基于MODIS数据的海表温度反演研究

发布时间：2018-04-19 07:16

  本文选题：海表温度 + 劈窗算法　； 参考：《中国科学技术大学》2017年硕士论文

【摘要】：海洋幅员辽阔,占地面积极大,约是全球面积的70%左右,海洋温度变化会在很大程度上影响地球的温度变化,因此,研究海洋表层温度至关重要。海洋蕴含着巨大的能量,海水的热容量非常大,其温度的微小变化在影响人类生存环境的同时,会给全球的天气带来非常大的改变,同时在一定程度上给局部地区的天气带来变化。本文分析了海表温度反演算法的进程。在深入分析大量已有成果的同时,结合MODIS的特点,以MODISL1B遥感影像数据作为数据源,以南海地区为示范区,对海表温度反演的过程作了一系列的研究。具体内容如下:(1)以Terra-MODIS数据作为数据源,基于如今已有的结果,通过对比和分析,最终从众多的反演算法中,选出最适合本文研究区域的海表温度反演算法,然后应用覆盖研究区域的实测数据做反演实例结果的检验。(2)云是造成红外反演误差的主要原因之一,因此,在反演之前,本文做了云的检测处理;由于海洋和陆地在红外波段发射率差异较大,会对反演SST造成影响,因此,本文做了陆地水体分离处理。(3)对于海表温度反演软件的设计和实现,以VS 2010为集成环境,基于C#语言进行界面设计,结合ArcGIS和ENVI/IDL二次开发平台,建立海表温度反演软件。在设计软件的过程中,本文给出构建海表温度反演软件的关键技术、功能和界面的设计方案,并通过调用GIS组件来实现海表温度反演软件反演结果的制图输出。(4)对于反演结果的验证和分析,以我国南海海域作为验证区域范围来进行。以2014年2月4日标准时间0250的影像和2014年8月8日标准时间0245的影像为例,利用软件进行海表温度计算,处理结果表明,此结果可以真实地反映试验范围内温度的分布。用浮标测量值作为验证,结果显示,反演结果与实测数据的相关系数约为0.92,平均误差MRE为0.69℃,均方根误差RMSE为0.41℃,最小误差为0.039℃,最大误差为1.354℃。表明了本文软件采用的海表温度反演算法能够适用于我国海域,反演结果以期对海洋开发和探测提供一定的理论依据。
[Abstract]:The ocean covers a large area, which is about 70% of the global area. The temperature change of the ocean will affect the temperature change of the earth to a great extent, so it is very important to study the ocean surface temperature.The ocean contains enormous energy, the heat capacity of seawater is very large, the small change of its temperature will affect the living environment of human beings, at the same time, it will bring great changes to the global weather.At the same time, to a certain extent to the local weather changes.The process of sea surface temperature inversion algorithm is analyzed in this paper.In this paper, a series of researches on the inversion of sea surface temperature (SST) are carried out in the South China Sea region and in the South China Sea region, taking the MODISL1B remote sensing image data as the data source and the South China Sea region as the demonstration area, combined with the characteristics of MODIS.The specific contents are as follows: 1) taking Terra-MODIS data as the data source, and based on the existing results, through comparison and analysis, the sea surface temperature inversion algorithm, which is the most suitable for the region studied in this paper, is selected from the numerous inversion algorithms.Then, the cloud is one of the main reasons for the infrared inversion error by using the measured data covering the study area as the test of the inversion examples. Therefore, before inversion, the cloud detection and processing are done in this paper.Because the difference between ocean and land emissivity in infrared band will affect the inversion of SST, this paper designs and implements the software of sea surface temperature inversion, and takes vs 2010 as the integrated environment.Interface design based on C # language, combined with ArcGIS and ENVI/IDL secondary development platform, the establishment of sea surface temperature inversion software.In the process of designing the software, this paper gives the key technology, function and interface design scheme of building the sea surface temperature inversion software.The GIS module is used to realize the mapping output of the inversion result of sea surface temperature inversion software. The verification and analysis of the inversion results are carried out in the South China Sea as the verification area.Taking the images of 0250 standard time on February 4, 2014 and 0245 images of August 8, 2014 as examples, the software is used to calculate the sea surface temperature. The processing results show that the results can truly reflect the temperature distribution in the test range.The results show that the correlation coefficient between the inversion results and the measured data is about 0.92, the average error is 0.69 鈩，

本文编号：1772108