Forecasting sea level fluctuations caused by storm winds at the ports in Odesa Region of the north-western part of the Black Sea

Abstract

Increasing the draught of ships that may be accepted by ports for loading at their loading berths is one of the main tasks aimed at development and freight turnover enhancement of sea trade ports located in Odesa Region of the north-western part of the Black Sea (cities of Chornomorsk, Odesa and Pivdennyi). An operational forecasting of short-term sea level fluctuations caused by storm winds presents a critical task for ensuring safe navigation across the ports’ water area and approach channels. The article is devoted to analysing and discussing the results of tests of a simplified 2D hydrodynamic model designed for forecasting such phenomena as upsurge and downsurge of the sea level caused by storm winds in the vicinity of sea ports in Odesa Region of the north-western part of the Black Sea. Spatio-temporal variability of wind conditions at the sea-to-atmosphere boundary was set based on the data retrieved from a 10-day synoptic forecast using global atmospheric prediction model GFS (Global Forecast System). The study analyses the results of forecast of significant (the ones exceeding 30 cm) short-term sea level drops and rises at the ports which were observed in 2016, 2017 and 2020.

It was established that, in case of use of the GFS forecast data, the pattern of sea level denivellations caused by storm winds and their amplitude in the majority of events start approximating to the observed values provided the forecast has a 4-day lead time. Therefore the accuracy of wind conditions variability forecast with application of the GFS model having a longer lead advance time is not sufficient for forecasting the sea level fluctuations caused by storm winds.  The study made it possible to get an acceptable equivalence between the values of sea level denivellation amplitudes which were forecast with a 1-to-3-day lead time and the ones observed afterwards. In particular, when the forecast lead time is equal to »2 days, in relation to the expected storm conditions, the average absolute error for the forecast of sea level fluctuations amplitude constituted 7-8 cm, while its permissible value was defined as 15 cm, and the average relative error – 16-18%. It allowed making a conclusion that a hydrodynamic model option, applied alongside with the forecasting information on wind conditions variability retrieved with the help of the GFS weather prediction model, may be used for operational forecasting of short-term sea level fluctuations caused by storm winds with the forecast lead time of up to 4 days.