Weather forecast and earth system models usually have a number of parameters, which are often optimized manually by trial and error. Several studies have proposed objective methods to estimate model parameters using data assimilation techniques. This paper provides a review of the previous studies and illustrates the application of ensemble-based data assimilation to the estimation of temporally varying model parameters in a simple low-resolution atmospheric general circulation model known as the SPEEDY model. As shown in previous studies, our results highlight that data assimilation techniques are efficient optimization methods which can be used for parameter estimation in complex geophysical models and that the estimated parameters have a positive effect on short-to medium-range numerical weather prediction.
The cloud microphysical structure for a freezing rain event between Jan 11 and Feb 4, 2008, over southern China is studied using the 30 km-mesh Weather Research and Forecasting (WRF) model simulations with four different microphysics schemes and CloudSat satellite observations. This 3-week-long freezing rain event, centered in the domain of 108°-113°E and 25°-28°N, has been extraordinarily rare over southern China during the recent 50 years. Except the Lin scheme, the other three microphysics schemes (Morrison, WSM6, and Thompson) yield WRF simulations that capture the temporal and spatial distribution of precipitation and surface air temperature associated with the freezing rain event, albeit the modeled center of precipitation is slightly drifted northward in comparison with the observation. The WRF simulations highlight the importance of the following characteristics of an atmospheric vertical thermal structure for forming freezing rain: above-freezing temperature in the middle troposphere (600-850 hPa) and below-freezing temperature in the lower troposphere (below 850 hPa). The temporal evolution of cloud structure and icing processing from Jan 11 to Feb 4 is also well simulated by WRF, which can be divided into two phases: (a) abundant liquid water is found below 700 hPa from Jan 20 to 26, 2008, and (b) liquid and solid water content coexist, and there are mixed-phase clouds in the whole column of atmosphere from Jan 26 to Feb 1, in particular, solid water dominated at the higher level, while liquid water dominated at the lower level. Such a change of cloud structure represents the typical ice-phase mechanism for the freezing rain such as on Jan 28, 2008, over the Hunan province. Further evaluation with CloudSat observation over the Hunan province shows that the WRF model can capture the warm and refreezing layers near the surface, although it overestimates the solid water content.
This paper presents a theoretical estimation of the strength of equatorial superrotation in planetary atmospheres by exploring a quasi-axisymmetric system that is zonally averaged primitive equations for a dry Boussinesq fluid on a rotating hemisphere with the effects of nonaxisymmetric eddies parameterized by eddy diffusion. The fluid is forced by Newtonian heating and cooling, and the horizontal eddy diffusion of momentum is assumed to be much stronger than the vertical one. In this system, the superrotation is maintained by the Gierasch mechanism, which possibly explains the superrotation of the Venus atmosphere by angular momentum transport due to the mean meridional circulation and horizontal diffusion. For the estimation, a quintic equation for a scalar measure of the superrotation strength is developed from the primitive equations. The quintic equation estimates the superrotation strength by its unique positive solution, which depends only on three nondimensional parameters: the external thermal Rossby number, the ratio of the radiative relaxation time to the timescale for the vertical diffusion, and the ratio of the planetary rotation period to the geometric mean of the timescales for the horizontal and vertical diffusion. The parameter dependence of the dominant dynamical balance is also investigated. The balance is a cyclostrophic, geostrophic, or horizontal diffusion balance, and in each balance, the equator-to-pole temperature difference is either nearly equal to that in the radiative-convective equilibrium state or significantly reduced by thermal advection. Steady-state or statistically steady-state solutions of the primitive equations are obtained by numerical time-integrations for a wide parameter range covering many orders of magnitude. The numerical solutions show that the theoretical estimates have a relative error of less than 50%, which is very small compared with the superrotation strength varying five orders depending on the external parameters, and show that the estimation is valid.
