非线性映射的气象数据可视化及其应用  

Nonlinear mapping based meteorological data visualization and application

作　　者：聂俊岚[1,2] 刘益萌 陈贺敏 

机构地区：[1]燕山大学信息科学与工程学院,河北秦皇岛066004 [2]燕山大学河北省计算机虚拟技术与系统集成重点实验室,河北秦皇岛066004

出　　处：《燕山大学学报》2015年第3期276-282,共7页Journal of Yanshan University

基　　金：国家"863"高技术研究发展计划;河北省科技支撑计划项目(1420311D);河北省自然科学基金资助项目(F2013203099);燕山大学青年教师自主研究计划课题(13LGB016);河北省高等学校科学技术研究重点项目(ZH2012016)

摘　　要：为解决传统基于2.5维面的数据分析手段单一,未充分结合地理环境信息,难以快速实时获得任意区域气象变化规律的问题,提出非线性数据映射的2.5维面可视化方法,经过坐标转换应用至球面并利用GPU进行快速绘制;从区域和时域两方面进行全球气象数据可视分析:通过用户交互提取任意区域信息,定量展示分析结果并进行反馈;引入时间维展现选定区域气象数据变化趋势,并用不同表现形式进行统计分析。实验表明,本文可视化效果及可视分析方法能直观清晰地表达和分析气象数据的时间和空间关系,直观的视觉效果、实时交互的手段有助于气象预测和数据分析。The traditional data analysis methods based on 2.5 dimensional surfaces are singularity, not fully integrated with geograph- ical information, and difficult to obtain the change characteristics and laws of meteorological data in any region.To solve the problem, the nonlinear data mapping method in 2.5 dimensional surface visualization is proposed.The data is applied to the sphere after coordi- nate transformation and then fast rendered with GPU.The global meteorological data visualization is analyzed from the region and the time domain： the data analysis results of the regional information which is extracted interactively are displayed and reversely displayed quantitatively.The time dimension is introduced to display the mutative trend of meteorological data in selected area, and the results of statistical analysis are showed in different ways.The experimental results show that visualization and the visual analysis methods in this paper can express and analyze the relation between time and space of meteorological data intuitively and clearly.Furthermore, the intui- tive visual effects and real-time interactive means make the meteorological forecast and data analysis more easier.

关 键 词：2.5维面 非线性数据映射 GPU 实时交互 可视分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]