Vol 55, No 1 (2019)

This paper presents a review and analysis of approaches to data assimilation in problems of geophysical hydrodynamics, from the simplest sequential assimilation schemes to modern variational methods. Special attention is paid to the study of the problem of variational assimilation in a weak formulation, in particular, to the construction of an optimality system and the estimation of the covariance matrices of the optimal solution errors. This is a new direction of research in which the author has obtained some results.

This paper updates the results obtained earlier with the climate model developed at the A.M. Obukhov Institute of Atmospheric Physics of the Russian Academy of Sciences (IAP RAS CM) and describing the impact of atmospheric sulfur dioxide on the terrestrial carbon cycle. Since no global data are available for the near-surface SO₂ concentration, a statistical model was used to reconstruct its concentration from appropriate data on surface sulfate concentrations; the coefficients of this model were tuned by using the output of the RAMS-CMAQ atmospheric chemistry and transport model. The results obtained in this study are generally consistent with data reported earlier. Specifically, the most significant SO₂ impact on the terrestrial carbon cycle was found to be for southeastern North America and for Europe. However, this impact for southeastern Asia obtained in this study is considerably weaker than the value obtained earlier, which can be explained by the excessive moisture content in the atmosphere used in the IAP RAS CM for this region.

The results of studying spatial–temporal CO₂ variations near St. Petersburg in 2014–2017 based on satellite measurements (OCO-2 satellite) and ground-based spectroscopic and local measurements are presented. According to satellite data, the full amplitude of the spatial–temporal variations for the average CO₂ mixing ratio (XCO₂) amount to 57.7 ppm (over 14%). The maximal XCO₂ spatial variations during 1 day of observations (March 17, 2015) were 46.8 ppm (more than 10%). A comparison of CO₂ satellite and ground-based spectroscopic measurements has shown that ground-based measurements in the NDACC observation system after the correction of systematic differences from the TCCON system can be used to validate satellite measurements. Ground-based local measurements of the near-surface CO₂ mixing ratio at Peterhof do not correlate either with spectroscopic ground-based or satellite measurements due to both mesoscale CO₂ variations and significantly different spatial averaging kernels of direct and remote measurements.

Conditions for the occurrence of the “voice of the sea” within the infrasonic range are studied and its parameters are determined from measurements performed in the waters of the Black Sea (2011 and 2016) and the Sea of Okhotsk (2017). Different parameters (mean correlations, acoustic-arrival spectra, directions, and phase velocities) of high-frequency infrasounds (1–10 Hz) recorded during the 2011 and 2016 experiments carried out in Katsiveli (Crimea) are compared. Wind conditions over the Black Sea waters and the conditions of propagation of acoustic waves along the direction of their arrivals during these measurements have been studied in detail. In both cases, atmospheric vortices are observed in the direction of infrasound arrivals, which cause the wind to change its direction over the sea surface. The interaction between two oppositely directed atmospheric vortices (according to data obtained in 2011) and a vortex observed to the west of the recording point (according to data obtained in 2016) result in the generation of infrasounds. The generation of microbaroms and the voice of the sea due to wind-direction variations, which cause the nonlinear interaction of surface waves propagating in opposite directions and the formation of their 2nd harmonics in the form of standing surface waves, is discussed. The most probable regions of infrasound generation are determined from an analysis of the profiles of wind velocity and direction along the path of infrasound arrivals and acoustic-pressure fields calculated using the parabolic-equation method according to the \({{C}_{{{\text{eff}}}}}\) profiles in the direction of infrasound propagation. In both cases, these regions coincide with zones in which the wind velocity had dropped to zero and the wind direction had reversed. The infrasound within a microbarom frequency range of 0.2–0.3 Hz and the (higher frequency) sea voice with a mean frequency of 5.5 Hz, which were simultaneously recorded from the same direction, are given as an example.

A train of internal waves ~16 m in height and 101–131 m in length, abnormally large for the Black Sea, is studied. The measurements have been conducted in the water area near Gelendzhik using a towed Acoustic Doppler Current Profiler (ADCP), an RBR concerto hydrological miniprobe, and a moored thermistor string and applying the spatial spectral analysis of a radar satellite image obtained at the moment of sea-truth measurements. Registered anomalous internal waves have a nonlinear character (soliton-like shapes of wave profiles, amplitude dispersion, changing distance between the crests, etc.). Vertical components of orbital current velocities in the internal waves reach 0.20 m/s. Quasimonochromatic spectral maxima due to the surface manifestations of internal waves having lengths corresponding to those measured by ADCP were registered using the satellite radar image. The source of this internal wave train with such anomalous amplitude was a cold atmospheric front and related low-pressure region.

A new version of the one-dimensional thermo-hydrodynamic and biogeochemical model LAKE2.1 is presented. The model is supplemented with a description of the dynamics and vertical distribution of salinity in an ice cover. Simulation results are compared to in situ and satellite data of water temperature and ice cover at Uvs Nuur Lake (Mongolia) from 2000 to 2015. It is shown that underestimating the mixed-layer depth by the model with standard turbulence closure k−ε during summer and autumn leads to a significant shift in the timing of the onset of ice. It is also demonstrated that, while neglecting the salinity of the lake, the freeze-up according to the model happens 16–17 days earlier than in reality. This error is removed if the effect of salinity on water density and freezing temperature is included. However, in this case, the model underestimates the maximal seasonal ice thickness on average by ≈0.2 m. In turn, this error decreases an order of magnitude if the dynamics and vertical distribution of salinity in ice are simulated in the model.