基于物候的水稻种植面积提取及时空分析

摘要水稻是我国主要的粮食作物之一，其种植面积的快速监测和自动提取对于我国粮食估产和保障国民粮食供给安全方面具有重要意义。采用遥感技术对植被的物候期进行分析研究，主要是指通过分析植被指数在时间序列上发生的明显变化，提取相应的关键时间节点及特征值。利用物候期的特点对水稻种植区域进行提取，估算面积并进行分析，对于估算黑龙江省水稻的种植面积，合理配置农业种植体系有很大的帮助。本文利用Landsat遥感影像数据，基于物候进行了水稻种植面积提取，对黑龙江省市2015年和2021年的水稻种植面积分别进行了提取，并对提取结果进行了分析。本文的主要工作及结论如下：(1)以Landsat数据为数据基础，对黑龙江省影像进行投影变换、大气校正、影像拼接、NDVI植被指数、地表水指数LSWI提取、影像裁剪等处理。(2)水稻关键物候期的提取：水稻不同生长发育期的生长特性不同，利用水稻生长过程中NDVI植被指数、地表水指数的分析，反演出水稻关键物候期模型，实现对水稻物候期的识别，并将识别结果与气象站点记录的水稻关键物候期进行对比分析，评价模型识别精度。(3)大范围水稻种植面积提取和时空变化分析：根据水稻不同生长发育期的生理特性在时间序列中所表现出来的光谱特征，比较其光谱特征与其他地物类型之间的差异，构建水稻信息提取模型，提取出水稻的种植区域。通过对比黑龙江省2015年和2021年的水稻面积分析其变化特征。关键词：水稻：黑龙江省；Landsat数据；面积提取；时空分析ABSTRACTRice is one of the main food crops in China,and the rapid monitoring and automatic extraction of its planting area is of great significance for estimating grain production and ensuring the security of national food supply in China.The use of remote sensing technology to analyze and study the phenology of vegetation mainly refers to the extraction of corresponding key time nodes and feature values by analyzing the significant changes in vegetation indices over time series.Extracting the rice planting area using the characteristics of phenology,estimating the area,and analyzing it are of great helpfor estimating the rice planting area in Heilongjiang Province and configuring the agricultural planting system reasonably.This article uses Landsat remote sensing image data to extract rice planting area based on phenology.The rice planting areas in Heilongjiang Province and cities in 2015 and 2021 were extracted,and the extraction resultswere analyzed.The main work and conclusions of this article are as follows:(1)Based on Landsat data,the image of Heilongjiang Province is processed by projection transformation,atmospheric correction,image stitching,NDVI vegetation index,LSWI extraction of surface water index,image clipping,etc.(2)Extraction of key phenological periods in rice:Different growth and developmentstages of rice have different growth characteristics.By analyzing the NDVI vegetationindex and surface water index during the rice growth process,a rice key phenologicalperiod model is inverted to achieve recognition of rice phenological periods.The recognition results are compared and analyzed with the key phenological periods recordedby meteorological stations to evaluate the recognition accuracy of the model.(3)Extraction of large-scale rice planting area and analysis of spatiotemporal changes:Based on the spectral characteristics exhibited by the physiological characteristics ofrice at different growth and development stages in time series,compare their spectralcharacteristics with other land types,construct a rice information extraction model,and extract the planting area of rice.By comparing the rice area in Heilongjiang Province in 2015 and 2021,we analyze its changing characteristics.Keywords:rice;Heilongjiang Province;Landsat data;area extraction,;spatiotemporalanalysis目录第1章绪论…1.1研究背景及意义1.1.1研究背景…11.1.2选题意义.11.2国内外研究