發(fā)布時間：2018-04-18 11:21

  本文選題：葉面積指數(shù) + 紅邊��； 參考：《中國科學(xué)院大學(xué)(中國科學(xué)院遙感與數(shù)字地球研究所)》2017年博士論文

【摘要】：葉面積指數(shù)(Leaf Area Index,LAI)是最重要的植被結(jié)構(gòu)參數(shù)之一,是作物長勢監(jiān)測、作物估產(chǎn)、肥水管理等精準(zhǔn)農(nóng)業(yè)必備的數(shù)據(jù)源。遙感技術(shù)為大面積、及時獲取LAI提供了有效手段。紅邊波段能夠用于研究植物養(yǎng)分及健康狀態(tài)監(jiān)測、植被識別和生理生化參數(shù)等信息,是定量遙感分析的理論基礎(chǔ)。利用不同遙感數(shù)據(jù)估測植被LAI各有其優(yōu)劣性,葉面積指數(shù)反演過程中需要充分挖掘包含紅邊波段的不同數(shù)據(jù)源的特點。例如,高光譜數(shù)據(jù)紅邊波段數(shù)量多、波段窄,但是存在波段間高度相關(guān)、數(shù)據(jù)冗余的問題;包含單個紅邊波段的多光譜數(shù)據(jù),紅邊波段較寬,比高光譜數(shù)據(jù)的紅邊波段缺少了許多細節(jié);包含多個紅邊波段的多光譜數(shù)據(jù),可以反映更多紅邊區(qū)域的光譜細節(jié),并且由于紅邊區(qū)域反射率迅速上升,紅邊區(qū)域內(nèi)的不同波段之間存在較大差別,在實際反演中需要進行合理選擇。本文針對不同遙感數(shù)據(jù)源的特點,圍繞紅邊波段進行葉面積指數(shù)反演研究,主要研究內(nèi)容及結(jié)論如下:(1)基于近地和航空高光譜數(shù)據(jù)紅邊波段的葉面積反演方法研究�；�于研究區(qū)域采集的近地、航空高光譜數(shù)據(jù)和田間同步試驗測量LAI數(shù)據(jù),探究航空和地面高光譜數(shù)據(jù)紅邊區(qū)域?qū)Χ�小麥LAI的反演能力。首先,建立高光譜植被指數(shù)反演模型,進而研究紅邊波段組合法和傳統(tǒng)波段組合、逐波段組合方法對植被指數(shù)反演LAI精度的影響,結(jié)果顯示在紅邊區(qū)域680-750nm波段范圍內(nèi),波段組合得到的植被指數(shù)與LAI的相關(guān)性非常高。最后,針對不同肥水條件下葉面積指數(shù)的特征光譜及參數(shù)隨不同試驗條件存在差異,本文基于航空和近地高光譜數(shù)據(jù),以及田間實測數(shù)據(jù),建立了基于高光譜植被指數(shù)MSAVI(Modified Soil-Adjusted Vegetation Index),NDVI(Normalized Difference Vegetation Index)和MTVI2(Modified Triangular Vegetation Index 2)的普適性強、精度高的冬小麥葉面積指數(shù)估算模型。(2)基于包含單個紅邊波段的多光譜衛(wèi)星數(shù)據(jù)反演作物葉面積指數(shù)方法研究。針對一般紅邊波段代替紅波段的改進植被指數(shù)多是基于單一時相、單一作物實現(xiàn)LAI估算中存在的對葉綠素含量的干擾因素考慮不足的缺陷,本文提出基于紅邊波段和紅波段進行組合改進的新植被指數(shù)ndviredre(red-edgenormalizeddifferencevegetationindex),msrredre(red-edgemodifiedsimpleratioindex)和ciredre(red-edgechlorophyllindex)。依據(jù)田間實測的不同生育時期的四種作物(小麥,大麥,苜蓿,玉米)的葉面積指數(shù)和與田間試驗準(zhǔn)同步的rapideye衛(wèi)星影像,建立基于植被指數(shù)的反演模型,結(jié)果證明本文提出的植被指數(shù)克服了在多時相和多種類型作物的情況下葉綠素含量的變化對lai反演的影響,有效提高了lai的反演精度,比一般紅邊波段代替紅波段的植被指數(shù)反演結(jié)果的決定系數(shù)提高至少10%。(3)基于包含多個紅邊波段的多光譜衛(wèi)星數(shù)據(jù)反演作物葉面積指數(shù)方法研究。面對包含多個紅邊波段的新發(fā)射多光譜衛(wèi)星在作物參數(shù)反演中的研究尚未成熟的情況,本文以搭載兩個紅邊波段的sentinel-2衛(wèi)星為例,針對不同紅邊波段之間光譜差異、多個紅邊波段的波段選擇等問題,采用三種葉面積指數(shù)反演的經(jīng)典方法:查找表、神經(jīng)網(wǎng)絡(luò)和植被指數(shù)法,建立冬小麥葉面積指數(shù)反演模型。作為對比,同時利用不包含紅邊波段的landsat8衛(wèi)星數(shù)據(jù)反演,由衛(wèi)星數(shù)據(jù)、農(nóng)學(xué)信息、地面實測等多元數(shù)據(jù),反演了北京順義區(qū)部分樣點的冬小麥葉面積指數(shù)。結(jié)果表明,具有更高的“時-空-譜”分辨率的sentinel-2衛(wèi)星,比landsat8衛(wèi)星反演精度更高。sentinel-2衛(wèi)星搭載的中心波長為705nm和740nm的兩個紅邊波段,比單個紅邊波段的多光譜數(shù)據(jù)(如rapideye)提供了更豐富的紅邊區(qū)域波譜信息,以及更多與lai高度相關(guān)的基于705nm和750nm的植被指數(shù)的選擇。本研究可以為搭載多個紅邊波段的多光譜衛(wèi)星數(shù)據(jù)在植被定量遙感中的應(yīng)用提供理論依據(jù)。本文的研究結(jié)論表明高光譜數(shù)據(jù)紅邊區(qū)域680-750nm波段范圍內(nèi),植被指數(shù)與lai的相關(guān)性非常高;基于包含單個紅邊波段的多光譜衛(wèi)星數(shù)據(jù)可以通過結(jié)合紅邊波段的改進植被指數(shù),來抑制葉綠素含量等因素的影響,提高lai的反演精度;多個紅邊波段的多光譜衛(wèi)星數(shù)據(jù)的紅邊波段之間反射率差異顯著,提供了比單波段多光譜數(shù)據(jù)更加豐富的紅邊波段信息,有利于豐富植被指數(shù)類型選擇和lai反演模型精度的提升。以上結(jié)論可以為高光譜數(shù)據(jù)、包含一個或多個紅邊波段的多光譜數(shù)據(jù)在作物葉面積指數(shù)反演中的應(yīng)用提供理論依據(jù),為作物生長狀態(tài)監(jiān)測、農(nóng)田管理決策提供可靠的參考信息。同時本文證明了sentinel-2衛(wèi)星搭載的兩個中心波長分別為705nm和740nm的紅邊波段,在葉面積指數(shù)反演中具有重要的應(yīng)用價值,可以為多光譜傳感器的波段設(shè)計提供參考依據(jù)。
