to study the impact of compound flood events on the regional water systems and the effect of potential measures against flood risks. The third case study considers the Dommel and Aa rivers, which both discharge under gravity to the river Meuse. In 1993 and 1995 high discharges from these rivers occurred simultaneously with high Meuse water levels, which resulted in large inundations around the city of Den Bosch. Correlations between precipitation in the three river basins and the timing of peak discharges in RACMO EC-EARTH correspond to historical observations. The effect of joint occurrence of high discharges in the three rivers on water level statistics in the city is assessed using hydraulic models.The final case study considers the area of Brabantse Delta. In case of extreme storm surge at the North Sea, the coastal flood defense structures are closed and Rhine and Meuse rivers cannot discharge to the sea. When this jointly occurs with extreme discharges, the Rhine and Meuse are forced to discharge into lake Volkerak. Discharge of smaller local rivers into this lake is blocked when lake water levels rise too high, posing flood risk in the area. Hydrological, hydrologic and hydrodynamic models are used to investigated the joint occurrence of high sea water levels, high Rhine river discharges and high local river discharges. [ABSTRACT FROM AUTHOR]