發(fā)布時(shí)間：2018-03-21 17:14

  本文選題：尺度效應(yīng)　切入點(diǎn)：尺度轉(zhuǎn)換　出處：《中國(guó)地質(zhì)大學(xué)(北京)》2017年博士論文　論文類型：學(xué)位論文

【摘要】：遙感數(shù)據(jù)與作物生長(zhǎng)模型在農(nóng)作物估產(chǎn)上能夠優(yōu)勢(shì)互補(bǔ)。然而,作物模型尺度與遙感觀測(cè)尺度不匹配是影響同化模型精度的重要因素,這將極大增加遙感反演和數(shù)據(jù)同化的不確定性。本研究以基于作物生長(zhǎng)模型與遙感數(shù)據(jù)同化的冬小麥估產(chǎn)為目標(biāo),針對(duì)遙感尺度這一重要科學(xué)問題,借助統(tǒng)計(jì)模型、物理方法、數(shù)據(jù)融合等技術(shù),剖析了同化觀測(cè)量(葉面積指數(shù),LAI)尺度效應(yīng)產(chǎn)生的主要原因,定量描述并校正了空間異質(zhì)性與模型非線性所引起的尺度效應(yīng)誤差,探討了多源遙感數(shù)據(jù)差異及其形成的內(nèi)在機(jī)理,研究了Landsat與MODIS之間的尺度轉(zhuǎn)換方法,建立了空間尺度差異校正的基本框架,并面向研究區(qū)對(duì)尺度校正框架進(jìn)行了適用性驗(yàn)證,利用數(shù)據(jù)融合技術(shù)對(duì)同化觀測(cè)量進(jìn)行了時(shí)間尺度擴(kuò)展,進(jìn)而耦合WOFOST模型與多時(shí)空尺度遙感數(shù)據(jù)對(duì)河北省衡水地區(qū)冬小麥產(chǎn)量進(jìn)行估測(cè)。論文的研究工作及主要結(jié)論如下:(1)在分析尺度效應(yīng)產(chǎn)生根源、明確多源遙感數(shù)據(jù)差異的基礎(chǔ)上,定量分析多空間尺度遙感反演農(nóng)作物同化觀測(cè)量LAI的總體差異。結(jié)果表明,多源遙感數(shù)據(jù)引起的差異高于尺度效應(yīng)帶來(lái)的誤差。(2)考慮同化觀測(cè)量LAI的空間異質(zhì)性,細(xì)化尺度效應(yīng)產(chǎn)生過(guò)程,結(jié)合小波變換和分形理論,定量分析冬小麥LAI不同反演過(guò)程對(duì)尺度效應(yīng)的貢獻(xiàn),并有效校正尺度效應(yīng)引起的誤差。(3)從系統(tǒng)內(nèi)在機(jī)理出發(fā),歸納分析Landsat和MODIS數(shù)據(jù)差異,提取可通過(guò)數(shù)理方法模擬的相關(guān)信息,基于點(diǎn)擴(kuò)散函數(shù)及粒子群優(yōu)化算法,定量校正遙感觀測(cè)數(shù)據(jù)差異,建立尺度轉(zhuǎn)換模型,降低尺度差異導(dǎo)致的不確定性。(4)基于多尺度遙感數(shù)據(jù)反演同化觀測(cè)量差異的定量分析,借助各類數(shù)理模型,構(gòu)建并完善空間尺度差異校正框架,將多尺度遙感定量反演同化觀測(cè)量LAI的總體不確定性降低了50%以上。(5)在空間尺度差異校正和時(shí)間尺度擴(kuò)展的基礎(chǔ)上,通過(guò)四維變分同化算法及SCE-UA優(yōu)化算法,耦合多尺度遙感信息與WOFOST作物生長(zhǎng)模型,對(duì)冬小麥進(jìn)行區(qū)域化時(shí)空同化估產(chǎn),在保證同化精度的前提下,大幅提高同化效率。
[Abstract]:The remote sensing data and crop growth model can be complementary in the estimation of crop yield. However, crop model scale and remote sensing observation scale, is an important factor affecting the assimilation model precision, which will greatly increase the retrieval and data assimilation uncertainty. Based on the yield of winter wheat crop growth model and remote sensing data assimilation based on the target in view of this, the scale of remote sensing of important scientific problems, physical methods by means of statistical model, data fusion technology, analyzes the assimilation measurements (leaf area index, LAI) mainly due to the scale effects, quantitative description of the error correction and scale effect of spatial heterogeneity and nonlinear model caused by the inherent mechanism of the difference of multi-source remote sensing data and the formation of the research on the transformation method between Landsat and MODIS scale, establish the basic framework of correction in different spatial scales, and surface To study on scale correction framework for the validation of measurements, the assimilation time scale expansion of the use of data fusion technology, and the coupling of WOFOST model and multi-scale remote sensing data on the yield of Winter Wheat in Hengshui area of Hebei province were estimated. The research work of this thesis and the main conclusions are as follows: (1) in the analysis of causes the scale effect, based on the difference of multi-source remote sensing data, the overall differences in quantitative analysis of multi spatial scale remote sensing crop assimilation measurements of LAI. The results show that the difference of multi-source remote sensing data caused by the above error scale effects. (2) considering the spatial heterogeneity of assimilation measurements of LAI, produced in the process of refining the scale effect, combined with wavelet transform and fractal theory, quantitative analysis of winter wheat LAI inversion process different contribution to the scale effect, scale effect and effective error correction caused by (3) from the Department. The system of internal mechanism, analyzed the difference between Landsat and MODIS data, the relevant information can be extracted by mathematical simulation method, point spread function and particle swarm optimization algorithm based on the difference of quantitative calibration of remote sensing observation data, a scale transformation model, reduce the scale difference leads to uncertainty. (4) quantitative analysis of differences of multi-scale remote sensing data based on the concept of retrieval and assimilation, with all kinds of mathematical model, construct and perfect the spatial scale difference correction framework of multi-scale quantitative remote sensing measurements of LAI assimilation overall uncertainty reduced by 50%. (5) based in correction in different spatial scales and time scales, the four-dimensional Variational Assimilation Algorithm and SCE-UA optimization algorithm, coupled multi-scale remote sensing information and crop growth model WOFOST, regional spatial assimilation yield of winter wheat, under the premise of ensuring the accuracy of assimilation, sharp Improve the efficiency of assimilation.

【學(xué)位授予單位】：中國(guó)地質(zhì)大學(xué)(北京)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2017
【分類號(hào)】：S512.11;S127

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 代璇;蘇弋;;景觀園林設(shè)計(jì)中空間尺度的應(yīng)用解析[J];現(xiàn)代園藝;2012年18期

2 劉翔;鄒志榮;;園林景觀空間尺度的視覺性量化控制[J];安徽農(nóng)業(yè)科學(xué);2008年07期

3 滕s，

本文編號(hào)：1644792