基于R语言的MODIS数据自动化处理方法研究

A R-language Based on Approach to Automated Process MODIS Data

卫星遥感数据具有覆盖空间范围广、观测周期短等优点,已经成为全球变化监测领域不可或缺的数据源。尤其是具有较高时空分辨率的MODIS系列数据产品,自发布以来已经广泛用于陆表过程模拟和植被动态变化监测等研究。本文介绍了一套利用R语言自动化进行数据下载,处理任意地点、时间段MODIS数据产品的流程方法,并以MODIS NDVI数据产品为例展示了具体的实现机制和过程。首先使用http协议发起数据请求,并使用二进制流方式下载数据,得到hdf格式数据以后利用R语言连接MRT(MODIS Reprojection Tool)工具,进行高效的遥感产品的格式转换、投影转换、拼接工作,最后利用R语言对影像进行时间趋势分析和空间探索分析。结果证明:利用R语言的网络数据抓取、传输和可执行程序黏合等灵活的功能,可以减少MODIS数据处理过程中的人为误差和研究人员手动处理数据的工作量。自动化的MODIS NDVI处理流程有助于生态学的研究,特别是在meta分析、观测网络和全球性的分布式实验中的应用。

Satellite remote sensing data have been a useful data source for global change studies due to its largely covered extent and highly observed cycle.MODIS data are widely used in land process modeling and vegetation observation studies.This paper introduces a new approach to automated download,process MODIS data at any location and period.MODIS NDVI data processing is described as an example.Firstly,http protocol is applied to request the data.then binary flow is obtained from website.The result is HDF files that can be processed using MRT tool(MODIS Reprojection Tool).MRT can been used to do data format transform,reprojection and merging tiles.Finally,R language can help to analysis time series trend and spatial pattern.Our experiment shows that R language is an effective tool to download and process MODIS data that can avoid the errors happening when these works are done manually.The automated workflow of MODIS NDVI data process is helpful in ecological studies,such as meta analysis,observation network and distributed experiments.