Multi-dimensional Dynamic Visualization And Analysis For Hydro-meteorological Field Data Based On Virtual Globe

In recent years,hydro-meteorological disasters have occurred frequently in China,which have brought great threats to people’s lives and property safety.With the continuous improvement of detection methods and technologies,hydro-meteorological observation data is characterized by multi-dimensional,dynamic,multi-scale,and massive.Real-time(near-real-time)visualization of these data is of great significance for the rapid study of hydro-meteorological changes,especially in disaster emergency environments.Multi-dimensional dynamic visualization uses a simulated threedimensional virtual geographic environment to visualize professional obscure hydrometeorological information in the form of images,animations,etc.,which enable researchers to analyze the overall process and situation of disasters and emergency commanders to assist in efficient decision-making.Therefore,it is of great significance to comprehensively study multi-dimensional dynamic visualization of hydrometeorological fields for analyzing the dynamic changes of air currents and water currents,and assessing the impact of changes in hydro-meteorological disasters.However,there are a series of challenges in studying unified multi-dimensional visualization analysis for hydro-meteorological data.First,the spatial scale and dimensions of hydrological and meteorological data are very different.Hydrological and meteorological phenomena are the most common environmental constituents on the earth.They affect each other and complement each other.However,due to its wide fluidity,the meteorological field is usually a large-scale three-dimensional data.And the hydrological field is usually distributed on the ground surface,such as rivers,lakes,etc.Hydrological observations are usually small-scale two-dimensional data.Therefore,in order to realize multi-dimensional visualization of hydro-meteorological data,it is necessary to build a unified data description and an efficient data organization model.Secondly,for the visualization of large-scale hydro-meteorological data,the modeling method for wide and massive data needs to be studied and optimized for good visual effects and efficiency.For example,for large-scale meteorological volume rendering,the traditional cube rendering method is difficult to simulate the shape of the large-scale arc-shaped volume,so it is necessary to study efficient block rendering or other irregular volume rendering methods.For multi-dimensional dynamic visualization,the increase of the time dimension means that the system needs to continuously visualize the data,which requires more efficient processing and visualization algorithms.Finally,because users have different backgrounds,skills,and interests,visual model interaction and analysis can meet users’ visual needs for specific areas or specific effects.At the same time,external information input,integration and rendering play an important role in fusing multi-source information.In order to solve the above problems,this paper put forwards the unified data organization,multi-dimensional dynamic visualization and interactive dynamic analysis methods of hydro-meteorological data based on a virtual globe.The core idea is to realize a unified description and efficient organization of hydro-meteorological data based on the multi-scale,multi-dimensional,dynamic visualization framework of the virtual globe,and then optimize a variety of multi-dimensional dynamic visualization models,and finally improve the interaction and analysis capabilities of the visualization models.The main research contents include:(1)A multi-level organization method of field data based on virtual globe is proposed.Voxel based model for hydro-meteorological data of different scales,dimensions and time series is realized;Multi-leveled data organization based on variables,time and space,and multi-resolution processing of memory-based hydrometeorological data,are implemented.(2)Multi-dimensional dynamic visualization models including particle tracking,streamline,texture mapping and volume rendering are improved.The particle tracking model based on geographic cellular automata achieves adaptive control of particle size and density,and improves dynamic visualization effect and efficiency.The streamline model is implemented based on the adaptive step sampling,which not only considers the global view distance but also takes into account the characteristics of local particle velocity and curvature changes.The adaptive step sampling improves the efficiency and effect of streamline rendering.The dynamic texture mapping model realizes a multi-resolution water surface mesh construction and multi-texture fusion,which achieves a good visual effect of river water surface.The GPU-accelerated large-scale volume rendering model considers the spatial distribution of the data volume,and realizes the block slicing or sphere ray slicing method.It optimize the volume slicing efficiency at different camera interaction and view angle,which improve the dynamic rendering efficiency and effect.(3)The interactivity and analysis capability of visualization models are improved.Users can interact with the parameters of the dynamic visualization model and combine multiple models for mixed rendering to achieve a specific visualization analysis effect.The ability of visualizing and analyzing the profile of a specific area,and the dynamic tracking and analysis capabilities of hydro-meteorological field(such as eddy tracking)are implemented.Expert users can involve and external information can be integrated.In order to verify the validity and practicability of the above methods,a fourdimensional hydro-meteorological visualization analysis system was developed in this paper.The methods proposed in this paper are verified based on two typical cases:large-scale three-dimensional “Typhoon Rananim” landing data and small-scale twodimensional Yangtze river flood observation data.The experimental results show that the multi-dimensional dynamic visualization analysis method proposed in this paper has the following advantages:(1)General hydro-meteorological data support.Unified description,modeling and visualization are implemented for hydrological and meteorological observation data at different scales,different dimensions,and different time series.(2)Integrating lots of efficient multi-dimensional dynamic visualization models.The system integrates ten commonly used vector field and scalar field visualization models,including particle tracking,streamlines,texture mapping,and volume rendering.According to the different characteristics of the models,LOD,adaptive sampling,GPU acceleration and other optimized technologies are applied to optimize the efficiencies and effects,which improve the spatial perception of the models.(3)Flexible interactivity and analysis for the visualization models.Users can interactively configure model parameters and freely combine multiple models as required,analyze specific areas or features of the data field,or integrate external information for plotting.