Analysis of temporal variability of measured and modeled vertical distributions of salinity and temperature in the Gulf of Finland during 10-year period

Abstract

The Gulf of Finland is the sub-basin of the Baltic Sea seriously affected by eutrophication. General Estuarine Transport Model (GETM) was used for modeling hydrophysical fields of the Gulf during the period from 1997 to 2005. The results of the hydrodynamic modeling are important input for ecosystem modeling, in which salinity and temperature variations play the most important role. An accurately simulated salinity field is to some extent a proof that the transport of passive bio-chemical tracers can also be simulated correctly. In this study validation of the results of GETM is performed and comparison with Modular Ocean Model (MOM) is provided. These two models differ in numerical schemes that are used for solving the model equations, in model setup and to some extent in forcing. At first the time series of surface and bottom temperature and salinity from GETM are visually compared with measurements. Long-term measurement data from three HELCOM monitoring stations representing western, central and eastern parts of the Gulf were used. In this study we focus on Taylor diagram that provides quick summary of the degree of patterns correspondence and allows seeing how well model simulates natural pattern. For statistical analysis the surface temperature and salinity have been given the values at a depth of 5 m, and the bottom salinity and temperature are the corresponding values at the lowest depth (about 60 m) at which measurements were carried out. The validation results were grouped similarly for both models. Modeled surface temperature showed good agreement with observed data in all three stations. Root mean square error (RMSE) was between 0.2 and 0.4, correlation coefficients between 0.94 and 0.98 and normalized standard deviations between 0.9 and 1.1 for the both models. Thus, seasonal cycles in the upper layer were reproduced well. Bottom temperatures and surface and bottom salinities were reproduced with lower quality. Bottom temperatures were better reproduced in the western and central Gulf than in the eastern Gulf. Surface salinity was simulated with the same quality in all stations by GETM, while MOM reproduced salinity better in the central Gulf compared to the eastern and western part. Bottom salinities were better simulated by MOM than by GETM. The latter showed larger variability due to higher spatial resolution.