Investigation of polar mesocyclones in Arctic Ocean using COSMO-CLM and WRF numerical models and remote sensing data Mikhail Varentsov (1) , Polina Verezemskaya (1) , Anastasia Baranyuk (2) , Elizaveta Zabolotskikh (3) and Irina Repina (4) (1) Lomonosov Moscow State University, Faculty of Geography, Department of Meteorology and Climatology, mvar91@gmail.com ; (2) V. I. Il’ichev Pacific Oceanological Institute FEB RAS; (4) Russian State Hydrometeorological University, Satellite Oceanography Lab., (4) A. M. Obukhov Institure of Atmosphere Physics RAS, Air-Sea Interaction Lab., Polar lows (PL), high latitude marine mesoscale cyclones, are an enigmatic atmospheric phenomena, which could result in windstorm damage of shipping and infrastructure in high latitudes. Because of their small spatial scales, short life times and their tendency to develop in remote data sparse regions (Zahn, Strorch, 2008), our knowledge of their behavior and climatology lags behind that of synoptic-scale cyclones. In case of continuing global warming (IPCC, 2013) and prospects of the intensification of economic activity and marine traffic in Arctic region, the problem of relevant simulation of this phenomenon by numerical weather and climate models is especially important. This research is devoted to investigation of the ability to simulate polar lows of two modern nonhydrostatic mesoscale numerical models, driven by realistic lateral boundary conditions: regional climate model COSMO-CLM and weather prediction and research model WRF. Fields of wind, pressure and cloudiness, simulated by models, were compared with remote sensing data and ground meteorological observations for three cases, when polar lows were observed in different regions: Norwegian sea (p. 4), Kara sea (p. 5) and Laptev sea (p. 6). 4. Norwegian sea PL case 5. Kara sea PL case Three differently–typed polar lows were observed over Norwegian and Barents seas at 29 th – 31 st March, 2013. Western one had baroclinic nature and two northern were triggered by convective mechanisms (Verezemskaya P., Master thesis’2014). 1. Abstract 2. Data & methods 6. Laptev sea PL case Polar low was observed over the Kara sea at 30 th of September 2008 in frontal (warm) sector of the synoptic-scale cyclone. Its developing in so unusual conditions were caused by upper-level potential vorticity forcing (Verezemskaya P., Master thesis’2014). WRF model in short experiments, driven only by boundary conditions, and CCLM model in longer experiments, driven also by spectral nudging of synoptic-scale circulation from reanalysis, successfully simulated observed morphology and evolution of PLs in three different geographical regions. Experiments with CCLM without spectral nudging shows different from observation evolution of the PLs - the importance of this option for correct PLs simulations was shown. Comparison between reanalysis fields, modelling data and remote sensing data shows that reanalysis significantly underestimate wind speed and vorticity in PL’s and provide unrealistic distribution of cloud water, and the both models are possible to improve and detail this fields, providing better data for PLs diagnostic and research. Research supported by Russian Foundation for Basic Research: proj. №13-05-41443 «Extreme atmospheric events in the polar regions and its connection to changes in ice conditions» 8. Results 3. Model verification vs observations 1. Böhm U. et al (2006), CLM - the climate version of LM: Brief description and long-term applications. COSMO Newsletter. Vol. 6. 2. IPCC Fifth Assessment Report: Climate Change 2013 (AR5) Rep.,Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. 3. Sadovnichy V. et al. (2013), "Lomonosov": Supercomputing at Moscow State University. In Contemporary High Performance Computing: From Petascale toward Exascale (Chapman & Hall/CRC Computational Science), pp.283-307, Boca Raton, USA, CRC Press. 4. Skamarock, W. C. et al (2008): A description of the Advanced Research WRF Version 3. NCAR Tech Notes- 468+STR. 5. von Storch, H., Langenberg, H., & Feser, F. (2000). A spectral nudging technique for dynamical downscaling purposes. Monthly weather review, 128(10), 3664- 3673. 6. Zahn, M., and H. von Storch (2008), A long-term climatology of North Atlantic polar lows, Geophys. Res. Lett., 35, L22702 7. Zabolotskikh, E. V., L. M. Mitnik, and B. Chapron (2013), New approach for severe marine weather study using satellite passive microwave sensing, Geophys. Res. Lett., 40, 3347–3350, doi:10.1002/grl.50664. 