This study reports on the concentration of ice crystals measured in midlatitude cirrus clouds by a balloonborne hydrometeor videosonde (HYVIS), which has the advantage of measuring small ice crystals in the size range of 10-100 μm more reliably. The cirrus clouds were generally associated with warm or stationary fronts of synoptic-scale lows. The microphysical dataset consisted of 37 launches from Tsukuba, Japan, during the observation period 1994-2007. On the basis of the comparison with concurrent data by other airborne instruments in the laboratory, the ice crystal concentrations can be measured by the HYVIS with an uncertainty factor of 2-3, although significant uncertainties are still included in the size range 10-30 μm. The reliability of the measured concentrations is supported by the observed size spectra of the dataset in this study and the simulated total concentrations of ice particles with a parcel model. Vertical profiles of size distributions of cirrus cloud ice crystals were obtained for clouds with top temperatures ranging from −33° to −72°C and base temperatures from −3° to −49°C. Ice crystal concentrations varied approximately from the order of 10−1 to 102 L−1. Median ice crystal concentrations were typically several tens per liter regardless of temperature or their vertical location. While the concentrations were sometimes the highest near the cloud top, some clouds had their maximum concentration near the cloud base. As ice particles near the cloud base were usually in sublimation zones, it is suggested that crystal breakup through the sublimation process enhanced the concentrations in some cases. There was a large difference between the measured concentrations and simulated ones in earlier modeling studies of cirrus cloud formation that treated the ice crystal generation process through homogeneous ice nucleation of aqueous solution droplets, although the measured ones are probably affected by other physical processes such as secondary ice formation and gravitational sedimentation and turbulent mixing of ice particles after the initial cloud formation. Furthermore, a strong temperature dependence expected from heterogeneous ice nucleation formulas at relatively warm temperatures (> −25°C) was not found over all temperature ranges. Some implications for ice nucleation mechanisms in cirrus clouds in comparison with recent modeling studies involving heterogeneous ice nucleation at temperatures below −40°C are briefly discussed.
A numerical simulation was performed here using a regional climate model (RegCM3) at 60-km horizontal resolution to estimate the maximum impact of urbanization in East Asia on the summer monsoon precipitation in China. A new bulk urban parameterization was added into the land-surface scheme to capture the thermal, dynamic, and hydraulic effects of urban land cover. The 9-year simulation results showed that the large-scale change from natural-surface to high-density urban land cover induced an enhancement in the southwesterly wind speed in the early monsoon season across eastern China. In July and August, the enhanced southwesterlies moved to northern China, while a weakened southwesterly wind speed appeared south of the Yangtze River valley (YRV). Meanwhile, a large reduction in rainfall occurred in southern China with most reduction accounted for by convective precipitation reduction. This corresponded to the reduced water-vapor supply from the weakened southwesterlies in southern China, decreased convergence in the YRV, reduced convective instability in the lower troposphere, and lessened evapotranspiration from the impervious urban surface. Accompanied by the increased southwesterlies in the west and northwest of the urban region, a moderate increase in rainfall occurred in northern China. The changes in southwesterly wind speed and rainfall in southern China were more evident in the weak East Asian summer monsoon (EASM) years than in the strong ones. This underlines the impact of change in large-scale intensive urban land use on regional climate in the weak phase of the EASM. As fractional urban land cover was reduced, the changes in wind and precipitation were still shown but the spatial coverage and magnitude of these changes were reduced.
To contribute to a better understanding of the global carbon cycle, a high precision continuous measurement system for atmospheric O2/N2 ratio was developed using a fuel cell oxygen analyzer. To obtain highly precise values of the atmospheric O2/N2 ratio, pressure fluctuations of the sample and standard air were reduced to within ±0.005 Pa, with temperatures stabilized to 32.0 ± 0.1°C. The analytical precision of the system was estimated to be ±1.4 per meg for 24-minute measurement as the standard deviation (1σ) of replicate analyses of the same sample air. This analytical precision is sufficient for clearly detecting very small spatiotemporal variations of the atmospheric O2/N2 ratio. A new set of secondary and working standard gases with specified O2/N2 ratios were also prepared by drying natural air to dew points lower than −80°C using a specially designed H2O traps and then adjusting its amount of O2. The prepared five secondary standard gases were repeatedly calibrated against our primary standard, and their O2/N2 ratios were confirmed to be stable with no appreciable trend for over 570 days at least. A non-dispersive infrared analyzer was also installed into the measurement system to allow simultaneous measurements of the atmospheric CO2 concentration. The analytical precision of the CO2 concentration was estimated to be ±0.03 ppm (1σ). Using the new system, we initiated a systematic observation of the atmospheric O2/N2 ratio at Aobayama, Sendai, Japan in February 2007. The observed measurements clearly showed seasonal and diurnal cycles, along with short-term variations on time scales of several hours to several days, caused by terrestrial biospheric and human activities.
