自動(dòng)多角度光譜觀測(cè)和遙感技術(shù)在碳源匯估算研究中的應(yīng)用
發(fā)布時(shí)間：2019-06-11 09:30
【摘要】：陸地生態(tài)系統(tǒng)碳循環(huán)是當(dāng)今全球變化研究的熱點(diǎn)問(wèn)題之一,遙感技術(shù)由于空間覆蓋廣并能快捷、非破壞性的對(duì)地表植被性質(zhì)進(jìn)行探測(cè),在人們探索碳循環(huán)過(guò)程中具有很好的應(yīng)用。鑒于此,本研究首先通過(guò)改進(jìn)遙感影像融合方法,為基于遙感的碳收支估算模型提供高時(shí)空分辨率地表反射率數(shù)據(jù)集;其次,在站點(diǎn)觀測(cè)水平上(站點(diǎn)尺度),利用多角度觀測(cè)數(shù)據(jù),計(jì)算冠層尺度光化學(xué)反射植被指數(shù)的算術(shù)平均值,評(píng)估光化學(xué)反射植被指數(shù)監(jiān)測(cè)光能利用率變化的能力,并分析影響光化學(xué)反射植被指數(shù)與光能利用率相關(guān)關(guān)系的外部(非生理)因素;然后利用通量塔觀測(cè)數(shù)據(jù)得到針對(duì)研究區(qū)不同土地覆被類(lèi)型的最大光能利用效率,進(jìn)而采用NCEP(National Centers for Environmental Prediction)氣象再分析資料和MODIS(Moderate-Resolution Imaging Spectroradiometer)光合有效輻射分量(fraction of Photosynthetically Active Radiation,fPAR)數(shù)據(jù),將站點(diǎn)觀測(cè)的總初級(jí)生產(chǎn)力(Gross Primary Productivity,GPP)擴(kuò)展到景觀尺度;最后,利用機(jī)器學(xué)習(xí)方法(回歸樹(shù)),footprint模型(Simple Analytical Footprint model on Eulerian coordinates,SAFE-f)和影像融合方法,結(jié)合通量塔觀測(cè)數(shù)據(jù)和遙感數(shù)據(jù),建立完全基于遙感數(shù)據(jù)的高分辨率凈生態(tài)系統(tǒng)碳交換量(Net Ecosystem Exchange of CO2)估算模型。得到了以下幾點(diǎn)主要結(jié)論:(1)優(yōu)化增強(qiáng)型時(shí)空自適應(yīng)反射率融合模型(Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model,ESTARFM),改進(jìn)后的影像融合算法預(yù)測(cè)的類(lèi)Landsat(Landsat-like)地表反射率具有更高的精度,所合成、預(yù)測(cè)的高時(shí)空分辨率的反射率數(shù)據(jù)可用于基于遙感數(shù)據(jù)的碳通量估算模型中。(2)利用自動(dòng)多角度光譜儀觀測(cè)的高頻率多角度反射率數(shù)據(jù),通量塔觀測(cè)數(shù)據(jù)分析表明:飽和水氣壓、淺層土壤溫度、總初級(jí)生產(chǎn)力、光合有效輻射對(duì)光化學(xué)反射植被指數(shù)和光能利用率的變化都有一定的影響。其中光合有效輻射對(duì)兩者的變化影響最大,分別為64%和22%;光化學(xué)反射植被指數(shù)和光能利用率之間具有較好的相關(guān)性,在不同的時(shí)間尺度上,半小時(shí)平均值決定系數(shù)(Coefficient of Determination,R2)為0.4084,日平均值決定系數(shù)R2為0.7349;在特定的環(huán)境因子條件下光化學(xué)反射植被指數(shù)和光能利用率的相關(guān)性表現(xiàn)會(huì)更好,光化學(xué)反射植被指數(shù)對(duì)探測(cè)脅迫狀態(tài)下的光能利用率的變化具有較好的敏感性,且敏感性隨著環(huán)境的改善,如飽和水汽壓VPD為20-25hPa、淺層土壤溫度為20-25℃、光合有效輻射為300-600 umol.m-2.s-1、總初級(jí)生產(chǎn)力40 umolCO2.m-2.s-1的條件下,兩者敏感性最佳。(3)采用基于通量塔觀測(cè)數(shù)據(jù)的光能利用率模型將站點(diǎn)尺度上的GPP上推到景觀尺度。其中,通過(guò)光能利用率模型反演得到研究區(qū)針葉林和闊葉林的最大光能利用率分別為0.8421 gCMJ-1、1.8082 gCMJ-1,決定系數(shù)R2分別為0.7000和0.8345。(4)根據(jù)渦度相關(guān)通量塔觀測(cè)數(shù)據(jù)、遙感時(shí)空融合得到的高空間分辨率高時(shí)間分辨率數(shù)據(jù),利用分類(lèi)回歸樹(shù)模型,采用機(jī)器學(xué)習(xí)技術(shù)構(gòu)建了高時(shí)空分辨率凈生態(tài)系統(tǒng)碳交換量的估算模型,估算的NEE結(jié)果較為合理。總體而言,本文利用遙感數(shù)據(jù)對(duì)陸地表層碳收支相關(guān)的光能利用率,GPP和NEE進(jìn)行了研究,通過(guò)研究生態(tài)系統(tǒng)碳通量和高光譜連續(xù)同步觀測(cè),評(píng)估了光化學(xué)反射指數(shù)監(jiān)測(cè)光能利用率變化的能力,并分析了影響光化學(xué)反射指數(shù)與光能利用率相關(guān)關(guān)系的外部(非生理)因素,利用影像融合方法提供了高時(shí)空分辨率的地表反射率數(shù)據(jù),為估算碳收支研究提供了輸入數(shù)據(jù),為建立高分辨率碳收支估算研究提供參考。
[Abstract]:The carbon cycle of terrestrial ecosystem is one of the hot issues in the research of global change, and the remote sensing technology has a good application in the process of exploring the carbon cycle because of the wide space coverage and the rapid and non-destructive detection of the surface vegetation. in view of that, the present study first provides a high time-space resolution surface reflectance data set for a remote sensing-based carbon revenue and expenditure estimation model by improving the remote sensing image fusion method, and secondly, using multi-angle observation data at the site observation level (site scale), calculating the arithmetic mean value of the canopy-scale photochemical reflectance vegetation index, evaluating the capability of the photochemical reflectance vegetation index to monitor the light energy utilization rate change, and analyzing the external (non-physiological) factors affecting the relation between the photochemically reflected vegetation index and the light energy