[Abstract]:Leaf area index (Leaf Area, Index, LAI) is one of the most important parameters of vegetation, is the crop growth monitoring, crop yield, fertilizer and water management and precision agriculture the necessary data source. Remote sensing technology for large area, timely access to LAI provides effective means. Red edge band can be used to study plant nutrient and health monitoring vegetation, identification and physiological and biochemical parameters and other information, is the theoretical basis for quantitative remote sensing analysis. Using different remote sensing data to estimate vegetation LAI each has its advantages and disadvantages, the need to fully tap the characteristics of different data sources including red edge band inversion of leaf area index in the process. For example, the bands of hyperspectral data of red edge number, narrow band however, there is a high correlation between bands, the data redundancy problem; multi spectral data contains a single red edge band, red edge band is wide, the red edge wavelength of spectral data is missing a lot of details included; Multi spectral data of a plurality of red edge spectral band, can reflect more details of the red edge region, and the reflectance of red edge area increased rapidly, there is a big difference between different bands of red edge region, in the actual retrieval needs reasonable choice. According to the characteristics of different remote sensing data sources, Research on leaf area index back around the red edge band, the main research contents and conclusions are as follows: (1) study on leaf area and ground inversion method of Airborne Hyperspectral Data Based on red edge band. The study area near acquisition based on Airborne Hyperspectral Data and field test data synchronous measurement of LAI, on air and ground hyperspectral data inversion of red edge area winter wheat LAI. First of all, the establishment of Hyperspectral Vegetation Index inversion model, and studies on the red edge band combination method and traditional band combination, by wavelength combination method on vegetation index Effect of LAI inversion accuracy of the results are displayed in red edge region 680-750nm wavelength range, the correlation between vegetation index band combination with LAI obtained is very high. Finally, according to the spectral characteristics and parameters of leaf area index of different water and fertilizer conditions with different test conditions are different, the air and hyperspectral data based on field test and data, established Hyperspectral Vegetation Index Based on MSAVI (Modified Soil-Adjusted Vegetation Index), NDVI (Normalized Difference Vegetation Index) and MTVI2 (Modified Triangular Vegetation Index 2) of the universal model for estimating winter wheat leaf area index with high accuracy. (2) study on the index method of multi spectral satellite data containing a single crop the red edge band. Leaf area based on improved vegetation index for red edge band instead of the red band is a single phase based on single crop Considering the deficiencies of interference on the chlorophyll content factors of LAI estimation, this paper proposes a new vegetation index ndviredre modified red edge band and red band (based on red-edgenormalizeddifferencevegetationindex), msrredre (red-edgemodifiedsimpleratioindex) and ciredre (red-edgechlorophyllindex). On the basis of the four kinds of crop field measured in different growth stages (wheat, barley, alfalfa, corn) leaf area index and rapideye satellite image synchronization and field experiment, establish