【摘要】：本文以渤黃東海為研究區(qū)域，以2003-2012年渤黃東�，F(xiàn)場數(shù)據(jù)和NOMAD數(shù)據(jù)集為基礎(chǔ)，深入地分析與探討了渤黃東海渾濁水體大氣校正模型、系數(shù)與散射系數(shù)反演模型、葉綠素a和懸浮物濃度反演模型、以及漫衰減系數(shù)反演模型；以1997-2013年SeaWiFS與MODIS衛(wèi)星的月平均遙感反射率觀測數(shù)據(jù)為基礎(chǔ)，利用本文構(gòu)建的遙感模型反演獲得了全球海洋與渤黃東海關(guān)鍵生物光學(xué)要素時空信息，并系統(tǒng)性地探討了渤黃東海關(guān)鍵生物光學(xué)要素遙感反演及時空變化。其主要結(jié)論如下： (1)構(gòu)建了適用于渤黃東海復(fù)雜光學(xué)特性的高精度大氣校正模型。一方面，基于交叉定標(biāo)的短波紅外大氣校正模型利用了5×5窗口平均的方法，降低短波紅外隨機(jī)誤差對遙感反射率反演精度的影響，同時利用短波紅外和近紅外波段交叉定標(biāo)的方法，降低短波紅外系統(tǒng)誤差對遙感反射率反演精度的影響；另一方面，針對1240nm波段的黑體假設(shè)在非常渾濁的水體中失效，，本文利用線性模型修復(fù)了1640nm波段數(shù)據(jù)的條帶，并構(gòu)建了16402130nm波段組合的大氣校正模型。結(jié)合實(shí)測遙感反射率數(shù)據(jù)可知，兩種大氣校正模型在遙感反射率精度方面較傳感器短波紅外大氣校正模型具有很大的改進(jìn)。 (2)構(gòu)建了適用于渤黃東海海水漫衰減系數(shù)反演的半分析模型。本文利用二流輻射傳輸原理，構(gòu)建了海水漫衰減系數(shù)與吸收系數(shù)和遙感反射率之間的關(guān)系，并結(jié)合雙波段半分析方法，最終確定了海水漫衰減系數(shù)與遙感反射率之間的關(guān)系。通過與渤黃東海實(shí)測數(shù)據(jù)對比可知，該模型較傳統(tǒng)模型精度提高了12%以上。 (3)構(gòu)建了適用于渤黃東海懸浮物濃度、葉綠素a濃度、吸收與散射系數(shù)、及海水漫衰減系數(shù)等關(guān)鍵生物光學(xué)要素反演的高精度神經(jīng)模型。本文利用分析方法構(gòu)建生物光學(xué)要素反演的概念模型，然后利用神經(jīng)網(wǎng)絡(luò)技術(shù)擬合概念模型中輸入和輸出參數(shù)之間的關(guān)系。利用全球?qū)崪y大數(shù)據(jù)集評估半分析與神經(jīng)網(wǎng)絡(luò)耦合模型的穩(wěn)定性和可靠性可知，與傳統(tǒng)神經(jīng)網(wǎng)絡(luò)模型和半分析模型比較，該模型在關(guān)鍵生物光學(xué)要素的反演方面具有更高的可靠性度，且適用的區(qū)域也更廣。 (4)1997-2013年期間渤黃東海關(guān)鍵生物光學(xué)要素的時空變化趨勢及其與全球變化的關(guān)系。受陸源和海陸相互作用的影響，各要素在渤黃東海呈近岸高，遠(yuǎn)岸地的分布格局。自1997年以來全球平均氣溫仍然以0.001℃mn-1速度升高，并導(dǎo)致全球海域葉綠素a濃度也逐年降低。作為全球海洋懸浮物的重要供給區(qū)域，渤黃東海葉綠素a與懸浮物濃度和全球葉綠素a和懸浮物濃度具有較高的時空相關(guān)性。
[Abstract]:Based on the field data and NOMAD data set from 2003 to 2012, the atmospheric correction model, the inversion model of coefficient and scattering coefficient of turbid water in the Bohai Sea, Yellow Sea and East China Sea are analyzed and discussed in this paper. Chlorophyll a and suspended matter concentration inversion models, as well as diffuse attenuation coefficient inversion models; based on monthly average remote sensing reflectance observations from SeaWiFS and MODIS satellites for 1997-2013, The spatiotemporal information of the key biological optical elements in the global ocean and the Bohai Sea and the Yellow Sea and the East China Sea are obtained by using the remote sensing model constructed in this paper. The remote sensing inversion and the temporal and spatial variations of the key biological optical elements in the Bohai Sea, the Yellow Sea and the East China Sea are systematically discussed. The main conclusions are as follows: (1) A high-precision atmospheric correction model for complex optical properties of the Bohai Sea, Yellow Sea and East China Sea is constructed. On the one hand, the short-wave infrared atmospheric correction model based on cross-calibration uses the method of 5 脳 5 window average to reduce the influence of the random error of short-wave infrared on the retrieval accuracy of remote sensing reflectivity. At the same time, the short-wave infrared and near-infrared band cross-calibration method is used to reduce the influence of the short-wave infrared system error on the precision of remote sensing reflectivity inversion. On the other hand, the blackbody hypothesis in 1240nm band is invalid in very cloudy water. In this paper, the linear model is used to repair the bands of 1640nm band data, and the atmospheric correction model of 16402130nm band combination is constructed. Combined with the measured data of remote sensing reflectivity, The two atmospheric correction models improve the precision of remote sensing reflectance greatly compared with the sensor short-wave infrared atmospheric correction model. (2) A semi-analytical model suitable for inversion of diffuse attenuation coefficient of sea water in the Bohai Sea, Yellow Sea and East China Sea is constructed. In this paper, the relationship between diffuse attenuation coefficient and absorption coefficient and remote sensing reflectivity of seawater is established by using the principle of two-stream radiation transfer. The relationship between diffuse attenuation coefficient and remote sensing reflectivity is finally determined with the method of two-band half-analysis. Compared with the measured data of the Bohai Sea, the Yellow Sea and the East China Sea, the accuracy of the model is improved by more than 12%. (3) the concentration of suspended matter, the concentration of chlorophyll a, the absorption and the scattering coefficient of suspended matter in the Bohai Sea, the Yellow Sea and the East China Sea are constructed. High precision neural model for inversion of key biological optical elements such as diffuse attenuation coefficient of seawater. In this paper, the analytical method is used to construct the conceptual model of bio-optical element inversion, and then the neural network technique is used to fit the relationship between the input and output parameters in the conceptual model. The stability and reliability of the coupling model of semi-analysis and neural network are evaluated by using the global measured big data set, which is compared with the traditional neural network model and the semi-analytical model. The model has a higher reliability in the inversion of key biological optical elements and has a wider range of applications. (4) the temporal and spatial trends of key biological optical elements in the Bohai Sea, the Yellow Sea and the East China Sea during 1997-2013 and their relationship with global changes. Influenced by the interaction between land source and sea and land, the distribution pattern of each element in the Bohai Sea, the Yellow Sea and the East China Sea is near shore high and far shore. Since 1997, the global average temperature has increased at the rate of 0.001 鈩

本文編號：2223919