本文選題：葉面積指數(shù) + MODIS��； 參考：《南京信息工程大學(xué)》2015年碩士論文

【摘要】：由于全球氣候變化、碳源匯變化驅(qū)動機(jī)制等研究的需要,葉面積指數(shù)(Leaf Area Index, LAI)作為氣候模式、碳循環(huán)模式等動態(tài)過程模型的關(guān)鍵輸入,LAI的研究和應(yīng)用更應(yīng)需要在區(qū)域及全球尺度上進(jìn)行,因此利用遙感衛(wèi)星數(shù)據(jù)進(jìn)行LAI反演,進(jìn)而生產(chǎn)全球遙感LAI產(chǎn)品數(shù)據(jù)得到越來越廣泛應(yīng)用。然而,由于不同的測量方法和儀器及冠層結(jié)構(gòu)的不同因素定義葉面積指數(shù),LAI產(chǎn)品能顯著變化,目前還不具有準(zhǔn)確性和一致性的全球和區(qū)域應(yīng)用產(chǎn)品。因此在應(yīng)用遙感LAI產(chǎn)品時,對LAI產(chǎn)品的精度評價顯得尤為重要。本研究以中國東北大興安嶺加格達(dá)奇林區(qū)、江蘇省南京市典型農(nóng)田為研究區(qū)�；诃h(huán)境衛(wèi)星遙感數(shù)據(jù)獲得典型植被指數(shù),利用環(huán)境衛(wèi)星植被指數(shù)和實測LAI構(gòu)建回歸分析模型分別反演林地、小麥和水稻LAI。最后,利用環(huán)境衛(wèi)星反演LAI數(shù)據(jù)通過尺度轉(zhuǎn)換對MODIS LAI產(chǎn)品和GLOBCARBON LAI產(chǎn)品進(jìn)行驗證,并分析LAI產(chǎn)品誤差來源,再對農(nóng)田區(qū)MODIS LAI產(chǎn)品和GLOBCARBON LAI產(chǎn)品進(jìn)行訂正。本研究的主要結(jié)論如下：(1)在中國東北大興安嶺加格達(dá)奇林區(qū),三種LAI數(shù)據(jù)在植被區(qū)域LAI值域范圍最大的為GLOBCARBON LAI數(shù)據(jù),其值在0.93～4.91, HJ-1kmLAI數(shù)據(jù)與MODIS LAI數(shù)據(jù)值域基本相同,GLOBCARBON LAI均值最高,其值比HJ-1kmLAI高0.29,誤差為11%,而MODIS LAI數(shù)據(jù)均值則比HJ-1kmLAI均值低0.28,誤差為11.8%,兩種遙感LAI數(shù)據(jù)產(chǎn)品在研究區(qū)精度誤差均在20%左右,但GLOBCARBON LAI存在高估現(xiàn)象,而MODIS LAI數(shù)據(jù)則為低于實測反演值。(2)在南京農(nóng)作物研究區(qū),小麥的GLOBCARBON LAI比HJ-30mLAI均值低1.18,誤差為44%,MODIS LAI比HJ-30mLAI均值低1.75,誤差為66%；在水稻區(qū),GLOBCARBON LAI比HJ-30mLAI均值低0.84,誤差為25%, MODISLAI比HJ-30mLAI均值低1.47,誤差為43%。通過結(jié)果分析可以看出,研究區(qū)小麥和水稻MODIS LAI、GLOBCARBON LAI的均值明顯低于環(huán)境衛(wèi)星反演得到的LAI值,存在嚴(yán)重低估現(xiàn)象。根據(jù)分析,由于南京農(nóng)田呈零星狀分布,地表異質(zhì)性嚴(yán)重,導(dǎo)致低分辨率GLOBCARBON LAI和MODIS LAI產(chǎn)品存在混合像元。(3)再對南京農(nóng)作物研究區(qū)GLOBCARBON LAI和MODIS LAI產(chǎn)品混合像元進(jìn)行分解,得出在小麥區(qū),GLOBCARBON LAI均值高估HJ-30mLAI為0.25,誤差從44%降到8.6%,MODIS LAI比HJ-30mLAI均值低0.29,誤差從66%降到10.9%；在水稻區(qū),GLOBCARBON LAI比HJ-30mLAI均值高0.28,誤差從25%降到7.6%,MODIS LAI比HJ-30mLAI均值低0.23,誤差從43%降到6.7%。從訂正后數(shù)據(jù)來看,訂正后的MODIS LAI和GLOBCARBON LAI極大的改善了混合像元的問題。但GLOBCARBON LAI存在高估現(xiàn)象,而MODIS LAI數(shù)據(jù)則為低于實測反演值。
[Abstract]:The Leaf Area Index (LAI) is the key input of the dynamic process model such as the climate model and the carbon cycle model as a result of the global climate change and the driving mechanism of carbon source and sink change. The research and application of LAI should be carried out on the regional and global scales. Therefore, the remote sensing satellite data is used for the LAI inversion, and then the remote sensing satellite data is used to inverse the LAI. The production of global remote sensing LAI product data is becoming more and more widely used. However, because of the different measurement methods and instruments and the definition of the leaf area index of the different factors of the canopy structure, the LAI product can change significantly. At present, the accuracy and consistency of the global and regional application products are not yet accurate. Therefore, when using remote sensing LAI products, the LAI products are applied to the products. This study takes the Jiagedaqi forest area of Greater Khingan Range in Northeast China and the typical farmland of Nanjing city of Jiangsu Province as the research area. Based on the environmental satellite remote sensing data, the typical vegetation index is obtained. The regression analysis model of the environmental satellite vegetation index and the measured LAI is used to reconstruct the woodland, and the wheat and rice LAI. are last, The LAI data of MODIS LAI products and GLOBCARBON LAI products are verified by scale conversion, and the error sources of LAI products are analyzed, and the MODIS LAI products and GLOBCARBON LAI products in farmland are revised. The main conclusions of this study are as follows: (1) three LAI data in the Jiagedaqi forest area of Greater Khingan Range, Northeast China. The maximum range of LAI range in the vegetation area is GLOBCARBON LAI data, its value is 0.93 ~ 4.91, HJ-1kmLAI data and MODIS LAI data range are basically the same, the mean value of GLOBCARBON LAI is highest, its value is 0.29 higher than HJ-1kmLAI, and the error is 11%, while the MODIS LAI data mean is 0.28 lower than the HJ-1kmLAI mean, and the error is 11.8%, two kinds of remote sensing data products. The accuracy error in the study area is around 20%, but the GLOBCARBON LAI is overestimated, while the MODIS LAI data is lower than the measured inversion value. (2) in the Nanjing crop research area, the GLOBCARBON LAI of wheat is 1.18 lower than the HJ-30mLAI mean, the error is 44%, MODIS LAI is 1.75 lower than HJ-30mLAI, and the error is 66%. In the rice region, GLOBCARBON LAI is compared to those in the rice region. The mean value of 0mLAI is 0.84, the error is 25%, and the MODISLAI is 1.47 lower than the HJ-30mLAI mean. The error is 43%. through the result analysis. The mean value of MODIS LAI and GLOBCARBON LAI in the study area is obviously lower than the LAI value obtained by the environmental satellite, and there is a serious underestimation. According to the analysis, the farmland is scattered in Nanjing and the ground surface is different. The quality is serious, resulting in a mixed pixel of low resolution GLOBCARBON LAI and MODIS LAI products. (3) then the mixed pixel of GLOBCARBON LAI and MODIS LAI products in the Nanjing crop research area is decomposed, and the average value of GLOBCARBON LAI is 0.25, the error is reduced from 44% to 8.6% in the wheat region, and the error is 0.29 lower than that of the average. From 66% to 10.9%, in the rice area, GLOBCARBON LAI is 0.28 higher than that of HJ-30mLAI, the error is reduced from 25% to 7.6%, MODIS LAI is 0.23 lower than the mean of HJ-30mLAI, and the error is reduced from 43% to 6.7%. from the revised data. The revised MODIS LAI and GLOBCARBON LAI greatly improve the problem of the mixed image element. The IS LAI data is lower than the measured inversion value.
【學(xué)位授予單位】：南京信息工程大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：S771.8;S127

