Annotation of the results obtained in 2019
Scientists of the Shirshov Institute of Oceanology of the Russian Academy of Sciences designed and implemented a unique 40-year retrospective reconstruction – hindcast – of atmospheric circulation in the North Atlantic (10°N - 80°N) - Russian Academy of Sciences North Atlantic Atmospheric Downscalling (RAS-NAAD) with a spatial resolution of 14 km and 50 vertical levels ( up to 50 hPa). The reconstruction was created using the regional non-hydrostatic model WRF-ARW 3.8.1 for the period 1979 - 2018. NAAD data includes all the basic parameters of the surface and free atmosphere (at model dry-hydrostatic sigma levels) with a 3-hour time resolution. The three-dimensional array meets the whole spectrum of research requirements of meteorologists, climatologists and oceanographers working in both the research and operational fields. The use of high resolution and a non-hydrostatic model in NAAD allow for the first time to relistically reproduce the mesoscale dynamics of the atmosphere, primarily in subpolar latitudes. Figure 1. NAAD computational domain. The new dataset can be used to diagnose extreme weather events, polar mesocyclones (including polar lows), tropical cyclones, and other atmospheric phenomena that are not reproduced in coarser spatial resolution data. As an example, we showb the diagnosis of intense polar low on March 2, 2008 near the southern tip of Greenland. NAAD (A, E) in more detail, in comparison with other data sources - reanalysis ERA-Interim (B, F), ERA5 (C, G), ASRv2 (D, H) reproduces the spatial structure (upper panels) and turbulent heat transfer (lower panels) between the ocean and the atmosphere, as well as the presence of a polar mesocyclone in the pressure fields at sea level (black contours) and wind (color). Figure 2. Diagnostics of the polar low on March 2, 2008. Wind speed at 10 m (color) and sea level pressure (isolines) according to NAAD (A), ERA-Interim (B), ERA5 (C), ASRv2 (D). Total surface turbulent heat flux from the ocean to the atmosphere (latent + sensible, in color) and sea level pressure (isolines) according to NAAD (E), ERA-Interim (F), ERA5 (G) and ASRv2 (H). NAAD datasets includes predictive and diagnostic variables on the surface and in the troposphere, provided on NAAD grids with resolutions of 14 and 77 km for the period 1979 - 2018. Coarser resolution is used to control the effects of the spatial step along the grid. The entire NAAD data archive is freely available for research purposes and is 150 TB with individual annual files ranging from approximately 140 MB in low resolution (77 km) to 3.3 GB in high resolution (14 km) for surface variables and approximately 165 GB for 3D fields. All NAAD data output is organized as annual NetCDF variable files and is available at www.naad.ocean.ru for download using FTP and OPeNDAP access.

 

