Guest blog by Cenk Sezen
Rain-on-snow (ROS) floods, that is, the combined occurrence of snowmelt and rainfall events can cause huge economic impacts and endanger human lives due to a significant increase in the magnitude of runoff compared to only snowmelt or only rainfall events.
Forecasting of ROS floods is difficult because of their complex dynamics.
For example, relatively warm air temperature and high humidity can affect the melting of the accumulated snow pack, whereas high runoff may occur as a response to accumulated snow in the catchment. With climate change, alterations in climate variables including air temperature, rainfall amount and snow depth at different elevations will affect the frequency, magnitude and seasonality of ROS events.
Scope of the Short-Term Scientific Mission
Projected future changes in frequency, magnitude and seasonality of ROS events were investigated in four catchments located in Slovenia (Europe) with support from the COST Action DAMOCLES. The Short-Term Scientific Mission was carried out by Cenk Sezen from Ondokuz Mayis University (Turkey) who visited the University of Ljubljana (Slovenia) between 04/03/2019 and 29/03/2019.
As a first step, measured historical data (i.e. rainfall amount, discharge, evapotranspiration, air temperature, hypsometric catchment curve) was prepared for the utilization of the CemaNeige GR6J lumped conceptual hydrological model in four catchments located in Slovenia (Soča, Sava Dolinka, Savinja and Gradaščica). In all four cases, snowmelt has an important impact on runoff generation, especially during winter. Firstly, CemaNeige GR6J was calibrated and validated using past observational data. For the hydrological modelling of the situation with climate change, temperature and precipitation data obtained from five different combinations of global climate models (GCMs) and regional climate models (RCMs), namely CNRM-CM5-LR/CCLM4-8-17, MPI-ESM-LR/CCLM4-8-17, MPI-ESM-LR/ RCA4, EC-EARTH/HIRHAM5, IPSL-CM5A-MR/WRF331F for the RCP4.5 scenario were used (As an input to this study for the projections we used outputs from the project “The estimation of the average and extreme meteorological and hydrological conditions in Slovenia over the 21st century” that was carried out by the Slovenian Environment Agency where they did all the pre-processing of climate models.). In the next step of the study, the ROS floods were identified using the discharge threshold that was determined depending on the information about actual flooding at the location of the selected discharge gauging stations (this information was also provided by the Slovenian Environment Agency that is responsible for flood forecasting in Slovenia). Once the events above the threshold were determined, the ROS floods were extracted from the simulated series (i.e. each event above the discharge threshold and with melt values larger than 0.1 mm/d in at least one of the CemaNeige GR6J model’s five elevation zones). Then, seasonality, frequency and magnitude of ROS floods were investigated for the periods of 1981-2010 (baseline period), 2011-2040, 2041-2070 and 2071-2100.
According to the hydrological modelling results, an increase in the number of the ROS floods is projected in selected catchments towards the future.
For instance, maximum increase rate reaches to %170 (for 2041-2070) in Soča, %367 (for 2071-2100) in Savinja, and %144 (for 2041-2070) in Gradaščica Rivers compared to the baseline period. As for the seasonality analysis, it was seen that the strength of seasonality changes according to the periods and climate projections in each catchment. Thus, no generalization could be made regarding the seasonality. In addition, the majority of the ROS floods occurred and will continue to occur in autumn. Moreover, the modelling results showed that the magnitude of ROS events appears to be high for the projections (particularly for 2041-2070 and 2071-2100) compared to the baseline period (1981-2010) for the Soča River case study. In addition, it was observed that the magnitude of the ROS floods could increase in case of the EC-EARTH/HIRHAM5, IPSL-CM5A-MR/WRF331F climate projections for Savinja River. Based on the results obtained in the scope of the short-term scientific mission, we can conclude that climate change will affect the ROS floods (e.g. increase in the number of flood events) but due to the complexity of the ROS mechanism, no clear generalization about the future characteristics of the ROS floods can be made. Therefore, we will continue to research this topic.