utilization ratio; The maximum light energy utilization efficiency of different land cover types of the study area is obtained by using the flux column observation data, and then the data of the photosynthesis effective radiation component (fPAR) of the NCEP (National Centers for Environmental Prediction) meteorological re-analysis data and the MODIS (Modern-Resolution Imaging Spectroradiometer) are adopted, The total primary productivity (GPP) of the site observation is extended to the landscape scale; and finally, the machine learning method (regression tree), the footprint model (Simple Analytical Footprint model on Eulian coorates, SAFE-f) and the image fusion method are utilized to combine the flux tower observation data and the remote sensing data, A high-resolution net ecosystem exchange of CO2 model based on remote sensing data is established. The following main conclusions are obtained: (1) The enhanced spatial-temporal adaptive reflectance fusion model (ESTARFM) is optimized, and the modified Landsat-like surface reflectance of the modified image fusion algorithm has higher accuracy, The predicted high temporal resolution reflectivity data can be used in a carbon flux estimation model based on remote sensing data. (2) using the high-frequency multi-angle reflectivity data observed by the automatic multi-angle spectrometer, the data analysis of the flux tower shows that the saturated water pressure, the shallow soil temperature, the total primary productivity, Photosynthesis of effective radiation has a certain effect on the change of the photochemically reflected vegetation index and the light energy utilization rate. The effect of photosynthetically active radiation on the change of the two groups was 64% and 22%, respectively. There was a good correlation between the photochemically reflected vegetation index and the light energy utilization rate. On different time scales, the average of the half-hour mean determination coefficient (R2) was 0.4084. the daily average determination coefficient R2 is 0.749; the correlation performance of the photochemically reflected vegetation index and the light energy utilization rate under a specific environmental factor condition can be better, and the photochemical reflectance vegetation index has better sensitivity to the change of the light energy utilization rate in the detection stress state, And the sensitivity is optimal under the condition that the saturated water vapor pressure VPD is 20-25 hPa, the shallow soil temperature is 20-25 DEG C, the photosynthetic effective radiation is 300-600 umol. m-2.s-1 and the total primary productivity is 40 umolCO2. m-2.s-1. (3) using a light energy utilization model based on the flux-tower observation data to push the GPP on the site scale to the landscape scale. The maximum light energy utilization ratio of coniferous forest and broad-leaved forest in the study area was 0.8421gCMJ-1, 1.8082 gCMJ-1, and the coefficient of determination R2 was 0.7000 and 0.8345, respectively. (4) according to the observation data of the vorticity correlation flux tower, the high-space resolution high-time resolution data obtained by the space-time fusion of the remote sensing, the classification regression tree model is utilized, the machine learning technology is adopted to construct the estimation model of the carbon exchange amount of the high-space resolution net ecosystem, The estimated NEE results are more reasonable. In general, the light energy utilization rate, GPP and NEE related to the land surface carbon revenue and expenditure were studied by the remote sensing data, and the ability of the