本文關(guān)鍵詞：黑龍江大興安嶺地區(qū)森林枯落物含水率遙感反演　出處：《東北林業(yè)大學(xué)》2016年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： MODIS 森林枯落物 含水率 水分敏感波段

【摘要】：森林枯落物含水率是影響森林火災(zāi)發(fā)生的重要因素,新興發(fā)展的遙感技術(shù)為森林枯落物含水率預(yù)測(cè)提供了快捷、方便而又可靠的途徑。其中的MODIS (Moderate Resolution Imaging Spectroradiometer)數(shù)據(jù)由于其周期短、宏觀性強(qiáng)、穩(wěn)定、經(jīng)濟(jì)的特點(diǎn),以逐漸成為自然資源探查、災(zāi)害預(yù)測(cè)預(yù)報(bào)、國(guó)土統(tǒng)計(jì)調(diào)查、環(huán)境保護(hù)等研究的重要手段。本研究利用MODIS BRDF(Bi-Directional Reflectance Function)和LAI (Leaf AreaIndex)數(shù)據(jù)結(jié)合基于實(shí)測(cè)光譜建立的森林枯落物含水率的估算模型對(duì)大興安嶺地區(qū)森林枯落物含水率進(jìn)行反演。主要的研究結(jié)果如下：(1)利用烘干法實(shí)測(cè)森林枯落物含水率和SVCHR-1024i全波段地物光譜儀采集其光譜數(shù)據(jù),并用統(tǒng)計(jì)方法建立原始、一階導(dǎo)數(shù)和去包絡(luò)線三種不同形式光譜的枯落物含水率的模型,模型的決定系數(shù)(R2)、平均相對(duì)誤差(MRE),均方根誤差(RMSE)分別為：0.422、0.17、0.54；0.489、0.33、1.15；0.566、0.306、0.45。通過(guò)相關(guān)系數(shù)分析得出水分敏感波段為468nm～868nm、 998nm～1028nm、1158～1218nm、1318-2248nm、 2378～2488nm (呈現(xiàn)極顯著)418nm～468nm、878～988nm、1018nm～1068nm、 1228nm～1248nm、2258nm～2308nm (呈現(xiàn)顯著),且在1498nm和1508nm處出現(xiàn)極值分別為0.69和-0.69。(2)借助陳鏡明教授的4-scale模型,結(jié)合MODIS BRDF和LAI數(shù)據(jù)產(chǎn)品對(duì)大興安嶺地區(qū)森林枯落物的反演,結(jié)果表明：在建立查找表對(duì)多次散射因子(M)和四分量(KT,KZT,KG,KZG)的求解、在M求解中,分別對(duì)0。、25。、45。、60。這四個(gè)觀測(cè)角度的模型擬合進(jìn)行評(píng)價(jià)和篩選,最后確定45。和60。為最優(yōu)角度。在選取的MODISBRDF圖像5、6波段和多次散射因子(M)建立的模型R2分別為：0.577、0.583;0.95、0.947。在四分量求解中建立的模型R2分別為：0.769、0.867、0.937、0.879;0.822、0.83、0.917、0.82。在之后反演制圖中模型的平均絕對(duì)誤差(MAE)等于61.4%,均方根誤差(RMSE)等于696.6%。結(jié)果表明：借助MODIS數(shù)據(jù)反演森林枯落物含水率能夠提供準(zhǔn)確的連續(xù)性好的枯落物含水率數(shù)據(jù),對(duì)林火預(yù)測(cè)預(yù)報(bào)具有一定的意義。
[Abstract]:Forest litter moisture is an important factor affecting forest fire, remote sensing technology developing moisture prediction provides a quick way for forest litter, convenient and reliable. The MODIS (Moderate Resolution Imaging Spectroradiometer) data because of its short cycle, macroeconomic stability, strong economic characteristics, to gradually become the natural resources exploration, disaster forecast, land survey, an important means of environmental protection and so on. This study uses MODIS BRDF (Bi-Directional Reflectance Function) and LAI (Leaf AreaIndex) to establish the estimation model of the measured spectrum of forest litter moisture on a forest in Greater Khingan Range area litter moisture inversion based on data combined. The main results are as follows: (1) the use of drying method to measure forest litter moisture and SVCHR-1024i full band spectrometer collection The spectral data, and use statistical methods to establish the original, first derivative and envelope to three kinds of spectra of the moisture content of litter in the model, the determination coefficient of the model (R2), the average relative error (MRE), root mean square error (RMSE) were: 0.422,0.17,0.54; 0.489,0.33,1.15; 0.566,0.306,0.45. through the analysis of the correlation coefficient the moisture sensitive bands of 468nm ~ 868nm, 998nm ~ 1028nm, 1158 ~ 1218nm, 2378 ~ 2488nm (1318-2248nm, 418nm showed significant) ~ 468nm, 878 ~ 988nm, 1018nm ~ 1068nm, 1228nm ~ 1248nm, 2258nm ~ 2308nm (significant), and the 1498nm and 1508nm value were 0.69 and -0.69. (2) by using the 4-scale model of Professor Chen Jingming, MODIS BRDF and LAI inversion, combined with data on a forest litter in Greater Khingan Range area. The results showed that: in the lookup table is established on multiple scattering factor (M) and four component (KT, KZT, KG KZG), the solution in solving M, respectively for 0., 25., 45., 60. model fitting the four observation angle evaluation and screening, to determine the final 45. and 60. for the optimal angle. In the 5,6 band MODISBRDF image selection and multiple scattering factor (M) model developed by R2 were: 0.577,0.583 R2; model 0.95,0.947. in the four component solution established respectively: 0.769,0.867,0.937,0.879; 0.822,0.83,0.917,0.82. after model inversion in mapping the mean absolute error (MAE) equal to 61.4%, the root mean square error (RMSE) is equal to the result of 696.6%. showed that with the help of MODIS data inversion of forest litter moisture can provide accurate good continuity of litter moisture content data, which has a certain significance for forest fire forecast.

