Monthly stream temperatures along the Danube River: Statistical analysis and predictive modelling with incremental climate change scenarios

Abstract

Abstract The aim of the study is to analyse changes and predict the course of mean monthly water temperatures of the Danube River at various locations for the future. The first part of the study involves conducting a statistical analysis of the annual and monthly average air temperatures, water temperatures, and discharges along the Danube River. The study examines long-term trends, changes in the trends, and multiannual variability in the time series. The second part of the study focuses on simulating the average monthly water temperatures using Seasonal Autoregressive Integrated Moving Average (SARIMA) models and nonlinear regression models (NonL), based on two RCP based incremental mean monthly air temperature scenarios. To assess the impact of future climate on stream temperatures, the historical long-term average of the monthly water temperature (1990–2020) was compared with scenarios S1 (2041–2070) and S2 (2071–2100). The simulation results from the two stochastic models, the SARIMA and NonL, showed that in scenario S1, the Danube River’s average monthly water temperature is projected to increase by 0.81/0.82°C (Passau), 0.55/0.71°C (Bratislava), and 0.68/0.56°C (Reni). In scenario S2, the models predict higher increases: 2.83/2.50°C (Passau), 2.06/2.46°C (Bratislava), and 2.52/1.90°C (Reni). Overall, the SARIMA model proved to be more stable and effective in simulating the increase in monthly water temperatures in the Danube River.