Publications

---

Annotation of the results obtained in 2017
The main scientific problem of the project is understanding of the chain of mechanisms through which the World Ocean affects the climate and weather on the continents. The modern understanding of these mechanisms makes it possible to say with confidence that the active role of the ocean is manifested on scales of several decades, while on the scales of years and decades the active role belongs to the atmosphere. In practice, this means that by observing a steady anomaly, for example, ocean surface temperatures over a decade, we can potentially that it will affect the climate of Europe, also on the scale of the decade. Or, that the change in the date of the onset of the spring flood, from year to year, is the result of the own dynamics of the atmosphere. However, even within this paradigm, we have not come to an understanding of how specific oceanic anomalies affect the continental climate. In our project, we explore the "lost" to some extent block of connections between the anomalies of the state of the ocean and the anomalies of atmospheric circulation and climate. Namely, we want to fully trace the chain of mechanisms associated with the transfer of the heat and moisture from the ocean to the atmosphere and the formation of climate anomalies (from energy exchange processes to the formation of extreme heat regimes and humidification on the continents). The "language" of the interaction of the air-sea is the exchange of heat and moisture, as well as the subsequent redistribution of heat in the atmosphere, due to various processes, such as cyclones in the midlatitudes. In the first year of the project, our studies were focused on the atmospheric mechanism, which controls the transfer of heat from the ocean to the atmosphere. We used the data of climatic reconstruction of the atmosphere state - reanalyses and developed a methodology for quantitative and quanlitative assessment of the role of atmospheric circulation and atmospheric cyclones in intensification or, conversely, the weakening of heat exchange between the ocean and the atmosphere. Our main result at this stage is the expansion of the traditional paradigm about the role of cyclones in the formation of heat exchange between the ocean and the atmosphere. Traditionally, cyclones are given leading role in the intensification of heat exchange between the ocean and the atmosphere. It is believed that the frontal structure of the cyclone - the presence of the warm and the cold sectors, always leads to an intensification of heat exchange between the ocean and the atmosphere, due to a sharp increase in the vertical gradients of temperature and humidity, if the cyclone spreads over the ocean. In our work, for the first time, we quantitatively and quantitatively compared turbulent heat fluxes from the ocean to the atmosphere and the trajectories and characteristics of all atmospheric cyclones. We have shown that not all atmospheric cyclones lead to a significant intensification of heat exchange between the ocean and the atmosphere, but only the closing cyclones in a series in the rear of which there is an anticyclone. Figure 1 illustrates the onset of atmospheric conditions leading to an intensification of heat exchange between the ocean and the atmosphere (Fig. 1, left and central panels) and heat exchange values close to the average in the North Atlantic (Fig.1, right panel). Figure 1. Anomalies of the total (latent + apparent) heat flux (color) calculated with respect to monthly mean LSHF values together with sea level pressure (SLP, lines) 18:00 03/01/2008 - left panel, 18:00 22/01 / 2008 - the central panel and 06:00 11/01/2008 - the right panel. The black point shows the positions of the NDBC buoy # 41048 The main difference between the cases presented in Fig. 1 (left and center panels) and Fig. 1 (right panel) is the presence and absence of an intense anticyclone over North America, with active storm track over the ocean. In Fig. 1 (right panel) over the Atlantic Ocean several cyclones are observed - a series of cyclones, but the heat fluxes are not generally high, with the exception of the region in the eastern part of the ocean. Interaction of the rear part of the cyclone and the anticyclone leads to the formation of an atmospheric dipole, providing a flow of cold air from the northern regions to the surface of the ocean. A sharp increase in the temperature and humidity gradients between the ocean surface and the atmosphere leads to a strong anomaly of the air-sea heat flux. Our studies have shown that the contribution of such regions (atmospheric dipoles) to the integral values of heat exchange between the ocean and the atmosphere is from 20 to 90% in various regions of the North Atlantic regionally. Thus, we demonstrated and numerically evaluated the importance of the interaction zones of cyclones and anticyclones for the intensification of the air-sea interaction processes. Such a mechanism was described for the first time. Another important result of our work is the development of a numerical experiment of the climatic reconstruction of atmospheric circulation in the North Atlantic with a resolution of 0.08 (14 km) degree for a period of more than 30 years (1979 to 2016). Available data of the state of the atmosphere - reanalyses are characterized by spatial resolution from 0.5 to 2.5 degree. Such a spatial resolution is not allowed to analyze processes whose scale "fails" into the distance between grid cells. For example, convective processes in the atmosphere, extremely high wind speeds, mesoscale cyclones, etc. One of the main objects of our research are atmospheric rivers - narrow corridors in which an abnormally high moisture transfer to the continents is formed. This phenomenon is the next step of moisture transfer from the ocean to the continents. For a detailed study of this mechanism, we developed the configuration of a numerical model of the WRF ARW atmospheric circulation with a spatial resolution of 0.08 degree (14 km). The scheme of the experiment is shown in Fig. 2. Figure 2. The region of numerical integration (in the center) and the scheme for setting up a numerical experiment with the WRF ARW model.

 