the inversion model of vegetation index based on vegetation index results presented in the paper overcomes the impact of changes in multitemporal and various types of crop chlorophyll content under the inversion of Lai, effectively improve the retrieval precision of Lai. The coefficient of determination of vegetation index inversion results than the general red edge band instead of the red band increased by at least 10%. (3) of the index method of multispectral satellite data includes a plurality of crop leaf area based on red edge band. Facing the new launch of multi spectral satellite contains more than one red edge band on crop in parameter inversion is not yet mature, this is equipped with two red bands of the sentinel-2 satellite as an example, according to the the spectral differences between red edge band, a plurality of red edge band band selection problem, the classical method using three kinds of inversion of leaf area index: look-up table, neural network and vegetation index method, established the winter wheat leaf area index inversion model. In contrast, at the same time using the landsat8 satellite data inversion does not contain red edge band from satellite data, information, agriculture, the measured multivariate data of Winter Wheat in Beijing, Shunyi District and some samples of leaf area index inversion. The results show that the higher the space-time and spectral resolution of sent The inel-2 satellite, the center wavelength accuracy of landsat8 satellite is higher than the.Sentinel-2 satellite for two red bands 705nm and 740nm, compared with multispectral data for a single red edge band (such as rapideye) with red edge area spectrum information more abundant, and more highly correlated with Lai and 705nm based on vegetation index the choice of 750nm. This study can provide a theoretical basis for multi spectral satellite data with a plurality of red edge band in the quantitative application of remote sensing in vegetation. The conclusion of this paper shows that the high spectral data in red edge area 680-750nm wavelength range, the correlation between vegetation index and Lai are very high; multi spectral satellite data contains a single red edge the band through a combination of improved vegetation index based on red edge band, to suppress the influence factors such as chlorophyll content, improve the retrieval precision of Lai; a plurality of red edge band multi spectral satellite number The difference between the reflectance of red edge band according to the significant, provides more information than the red edge band single band multi spectral data, enrich the types of vegetation index and Lai inversion accuracy of the model. The above conclusions can enhance the hyperspectral data, and provide a theoretical basis for the application of multi spectral data contains one or more red edge the band in the crop leaf area index inversion, for crop growth condition monitoring, to provide a reliable reference information for decision making in farmland management. At the same time this paper proves that two wavelength sentinel-2 satellite were red edge band 705nm and 740nm, which has important application value in the inversion of leaf area index, can provide a reference as the band design of multispectral sensors.

