地基GPS反演大气可降水量及其在台风分析中的应用研究

发布时间：2018-01-27 10:25

本文关键词： 地基GPS气象学对流层延迟大气可降水量台风分析　出处：《东华理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：地基GPS反演大气可降水量作为一项新兴的大气探测技术,在空间天气与GNSS遥感等方面得到广泛的应用研究。该技术以高精度、高时空分辨率、近时间观测等特点而备受的关注,其在数值天气预报、天气特征变化分析等研究中的作用越来越重要。本文基于香港卫星参考网的观测数据,重点分析地基GPS反演大气可降水量过程中所涉及的问题以及台风影响下GPS水汽的变化特征。主要研究内容如下:(1)阐述GPS反演大气可降水量的基本原理,并以南昌为试验区,利用南昌探空站资料建立南昌加权平均温度模型,与其他模型比较,结果表明,要获取高精度的可降水量信息,必须使用基于该地区气象资料建立的加权平均温度模型,同时,依据探空数据计算可降水量的原理,对解算过程中值得重视的问题进行分析。(2)分析论证GPS反演可降水量过程中模型选择的问题。对Saastamoinen模型、Hopfield模型以及Black模型进行精度和适用性评估。结合IGS站的观测数据,针对映射函数模型NMF、GMF和VMF1进行分析,以ITRF08框架坐标和CODE提供的ZPD产品为真值,比较模型的定位精度与对流层延迟解算效果,结果表明:GMF和VMF1在U方向精度和对流层延迟解算精度相当,且高于NMF模型。此外,针对影响对流层延迟解算的影响因子进行分析。(3)根据地基GPS反演可降水量的流程,对地基GPS反演可降水量的精度进行验证。此外针对影响可降水量解算的因素进行分析,结果表明:海潮改正引起的香港地区可降水量误差在1-3mm之间,对内陆几乎没有影响;快速星历和最终星历的解算精度相当,超快速星历的解算精度随着观测部分时间的增加而提高,最大差值为1.100mm,基本满足实时数值天气预报的要求;与Hopfield模型和Black模型相比,Saastamoinen模型更适合香港地区,此外在无实测气象数据时,GPT/SAAS模型比EGNOS适用性更好;对GPS_GLONASS组合系统反演精度进行分析,组合导航系统反演精度达到可降水量精度要求,平均偏差为-0.5380,均方差为2.4788,相关系数为91.68%,但反演精度却弱于地基GPS反演的精度。(4)结合香港卫星参考站网数据和香港天文台提供的雷达回波图像以及实际雨量分布信息,对影响香港地区的四次台风进行可降水量分析,结合台风期间温度、气压、相对湿度等气象参数的特征变化,对台风的预报与分析起到更好的研究作用。
[Abstract]:The retrieval of atmospheric precipitable water by ground-based GPS, as a new atmospheric detection technology, has been widely studied in space weather and GNSS remote sensing. This technique is of high precision and high spatial and temporal resolution. Recent time observation has attracted more and more attention. It plays an increasingly important role in the research of numerical weather forecast and analysis of weather characteristics. This paper is based on the observation data of Hong Kong satellite reference network. This paper focuses on the analysis of the problems involved in the retrieval of atmospheric precipitable water by ground-based GPS and the variation characteristics of GPS water vapor under the influence of typhoon. The main research contents are as follows: 1). The basic principle of retrieving atmospheric precipitable water by GPS is described. Taking Nanchang as experimental area, the weighted average temperature model of Nanchang is established by using the data of Nanchang sounding station. Compared with other models, the results show that high precision information of precipitable water should be obtained. A weighted average temperature model based on meteorological data in the region must be used, and the principle of calculating precipitable water based on sounding data must be used. The problem of model selection in the process of GPS inversion of precipitable water is analyzed and demonstrated. The Saastamoinen model is analyzed. The accuracy and applicability of Hopfield model and Black model are evaluated. Based on the observation data of IGS station, the mapping function model NMF-MF and VMF1 are analyzed. Taking the ITRF08 frame coordinate and the ZPD product provided by CODE as the true value, the positioning accuracy of the model and the effect of the tropospheric delay solution are compared. The results show that the accuracy of the U direction and the tropospheric delay solution of the VMF1 is similar to that of the NMF model, and the accuracy of the U direction is higher than that of the NMF model. Based on the analysis of the factors affecting the solution of tropospheric delay, the flow of retrieving precipitable water based on foundation GPS is presented. The accuracy of retrieving precipitable water by GPS is verified. In addition, the factors affecting the calculation of precipitable water are analyzed. The results show that the error of precipitable water in Hong Kong caused by the tide correction is between 1 and 3 mm, which has little effect on the inland area. The resolution accuracy of the fast ephemeris and the final ephemeris is about the same. The resolution accuracy of the super-fast ephemeris increases with the increase of the observation time, and the maximum difference is 1.100mm. It basically meets the requirement of real time numerical weather forecast. Compared with the Hopfield model and the Black model, the Saastamoinen model is more suitable for the Hong Kong region, in addition, in the absence of measured meteorological data. The applicability of GPT/SAAS model is better than that of EGNOS. The inversion accuracy of GPS_GLONASS integrated system is analyzed. The inversion accuracy of integrated navigation system reaches the requirement of precipitable water accuracy, the average deviation is -0.5380, and the mean deviation is 2.4788. The correlation coefficient is 91.68%. However, the accuracy of inversion is weaker than that of ground-based GPS inversion. It combines the data of Hong Kong satellite reference station network, radar echo images provided by the Hong Kong Observatory and the actual rainfall distribution information. The precipitation analysis of four typhoons affecting Hong Kong area, combined with the changes of meteorological parameters such as temperature, air pressure and relative humidity during typhoons, plays a better role in the prediction and analysis of typhoons.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P228.4;P412

【参考文献】