| As the most abundant greenhouse gas, water vapor plays an important role in maintaining livable temperature environment and bridging the earth water cycle. The distribution and variation of water vapor in the atmosphere can greatly affect the climate and weather system. Unlike other greenhouse gases, for example, CO2, the content of water vapor is pretty variable in both time and space domain, which makes it difficult to be precisely measured, modeled and predicted.China is located in the junction of the largest continental plate (Eurasian Plate) and the largest ocean (Pacific Ocean) with wide geographical range and complex terrains. The climate and weather types are diverse in China, resulting in unique and complex variations of water vapor content in the atmosphere over this region. Some large-scale climate phenomenons, such as ESNO and East Asian Monsoon, have significant influences on Chinese climate and weather system. In addition, the generation, development and dissipation of many natural hazard weather systems, such as the typhoon and storm, are closely related to the variations of water vapor in the atmosphere. Therefore, studying and monitoring the distribution and variation of water vapor over China are dispensable in understanding the climate system and improving the weather forecast.Ground-based GNSS can be used to measure water vapor content based on the signal delays passing through the atmosphere. Compared to other tools, ground-based GNSS have the advantages of high accuracy, all-weather conditions, low-cost, high temporal resolution and homogeneous in long-term measurements. In this thesis, the ground-based GNSS station observations from CMONOC network as well as the ground synoptic station records, reanalysis products and radiosonde observations, were used to study the water vapor distribution over china, different periodical signals, variation mechanisms, relationships with other climate and weather phenomenons, radiosonde and reanalysis product errors, tropospheric models and the application of ground-based GNSS in improving weather forecast of disastrous weathers. The main conclusions drawn from this thesis are summarized as followings:(1) The PPP module of PANDA package developed by Wuhan University was used to process the GNSS observations from CMONOC network covering the period from 1999 to 2015. The mean accuracy of the derived ZTD product is around 3.9 mm, which is in the same level compared to IGS final ZTD products. Due to the incomplete pressure records at GNSS stations, the pressure measurements from nearby synoptic stations and the reanalysis products were also used to derive the air pressure at all GNSS stations. The reanalysis products were also utilized to calculate the water vapor weighted mean temperature (Tm). Error RMS for the derived air pressure and Tm are 0.7 hPa and 1.8 K, respectively. The ZTD were then converted to PW, and error analysis show the average error RMS of the final PW solutions are about 0.98 mm.(2) Based on the derived PW products, the geophysical distribution and variation of water vapor content over China between 1999 and 2015 were analyzed. The content of water vapor was shown to be positively related to the surface air temperature, with moistest region in southern China (?50 mm of PW) and driest in Tibetan Plateau (less than 20 mm of PW). The annual and semi-annual variation amplitudes are maximum in the middle and lower reaches of Yangtze River and North China Plain, respectively, while minimum in Tibetan Plateau and southwestern China. The time of minimum water vapor content in a day is relatively stationary, but changing in the time along with seasons for maximum water vapor content.(3) Based on the derived PW products, the correlation between the observed PW and modeled PW (only with temperature changes considered) was studied and used to determine the main factor (thermodynamic or dynamic process) contributed to the water vapor variations in different climate type regions. The results showed that the water vapor content variations in southern South China Sea are almost controlled by dynamic processes while mainly controlled by thermodynamic processes in subtropical and humid monsoon regions. For other regions, both two factors are important. The correlation between long-term water vapor variation at coastal stations and SOI index was also analyzed to study the impact of ENSO. Results showed that ENSO mainly has direct influences on the long-term variations of water vapor content over tropical coastal regions.(4) Based on the derived PW products, the errors of radiosonde humid observations over China from 1999 to 2015 were assessed according to the radiosonde types (GZZ2, GTS1, GTS 1-1 and GTS 1-2). Results showed that measurements from GZZ2 which was widely used in the early time mainly contain wet biases while measures from the later used GTS1 and GTS 1-1 contain dry biases and biases of GTS 1-2 are not significant. This results in artificial downward trend signals for radiosonde stations with type changes. In addition, due to the assimilation of the uncorrected radiosonde data, the long-term trend estimations based on the reanalysis products are also contaminated by this inhomogeneity issue in radiosonde data.(5) The CPT model was built based on ERA-Interim products where the model complexity and accuracy were adequately considered. Compared to the widely used GPT and GPT2 (5° and 1°) model, the pressure error RMS in China decrease from 5.92,5.14 and 5.04 hPa to 7.76 hPa, and the temperature error RMS decrease from 5.95,4.25 and 4.14 K to 4.07 K. On ther other hand, compared to ITG model, the number of CPT model coefficients decrease significantly. The prior Tm model and Tm-Ts linear model in China were also constructed, and the error RMS decrease from 4.45 and 4.19 K (for Bevis model and GPT2 model, respectively) to 3.81 and 2.97 K, respectively.(6) The real time ZTD grid products in China were produced, which can provide real time ZTD corrections to any ground station with accuracy of?1.5 cm. The grid products can significantly accelerate real time PPP convergence:for BDS real time PPP, the convergence time in 3 dimensions can be shorten by 20% and can be much better for vertical component (-50%).(7) The ground-based GNSS measurements were used to study the spatial and temporal variations of ZTD over stations in the key regions for June 23,2016 Jiangsu Funing super tornado event. The WRF 3DVAR data assimilation system was used to assimilate the ground-based GNSS observations and to prove the improvements contributed by the ground-based GNSS observations to the short-term forecasts of key variables (accumulated precipitation) in this event. |
PDF Full Text Request