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 劉小平,鄧孺孺,彭曉鵑;城市綠地遙感信息自動提取研究——以廣州市為例[J];地域研究與開發(fā);2005年05期

2 馬俊飛;呂昌河;王茜;;基于MODIS NDVI時序數(shù)列的柴達(dá)木盆地土地覆蓋分類研究[J];干旱區(qū)地理;2008年03期

3 方秀琴,張萬昌;葉面積指數(shù)(LAI)的遙感定量方法綜述[J];國土資源遙感;2003年03期

4 劉洋;劉榮高;陳鏡明;程曉;鄭光;;葉面積指數(shù)遙感反演研究進(jìn)展與展望[J];地球信息科學(xué)學(xué)報;2013年05期

5 李石華;角媛梅;;環(huán)境與災(zāi)害監(jiān)測預(yù)報小衛(wèi)星A星CCD影像質(zhì)量評價[J];紅外技術(shù);2009年03期

6 譚一波;趙仲輝;;葉面積指數(shù)的主要測定方法[J];林業(yè)調(diào)查規(guī)劃;2008年03期

7 ;New Vegetation Index and Its Application in Estimating Leaf Area Index of Rice[J];Rice Science;2007年03期

8 王希群,馬履一,賈忠奎,徐程揚(yáng);葉面積指數(shù)的研究和應(yīng)用進(jìn)展[J];生態(tài)學(xué)雜志;2005年05期

，

本文編號：1954400