photochemical reflection index to monitor the light energy utilization rate was evaluated by studying the carbon flux and high-spectrum continuous synchronous observation of the ecosystem. The external (non-physiological) factors that affect the relation between the photochemical reflection index and the light energy utilization rate are analyzed, the surface reflectance data with high temporal and spatial resolution is provided by the image fusion method, and the input data is provided for estimating the carbon and expenditure research. And provides a reference for establishing a high-resolution carbon budget estimation study.
【學(xué)位授予單位】：中國(guó)礦業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：X171;X87
本文編號(hào)：2497086
[Abstract]:The carbon cycle of terrestrial ecosystem is one of the hot issues in the research of global change, and the remote sensing technology has a good application in the process of exploring the carbon cycle because of the wide space coverage and the rapid and non-destructive detection of the surface vegetation. in view of that, the present study first provides a high time-space resolution surface reflectance data set for a remote sensing-based carbon revenue and expenditure estimation model by improving the remote sensing image fusion method, and secondly, using multi-angle observation data at the site observation level (site scale), calculating the arithmetic mean value of the canopy-scale photochemical reflectance vegetation index, evaluating the capability of the photochemical reflectance vegetation index to monitor the light energy utilization rate change, and analyzing the external (non-physiological) factors affecting the relation between the photochemically reflected vegetation index and the light energy utilization ratio; The maximum light energy utilization efficiency of different land cover types of the study area is obtained by using the flux column observation data, and then the data of the photosynthesis effective radiation component (fPAR) of the NCEP (National Centers for Environmental Prediction) meteorological re-analysis data and the MODIS (Modern-Resolution Imaging Spectroradiometer) are adopted, The total primary productivity (GPP) of the site observation is extended to the landscape scale; and finally, the machine learning method (regression tree), the footprint model (Simple Analytical Footprint model on Eulian coorates, SAFE-f) and the image fusion method are utilized to combine the flux tower observation data and the remote sensing data, A high-resolution net ecosystem exchange of CO2 model based on remote sensing data is established. The following main conclusions are obtained: (1) The enhanced spatial-temporal adaptive reflectance fusion model (ESTARFM) is optimized, and the modified Landsat-like surface reflectance of the modified image fusion algorithm has higher accuracy, The predicted high temporal resolution reflectivity data can be used in a carbon flux estimation model based on remote sensing data. (2) using the high-frequency multi-angle reflectivity data observed by the automatic multi-angle spectrometer, the data analysis of the flux tower shows that the saturated water pressure, the shallow soil temperature, the total primary productivity, Photosynthesis of effective radiation has a certain effect on the change of the photochemically reflected vegetation index and the light energy utilization rate. The effect of photosynthetically active radiation on the change of the two groups was 64% and 22%, respectively. There was a good correlation between the photochemically reflected vegetation index and the light