| Cloud observation and study have great significances to weather forecasting, climate prediction and weather modification. In recent years, millimeter-wave radars have developed rapidly and become the main cloud dectection device. In 2013, Chinese Academy of Meteorological Sciences developed a Ka-band solid-state multi-mode cloud radar, and it had acquired a lot of valuable data during 2014~2015 summer, at the “South China Monsoon cloud-precipitation Experiment” and “China’s 3th Tibetan Plateau Atmospheric Science Experiment”. This dissertation focuses on the researches of cloud radar Doppler spectra processing, radar detection ability, radar data quality, data quality control and retrieval of air vertical speed in Tibet Plateau clouds. The main contents and conclusions include:(1) Emphatically introduced the cloud radar system features, technical specifications, signal processing methods and observation mode design. SNR method and individual signal source method were utilized for radar calibration. The results show that solid-state transmitter make radar system relatively stable, as well as a long continuous observation capability. Radar has a satisfactory sensitivity, the minimum detectable reflectivity factor can approach-39.5d BZ at 5km. Three different kinds of observation modes jointly run not only to ensure the comprehensiveness of radar detection of different cloud types, but also to ensure the reliability of measurements. Input and output of radar horizontal- channel and vertical-channel signal have high linear relationship, radar calibration can be performed accurately.(2) Doppler spectra data processing methods of Ka-band cloud radar had been researched, detailed contents include, spectra data simulation, quantitative evaluation of noise level, cloud signal identification and spectral moment computation. The results show, as the spectral width under 2 m/s, 1m/s and 4m/s, corresponding to Cirrus, Boundary layer and Precipitation modes respectively, noise level calculation was quite accurate with small deviations. Multi-mode design can avoid the noise level grossly overestimate. In terms of accuracy, stability and adaptability of noise level, segmentation method is the best choice. Positive and negative changes in skewness and kurtosis reflect changes in cloud droplet distribution and phase in the growth of ice clouds and transformation process of cloud to precipitation, they are very useful indicators.(3) Theoretical methods and real measurements, detected by different band radars, were used to study the cloud observation ability and data quality of cloud radar. The results show, actual weak echo detection ability of this radar is consistent with the theoretical calculation. For strong echo, as the particle radius between 0.6 to 1.24 mm, Mie scattering can make radar echo stronger than true value for 0~1.21 d B; as the particle radius greater than 1.24 mm beginning, Mie scattering can make radar echo much weaker than the true value, the bigger the radius, the more serious weak. In terms of cloud structure, shape, and intensity distribution, cloud radar measurements are consistent with other instruments; howerver, the sensitivity is much higher. Cloud radar can be able to detect lower cloud bases, higher cloud tops and more cloudages. Comparing cloud radar with other instruments on the reflectivity, mean Doppler velocity and spectrum width, the results show, for altostratus cloud, cloud radar has a mean deviation of reflectivity under 0.5d B, mean deviation of mean Doppler velocity is 0.1m/s, mean deviation of spectrum width is less than 0.1m/s. For strong cumulus cloud, Mie scattering make strong echo center weak for 0~4d B, mean deviation of mean Doppler velocity is 0.2m/s, mean deviation of spectrum width is less than 0.2m/s. For weak precipitation cloud, reflectivity of cloud radar gradually become weak from 27 d BZ, bright band can be weak up to 3d B; the bias of mean Doppler velocity is negligible, and the bias of spectrum width does not exceed 0.1m/s. Compared to longer wavelength radars, cloud radar Doppler spectra can detect weak cloud signals, cloud base signals are stronger, cloud middle signals are coincident, while, cloud top signals are weaker to a certain extent caused by Mie scattering in the strong cumulus. For spectral intensity distribution and bimodal/trimodal features, cloud radar is consistent with other radars. Taking into account the different radar beam width, wind shear and turbulence, data quality of cloud radar are considered to be quite reliable.(4) Data quality control methods for velocity aliasing, low-level plankton clutter, noise and radial interference of cloud radar were researched. Actual cases were used to verify and analyze the effects. The results show, the proposed velocity dealiasing method have a well effect, the success rate of 15 cases of different cloud processes both reach 100%. Statistical Reflectivity threshold and linear depolarization rate threshold can filter out 100%/97.84% of plankton clutter in South China / Tibetan Plateau.(5) A retrieval method of air vertical speed in Tibetan Plateau cumulus was studied using cloud radar multi-mode Doppler spectral data. Reflectivity changes, positive speed positions in mean Doppler velocity, large spectrum width areas and falling speed relationship with reflectivity were used to analyze the retrievals. The results show that, the inversions are reasonable and reliable. In terms of accuracy, for light cumulus and cumulus, retrieval deviations can be controlled under one velocity resolution, the bias of weak low-level cumulus is smallest, while, the bias of cumulonimbus is relatively larger, but compared with air vertical velocity, the retrieval bias can be controlled in an order of magnitude less. Z-0ω relationship was used to examine the retrieval reliability preliminary, the results show, retrievals were consistent with the theoretical speculations. |
PDF Full Text Request