Publications

---

Annotation of the results obtained in 2018
The main scientific problem that was addressed by this study in the framework of the project over the past year is to obtain reliable high-resolution data on the state of the atmosphere in the North Atlantic on a climatic time scale. Such data are necessary for the subsequent detailed study of atmospheric rivers, cyclonic activity and processes of the air-sea interaction. Studies of atmospheric dynamics on a synoptic scale have long and successfully been carried out using information provided by atmospheric reanalysis. However, the climatic role of subsynoptic and mesoscale processes in modern science is still «terra incognita». To take into account these processes, it is necessary to have high-resolution data on the state of the atmosphere over a long (climatic) period of time. In this study, the main tool is the regional mesoscale dynamic model of the atmosphere WRF (Weather Research and Forecasting) version 3.8.1. In fact, using the data of modern reanalyzes of relatively coarse resolution, we conduct a high-resolution reconstruction of the atmospheric circulation. Since our study regiona is the North Atlantic, the data obtained from the long-term (from 1979 to 2018) atmospheric experiment were called NAAD (North Atlantic Atmospheric Downscaling). When performing the experiment, we use an optimally selected set of parameterizations of subgrid processes, as well as to prevent the “drift” of the model solution during long-term integration, the spectral nudging procedure is used. From the boundary conditions (ERA Interim reanalysis), we retain oscillations with a long wavelength of more than 1,100 km in the fields of the geopotential, components of wind speed and potential temperature. Thus, the mesoscale model “preserves” the synoptic mode of circulation from atmospheric reanalysis and at the same time reproduces mesoscale dynamics. The resulting dataset can be used in a wide range of scientific tasks, so in the future it is planned to organize open access to this data to other research teams. The calculation was carried out on the resources of the SAIL laboratory (IORAS) using an average of 512 cores for more than 365 days. The volume of NAAD is about 150 TV of disk space. At the moment, the website naad.ocean.ru has been created, which is at the stage of filling. One of the most sensitive to the spatial and temporal resolution of the model processes are the processes associated with convection and, as a result, convective precipitation. Figure 1 shows the maps of average precipitation for July (the most developed convection period) of 2015. Due to the complexity of the processes associated with precipitation, it does not make sense to expect complete coincidence. However, it is obvious that the high-resolution experiment NAAD (A) is best matched by the GPM - Global Precipitation Mission (D) remote sensing data than all other sources reviewed, including the most up-to-date ERA5 (F) reanalysis. Fig. 1 Monthly average daily accumulated rainfall for July 2015 (mm day-1) in NAAD HiRes (A), LoRes (B), ERA-Interim (C), GPM (D), GPCP (E) and ERA5 (F). Thus, we conclude that the precipitation in the NAAD data are well consistent with observational data, better than all available databases, including reanalyses, and can be used to assess various hydrological characteristics. Including those for the study of atmospheric rivers. NAAD also (Fig. 2) reproduces well the position of the main mid-latitude cyclone storm tracks in the North Atlantic in accordance with the storm track climatology based on global reanalyses, but the local number of cyclones in NAAD HiRes is 30-60% more than LoRes. It is noteworthy that NAAD LoRes also shows a slightly higher number of cyclones as compared with ERA-Interim, although the differences are within 3-5%. Figure 2 shows the winter (DJF) time series of the integral number of cyclones in the area of calculating the model of different intensity (determined by the pressure value at the center of the cyclone). NAAD HiRes allows identifying 2 times more cyclones compared to LoRes, which demonstrates high consistency with global reanalyses (Fig. 2a). It is important to note that these differences between HiRes and all other products (including LoRes) are formed mainly by moderately deep and shallow cyclones (Fig. 2b, d), while the number of deep cyclones in HiRes is better consistent with LoRes and reanalysis, and exceeds their estimates by 10-15%. Fig. 2 Winter (DJF) (1979-2018) number of cyclones in NAAD-HiRes (A) and the difference in number of cyclones between NAAD-LoRes and ERA-Interim (B) and NAAD-HiRes and ERA-Interim (C). Color - the number of cyclone trajectories per season (DJF) that passed through a circle with a radius of 2 ° latitude (~ 155,000 km2) It is also importran to note that the effective cyclone radius, which characterizes the cyclone size, is approximately 50–100 km less in the HiRes NAAD compared to the NAAD LoRes, and also 100–150 km less in comparison to the ERA-Interim (the figure is not shown). For the identification of atmospheric rivers (AR), an approach was developed that allows them to be detected, segmented and filtered, and has received the code name "SAIL_v1". According to an informal definition, the atmospheric river is a narrow, extended synoptic phenomenon associated with anomalous moisture transfer. The method of detection, segmentation and filtration developed by the AR uses an approach based on identifying IVT anomalies - Integrated Vapor Transport (above the 80th percentile), screening out false candidates for atmospheric rivers and then analyzing all connected regions where the moisture transfer anomaly exceeds the specified value percentile. Figure 3 shows an example of an atmospheric river according to NAAD data, for which a bilinear spline approximation of the middle axis was constructed and the segments perpendicular to this axis along which the width of the atmospheric rivers is estimated are shown. The result of these estimates is a set of widths, then this sample is subjected to filtering anomalies. At the same time, as can be seen, for example, in Figure 3 (c), the distribution of the widths is asymmetrical, therefore an empirical median value is taken as an estimate of the width of the atmospheric river. On this basis we quatify the geometric characteristics of the atmospheric rivers such as length, width, and ratio of length to width. (b) (a) (c) Fig. 3 (a) Figure of the AR, for which the estimated widths along the entire median axis (shown by the red line) along segments (shown by green lines) perpendicular to the bilinear spline approximation of the median axis; (b) ARK widths estimated at the points of the median axis; (c) histogram of the widths of the ARK after filtering the anomalies. The empirical median value is minimized.

 

Publications

---