In this study, a linear baroclinic model (LBM) is developed from a three-dimensional (3D) spectral primitive equation model. With this LBM, we investigate the linear stability problem for various zonally varying basic states on a sphere. For a zonal climate basic state, we confirm that the traditional Charney and dipole Charney modes appear as the most dominant unstable modes in the synoptic to planetary scales. For a zonally varying basic state, we find that these unstable modes are modified by the regionality of the local baroclinicity of the basic state. Given the zonally varying barotropic basic state, we find that the barotropically most unstable standing mode appears to be the Arctic Oscillation (AO) mode. In this study, the eigensolution of the LBM is regarded as a generalized extension at the 3D normal mode at the motionless atmosphere to those of an arbitrary climate basic state. As an application of the LBM, various zonally varying basic states associated with the positive and negative AO indices are substituted into the LBM to find the response of the baroclinic eddies. According to the result, the positive feedback dominates in the Atlantic sector for positive AO index because of the presence of enhanced double-jet structure. When the AO index is negative, the eddy momentum flux converges in the mid-latitudes to shift the subtropical jet poleward in the Atlantic and Pacific sectors because of the intensified baroclinic instability. The positive feedback operates in a different way in the Atlantic and Pacific sectors depending on their double or single westerly jets. It is concluded that the baroclinically unstable modes are modified by the positive/negative AO index, so that the induced local eddy momentum flux shows a positive feedback to the AO.
Characteristics of raindrop size distribution (DSD) during the Meiyu season are studied using ground-based disdrometer measurements carried out in eastern China (Nanjing) from 2009 to 2011. The observational spectra are divided into convective and stratiform types. The results show that the histograms of the logarithm of the generalized intercept parameter (log10NW) and mass-weighted mean diameter of raindrops (Dm) are negatively and positively skewed, respectively, for both convective and stratiform rain. The absolute value of the skewness coefficient is higher for convective rain than for stratiform rain, in particular for the log10NW distribution. The mean log10NW and Dm values are 3.80 and 1.71 mm for convective rain and 3.45 and 1.30 mm for stratiform rain, respectively. The shape (μ)—slope (Λ) relationship of the gamma distribution and the radar reflectivity (Z)—rain rate (R) relationship are also derived for convective rain. The Z-R relationship is found to be Z = 368R1.21. The interpretation of the statistical parameters obtained in this study and possible mechanisms that yield difference and similarity in comparison with those in previous studies are discussed.
We used a 20-km-mesh non-hydrostatic regional climate model (NHRCM) nested in the JRA-25 reanalysis to reproduce snow phenomena of the present climate around Japan during 1985-2004 winter (December to March). To verify the NHRCM, we used monthly mean values for snow depth, surface air temperature, total precipitation, and snowfall derived from observations of the Automated Meteorological Data Acquisition System of the Japan Meteorological Agency. We obtained the following results: (1) Surface air temperatures were simulated over Japan, with cold biases of about 1°C on the Pacific Ocean side of Tohoku district and warm biases of 1-2°C on the Japan Sea side. (2) The model generally underestimated monthly total precipitation, snowfall, and snow depth, and the regression coefficients of the modeled values against observed values were only 0.3, 0.3, and 0.4, respectively. The NHRCM showed large negative biases along the Japan Sea coast from Hokkaido to the Chugoku district and small positive biases from eastern Hokkaido to the Pacific Ocean side of Tohoku district. (3) These biases decreased at elevations above 500 m. (4) Both modeled and observed surface air temperatures in winter were distributed bimodally. We derived statistical regression equations as a substitute for dynamical downscaling by the NHRCM to provide snowfall and snow depth according to region and elevation range.