【學(xué)位授予單位】：中國科學(xué)院大學(xué)(中國科學(xué)院遙感與數(shù)字地球研究所)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2017
【分類號】：S127

【相似文獻】

相關(guān)期刊論文 前10條

1 張瑞忠;田嵐;鄭家蘭;;大豆葉面積指數(shù)與豐產(chǎn)性能關(guān)系的初步研究[J];東北農(nóng)學(xué)院學(xué)報;1962年03期

2 J.G.布哈特;棉花不同形態(tài)品種葉片生長與葉面積指數(shù)的研究[J];湖北農(nóng)業(yè)科學(xué);1975年08期

3 莫家讓;;水稻葉面積指數(shù)簡易測定法[J];農(nóng)業(yè)科技通訊;1975年09期

4 趙福;;葉面積和葉面積指數(shù)的計算[J];內(nèi)蒙古農(nóng)業(yè)科技;1975年07期

5 ;怎樣用鮮重法測定小麥的葉面積指數(shù)[J];甘肅農(nóng)業(yè)科技;1976年04期

6 鄭春和;;怎樣用鮮重法測定小麥的葉面積指數(shù)[J];植物學(xué)雜志;1976年01期

7 ;介紹一種小麥生育后期和玉米不拔取樣株測定葉面積指數(shù)的方法[J];北京師范大學(xué)學(xué)報(自然科學(xué)版);1977年02期

8 沈瑞庭;李錫坤;;介}9幾種葉面積的n,定方法[J];蠶桑茶葉通訊;1980年01期

9 胡耀華;伍業(yè)綱;王釗;舒宜通;;關(guān)于巴西橡膠樹林葉面積指數(shù)的研究[J];熱帶作物學(xué)報;1981年02期

10 W·G·貝利;R·B·斯圖爾特;朱胤椿;;葉面積的測定方法[J];青海農(nóng)林科技;1982年03期

相關(guān)會議論文 前10條

1 姚冬萍;;融合多源遙感數(shù)據(jù)估算地表葉面積指數(shù)[A];中國地理學(xué)會百年慶典學(xué)術(shù)論文摘要集[C];2009年

2 陳雪洋;蒙繼華;吳炳方;朱建軍;紐立明;杜鑫;;基于環(huán)境星高光譜數(shù)據(jù)紅邊參數(shù)的冬小麥葉面積指數(shù)反演[A];遙感定量反演算法研討會摘要集[C];2010年

3 高帥;;利用ENVISAT-ASAR數(shù)據(jù)的植被葉面積指數(shù)反演研究[A];遙感定量反演算法研討會摘要集[C];2010年

4 張娜;;半干旱天然草地葉面積指數(shù)反演模型[A];第十五屆全國遙感技術(shù)學(xué)術(shù)交流會論文摘要集[C];2005年

5 孫永華;宮輝力;李小娟;浦瑞良;周德民;;基于高光譜的三江平原濕地葉面積指數(shù)反演研究[A];遙感定量反演算法研討會摘要集[C];2010年

6 曹中盛;朱艷;田永超;曹衛(wèi)星;姚霞;;小麥葉面積指數(shù)估測的最佳高光譜參數(shù)研究[A];中國作物學(xué)會2013年學(xué)術(shù)年會論文摘要集[C];2013年

7 呂蕓帙;楊旭;駱祖瑩;;基于無線傳感器網(wǎng)絡(luò)的植被葉面積指數(shù)測量方法[A];第六屆中國測試學(xué)術(shù)會議論文集[C];2010年

8 夏帆;;基于MODIS數(shù)據(jù)的長江流域生態(tài)環(huán)境質(zhì)量評價[A];第十五屆全國遙感技術(shù)學(xué)術(shù)交流會論文摘要集[C];2005年

9 劉倩;;2000-2010年三江源地區(qū)MODIS葉面積指數(shù)及FPAR產(chǎn)品驗證分析[A];遙感定量反演算法研討會摘要集[C];2010年

10 楊鵬;李春強;;基于MODIS數(shù)據(jù)的河北省冬小麥葉面積指數(shù)模型研究[A];第31屆中國氣象學(xué)會年會S5 干旱災(zāi)害風(fēng)險評估與防控[C];2014年

相關(guān)重要報紙文章 前1條

1 李濟 供稿;理科綜合測試模擬（四）[N];山西科技報;2003年

相關(guān)博士學(xué)位論文 前5條

1 樂源;基于光譜數(shù)據(jù)的植被重點參數(shù)反演方法研究[D];武漢大學(xué);2014年

2 溫一博;基于遙感技術(shù)的森林碳循環(huán)和地表蒸散模擬研究[D];東北林業(yè)大學(xué);2016年

3 謝巧云;考慮紅邊特性的多平臺遙感數(shù)據(jù)葉面積指數(shù)反演方法研究[D];中國科學(xué)院大學(xué)(中國科學(xué)院遙感與數(shù)字地球研究所);2017年

4 駱社周;激光雷達遙感森林葉面積指數(shù)提取方法研究與應(yīng)用[D];中國地質(zhì)大學(xué)(北京);2012年

5 郝雅s，

本文編號：1768180