energy utilization rate. On different time scales, the average of the half-hour mean determination coefficient (R2) was 0.4084. the daily average determination coefficient R2 is 0.749; the correlation performance of the photochemically reflected vegetation index and the light energy utilization rate under a specific environmental factor condition can be better, and the photochemical reflectance vegetation index has better sensitivity to the change of the light energy utilization rate in the detection stress state, And the sensitivity is optimal under the condition that the saturated water vapor pressure VPD is 20-25 hPa, the shallow soil temperature is 20-25 DEG C, the photosynthetic effective radiation is 300-600 umol. m-2.s-1 and the total primary productivity is 40 umolCO2. m-2.s-1. (3) using a light energy utilization model based on the flux-tower observation data to push the GPP on the site scale to the landscape scale. The maximum light energy utilization ratio of coniferous forest and broad-leaved forest in the study area was 0.8421gCMJ-1, 1.8082 gCMJ-1, and the coefficient of determination R2 was 0.7000 and 0.8345, respectively. (4) according to the observation data of the vorticity correlation flux tower, the high-space resolution high-time resolution data obtained by the space-time fusion of the remote sensing, the classification regression tree model is utilized, the machine learning technology is adopted to construct the estimation model of the carbon exchange amount of the high-space resolution net ecosystem, The estimated NEE results are more reasonable. In general, the light energy utilization rate, GPP and NEE related to the land surface carbon revenue and expenditure were studied by the remote sensing data, and the ability of the photochemical reflection index to monitor the light energy utilization rate was evaluated by studying the carbon flux and high-spectrum continuous synchronous observation of the ecosystem. The external (non-physiological) factors that affect the relation between the photochemical reflection index and the light energy utilization rate are analyzed, the surface reflectance data with high temporal and spatial resolution is provided by the image fusion method, and the input data is provided for estimating the carbon and expenditure research. And provides a reference for establishing a high-resolution carbon budget estimation study.
【學(xué)位授予單位】：中國(guó)礦業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：X171;X87
【參考文獻(xiàn)】
相關(guān)期刊論文 前10條
1 于貴瑞;張雷明;孫曉敏;;中國(guó)陸地生態(tài)系統(tǒng)通量觀測(cè)研究網(wǎng)絡(luò)(ChinaFLUX)的主要進(jìn)展及發(fā)展展望[J];地理科學(xué)進(jìn)展;2014年07期
2 鄭騰飛;于鑫;包云軒;;多角度高光譜對(duì)光化學(xué)反射植被指數(shù)估算光能利用率的影響探究[J];熱帶氣象學(xué)報(bào);2014年03期
3 黃永喜;李曉松;吳炳方;董泰鋒;;基于改進(jìn)的ESTARFM數(shù)據(jù)融合方法研究[J];遙感技術(shù)與應(yīng)用;2013年05期
4 任小麗;何洪林;劉敏;張黎;周磊;于貴瑞;王輝民;;基于模型數(shù)據(jù)融合的千煙洲亞熱帶人工林碳水通量模擬[J];生態(tài)學(xué)報(bào);2012年23期
5 于貴瑞;方華軍;伏玉玲;王秋鳳;;區(qū)域尺度陸地生態(tài)系統(tǒng)碳收支及其循環(huán)過(guò)程研究進(jìn)展[J];生態(tài)學(xué)報(bào);2011年19期
6 于貴瑞;王秋鳳;朱先進(jìn);;區(qū)域尺度陸地生態(tài)系統(tǒng)碳收支評(píng)估方法及其不確定性[J];地理科學(xué)進(jìn)展;2011年01期
7 王興昌;王傳寬;于貴瑞;;基于全球渦度相關(guān)的森林碳交換的時(shí)空格局[J];中國(guó)科學(xué)(D輯:地球科學(xué));2008年09期
8 同小娟;李俊;王玲;;農(nóng)田光能利用效率研究進(jìn)展[J];生態(tài)學(xué)雜志;2008年06期
9 吳朝陽(yáng);牛錚;湯泉;;利用光化學(xué)植被指數(shù)估算葉片的光能利用率[J];蘭州大學(xué)學(xué)報(bào)(自然科學(xué)版);2008年02期
10 陳晉;唐艷鴻;陳學(xué)泓;楊偉;;利用光化學(xué)反射植被指數(shù)估算光能利用率研究的進(jìn)展[J];遙感學(xué)報(bào);2008年02期
相關(guān)碩士學(xué)位論文 前1條
1 胡淼淼;北京奧林匹克森林公園植物景觀與生態(tài)效益初探[D];北京林業(yè)大學(xué);2009年
，本文編號(hào)：2497086
本文鏈接：http://sikaile.net/shengtaihuanjingbaohulunwen/2497086.html
最近更新
教材專(zhuān)著
