The Three-dimensional Reconstruction Based On Meteorological Radar Data

Weather forecasting has been an important thing which is related to people’s livelihood. Over the years, many weather-related scientific and technical personnel are dedicated to research in the field, hoping to help weather forecasters make accurate and real-time forecast especially for convective weather and reduce the losses caused by natural disasters. Up to now, there have been many research results in the field of meteorology, which make great contributions to weather forecasting. However, almost all the research is carried out in two-dimensional space, which is lack of intuitive traits and has low degree of visualization. Therefore, this paper is dedicated to the research on three-dimensional modeling.In this thesis, firstly, the data which are needed in modeling are extracted from meteorological radar base data; secondly, the research object is modeled by related modeling methods; thirdly, three-dimensional model is rendered and displayed with OpenGL. The main work includes:1 The contours in correlated faults are extracted from meteorological radar base data. The contours are sampled by string sampling method proposed in this thesis. The test shows that the sampling speed is fast and the distribution of sampling points is reasonable. Now good basic data can be provided for the three-dimensional modeling.2 The sampled contours data which are obtained in step one are modeled on three-dimensional space by a series of algorithms. The contours splicing algorithm based on optimization of line slope is proposed to splice together the contours in adjacent faults in this paper. The test shows that good results can be obtained by this method.3 The three-dimensional reconstructed model is rendered and displayed with OpenGL. The model can also be rotated and scaled. At the same time, some basic features of modeling objects such as height and distance-related are extracted.Many samples are used in this paper to test the system. The test shows that the reconstruction results and the actual situation are basically consistent. According to the results of three-dimensional reconstruction, we can obtain some important information needed in weather forecasting directly, such as the rising and sinking of the airflow, the state of the cloud anvil and so on. In addition, through the three-dimensional reconstruction of monomer in time series, we can obtain all stages of development of the monomer, and then obtain the whole movement process of the monomer. We also can provide an important basis for the prediction of the motion state of the monomer in the next moment.