【學(xué)位授予單位】：東北林業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類號(hào)】：S762.1

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 金森;周勇;;昆明典型地表死可燃物含水率預(yù)測(cè)模型的研究[J];中南林業(yè)科技大學(xué)學(xué)報(bào);2014年12期

2 于宏洲;金森;邸雪穎;;以時(shí)為步長(zhǎng)的塔河林業(yè)局白樺林地表死可燃物含水率預(yù)測(cè)方法[J];林業(yè)科學(xué);2013年12期

3 林海軍;李霞;牛婷;趙釗;;不同地下水埋深條件下胡楊葉片光譜特征分析[J];干旱區(qū)資源與環(huán)境;2013年11期

4 劉昕;邸雪穎;;三種方法對(duì)森林地表可燃物含水率的預(yù)測(cè)評(píng)價(jià)[J];森林工程;2013年02期

5 楊曦光;于穎;黃海軍;范文義;;森林冠層氮含量遙感估算[J];紅外與毫米波學(xué)報(bào);2012年06期

6 陳紅艷;趙庚星;李希燦;朱西存;隋龍;王銀娟;;基于小波變換的土壤有機(jī)質(zhì)含量高光譜估測(cè)[J];應(yīng)用生態(tài)學(xué)報(bào);2011年11期

7 張歡;;基于SuperMap的森林火災(zāi)監(jiān)測(cè)系統(tǒng)的實(shí)現(xiàn)[J];森林工程;2011年05期

8 包玉龍;張繼權(quán);趙云升;佟志軍;劉興朋;;基于高光譜的草地可燃物含水率估測(cè)方法[J];紅外;2011年07期

9 張慧娟;李耀翔;張鴻富;張亞朝;李湃;;基于近紅外光譜不同波段的紅松木材含水率預(yù)測(cè)分析[J];東北林業(yè)大學(xué)學(xué)報(bào);2011年04期

10 張佳華;許云;姚鳳梅;王培娟;郭文娟;李莉;YANG LiMin;;植被含水量光學(xué)遙感估算方法研究進(jìn)展[J];中國(guó)科學(xué):技術(shù)科學(xué);2010年10期

，

本文編號(hào)：1405033