8. Zabolotskikh E. V., Gurvich I. A., Chapron B., (2015). New regions of polar lows spread in Arctic as a result of sea ice area decreasing. Izvestia, Atmoscperic and Oceanic Physics. V. 2, 64–77 9. References Fig. 4.1. Reflected radiance by MODIS at 12:20 30 th of March. Blue line - sea ice border according ERA-Interim data. Fig. 4.2. Fields cloud liquid water content, sea-level pressure (SLP), horizontal wind speed and horizontal vorticity at 03 UTC 30 th of March: ERA-Interim reanalysis (cloud liquid water content data for 06 UTC), results of experiments with COSMO-CLM and WRF models, AMSR remote sensing data (at 03:12 UTC ± 25 min). ERA-Interim and AMSR data is interpolated to the COSMO-CLM grid. Wind spd. (m/s), SLP (hPa) Cloud liquid water, kg/m 2 ERA-Interim WRF CCLM SN AMSR (sat.) Quick Scat (sat.) Fig. 5.2. Reflected radiance by MODIS at 07:10 30 th of March. Blue line - sea ice border. Fig. 5.1. Fields of cloud liquid water, SLP, hor. wind speed and hor. vorticity at 21 UTC 30 th of September: ERA-Interim reanalysis (cloud liquid water for 18 UTC), results of experiments with COSMO-CLM and WRF models, AMSR radiometer data (at 22:06 UTC ± 25 min) and QuickSCAT scatterometer data (20:30 ± 15 min). ERA-Interim, AMSR and QuickSCAT data is interpolated to the COSMO-CLM grid. ERA-Interim WRF CCLM-SN AMSR Wind spd. (m/s), SLP (hPa) Cloud liquid water, kg/m 2 Fig. 2.1. Domains used for the numerical experiments with CCLM and WRF models CCLM Norw. sea WRF Norw. sea CCLM Kara sea WRF Kara sea CCLM Laptev sea COSMO-CLM (CCLM) - nonhydrostatic regional climate model (Böhm et all, 2006), developed by CLM-community and DWD: • Experiments launched for two month, starting at least one before the analyzed PL case; • Resolution: 0.1 lat/lon ⁰ of the rotated coordinate system (aprox. 11 km) – 320x220 grid cells, 40 vertical levels; • Lateral boundary conditions and SST from ERA-Interim reanalysis (0.75 ⁰ res.); • Experiments with spectral nudging technique (CCLM SN) according (Storch et al, 2000) with minimum wavelength to nudge about 500 km and without in (CCLM WN) were examined. • Version 5.0_clm2 with standard options, Tiedke convection; WRF – well-known weather forecast and research model (Skamarock et al, 2008): • Experiments lunched for several days starting just before the analyzed PL case; • Lateral boundary conditions and SST from ERA-Interim reanalysis (0.25 ⁰ res.); • Resolution: Norway and Barents seas - 10 km, 40 vertical levels, Kara sea - 5 km, 50 vert levels. • Version 3, No spectral nudging, radiation parameterizations from CAM (Community atmospheric model), boundary layer- Monin-Obukhov, microphysics - Goddard center, convection - only shallow by Grell-Devenyi, urbulence in PBL - Mellor-Yamada- Janich. In Kara sea modeling convection was resolved directly. Models were run on Lomonosov supercomputer of Moscow Stare University (Sadovnichy at. al, 2013). Remote sensing data: satellite cloud fields of MODIS (AQUA and TERRA), band №5 (1230 – 1250 µm); data from microwave radiometer AMSR-E and AMSR-2 microwave radiometer data (MODIS Aqua, GCOM-W1) for wind speed on 10 meters level and integrated atmospheric cloud liquid water; data from QuikSCAT scatterometer for wind speed and direction on 10 meters level. Hor. vorticity, sec -1 Hor. vorticity, sec -1 ERA-Interim CCLM-SN AMSR Hor. vorticity, sec -1 Wind spd. (m/s), SLP (hPa) Cloud liquid water, kg/m 2 0 5 10 15 20 2 3 4 Wind speed, m/s Month Obs CCLM WN CCLM SN -20 -10 0 10 2 3 4 Temperature, ⁰C Month Obs CCLM WN CCLM SN Norwegian sea domain (Slettness weather station in northern Norway): Kara sea domain (Troinoy island weather station): 0 5 10 15 20 8 9 10 Wind speed, m/s Month Obs CCLM WN CCLM SN Laptev sea domain (Kotelny island weather station): -10 -5 0 5 10 15 8 9 10 Temperature, ⁰C Month Obs CCLM WN CCLM SN WN SN T bias, ⁰C -4.3 -4.0 T correl. 0.77 0.84 SPD bias, m/c -1.5 -2.0 SPD correl. 0.37 0.52 WN SN T bias, ⁰C -4.6 -2.2 T correl. 0.91 0.93 SPD bias, m/c -0.3 -0.44 SPD correl. 0.50 0.78 0 5 10 15 20 8 9 10 Wind speed, m/c Month Obs CCLM WN CCLM SN -20 -10 0 10 20 8 9 10 Temperature, ⁰C Month Obs CCLM WN WN SN T bias, ⁰C -2.6 -2.6 T correl. 0.9 0.91 SPD bias, m/c -0.0 0.5 SPD correl. 0.41 0.72 Fig. 6.2. Reflected radiance by MODIS at 04:45 13 th of October. Blue line - sea ice border according ERA-Interim. Fig. 6.1. Fields of cloud liquid water, SLP, hor. wind speed and hor. vorticity at 00 UTC 13 th of October: ERA-Interim reanalysis. results of experiments with COSMO-CLM, AMSR radiometer data at 01:27 from (Zabolotskih et. al. 2015) Polar low was observed over the Laptev sea at 12 th – 13 st October 2007 in frontal part of pressure depression at occlusion front (Zabolotskih et. al. 2015) 7. Effect of spectral nudging (example for Norwegian sea case) 12Z 29.03 00Z 30.03 12Z 30.03 12Z 31.03 SN WN