相關誤差神經網路之應用於輻射量測植被和土壤含水量

簡易檢索 / 詳目顯示

回結果列表

研究生：	劉說芳 Shou-Fang Liu
論文名稱：	相關誤差神經網路之應用於輻射量測植被和土壤含水量 Retrieval of Crop Biomass and Soil Moisture from Brightness Temperatures by Using Backpropagation Neural Networks with Error Correlation
指導教授：	王文俊 Wen-June Wang 劉說安 Yuei-An Liou
口試委員:
學位類別：	博士 Doctor
系所名稱：	資訊電機學院 - 電機工程學系 Department of Electrical Engineering
畢業學年度：	90
語文別：	英文
論文頁數：	123
中文關鍵詞：	地表參數、微波遙測、神經網路
外文關鍵詞：	Surface Parameters, Remote Sensing, Neural Networks
相關次數：	點閱：10 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

本論文探討使用微波輻射量測地表參數，發展一個具相關誤差之倒傳遞神經網路，處理輻射亮溫與植被和土壤含水量的非線性映射。該神經網路模擬實際的生理現象，例如皮層上的某一點受到刺激，該點內具有最多神經末梢的神經纖維受到刺激，鄰近的神經纖維接受到較輕微的刺激，愈遠的纖維受到的刺激愈弱，即可建構為一個具相關誤差之倒傳遞神經網路。隨後將該神經網路應用在遙測領域----從亮溫反演植被和土壤含水量。土壤水分支配了地表與大氣間輻射能如何分割成感熱及潛熱。因此，在水文、氣象、農業和生物地球化學的理論上，扮演極為重要的角色，成為在遙測領域中很感興趣的參數。
本論文的具體成果可分述如下：
1. 具相關誤差之倒傳遞神經網路的發展
具相關誤差之倒傳遞神經網路(A Backpropagation Neural Network with Error Correlation)依照輸出層神經元間距離遠近定出不同的相關性，用來描述鄰近神經元的交互作用。實驗証明適當的相關誤差能提高神經網路訓練的精確度與準確度。具相關誤差之倒傳遞神經網路具有二個調整參數，首先距離參數α是用來調整相關誤差的的有效距離，其次強度參數λ是用來調整相關誤差的強度。在距離α外的神經元當作是無關的。
2. 應用在遙測領域 ~ 從亮溫反演土壤含水量:以模式模擬結果為例
在這個研究中，使用相關誤差之倒傳遞神經網路，從模擬亮溫反演土壤表面的含水量。使用的頻率包括先進微波掃瞄輻射計AMSR (Advanced Microwave Scanning Radiometer) 之6.9、10.7 GHZ 和土壤水分海洋鹽度感應器SMOS (Soil Moisture and Ocean Salinity sensor) 之1.4 GHZ。本研究使用的模式為LSP/R (Land Surface Process/ Radiobrightness) ，該模式提供一系列的土壤水分和亮溫資料。AMSR 的觀察角度為55度，而SMOS的觀察角度則介於0到55度間，為了方便本研究，吾人使用了0、10、20、30、40、50度等多重觀察角度。這些多重的頻率和角度，允許吾人設計不同的觀測方式來檢驗它們對於土壤含水量的敏感性。例如; L-band 單一觀看角度視為L-band 1-D觀測方式。同時，它可組合其他觀看角度，變成L-band 2-D 或多維的觀測方式，又或者結合AMSR 6.9、10.7 GHZ，變成多頻率/維度的觀測方式。本研究中，顯示L-band 1-D觀測方式亮溫對於土壤含水量是敏感的，且此敏感性可以用整合L-band第二觀察角度或AMSR 頻道來增加。
3. 應用在遙測領域 ~ 從亮溫反演植被和土壤含水量:以田野資料為例
在這個研究中使用法國PORTOS(PORTOS-93，96 experiment)輻射計在小麥田小麥生長期連續三個月收集的田野資料。從PORTOS-93資料中隨機選取作為神經網路訓練和測試，除了從亮溫反演土壤含水量並進一步反演植被含水量，我們發現利用L-band 2-D觀測方式時，可得到最好的植被和土壤含水量。此外，訓練過的神經網路以PORTOS-96資料作進一步的評估。研究顯示，反演土壤含水量的平均誤差率約為4%，植被總含水量平均誤差率約0.239kg/m2 (0.160kg/m2，93年和0.319kg/ m2，96年) ，雖然只使用1993年資料於網路訓練，因1996年的反演結果令人相當滿意，使我們對於使用神經網路從亮溫反演植被和土壤含水量充滿信心。

This dissertation is intended to investigate the sensing of surface parameters by microwave radiometry. A backpropagation neural network with error correlation (BNNEC) is developed to manage the nonlinear relationship between surface parameters and radiometric signatures. Its performance of retrieving plant water content (PWC) and soil moisture content (SMC) from brightness temperatures is examined by using both predictions from model simulations and measurements from field experiments.
The backpropagation neural network with error correlation (BNNEC) incorporates a novel rule, error correlation learning, to train a feedforward neural network. The correlated error terms associated with nearby neurons differ from those of the existing neural networks without considering of the correlations. That is, the output layers of its neurons are trained simultaneously and interactively through the error correlation terms. The BNNEC is applied to retrieve soil moisture from simulated brightness temperatures and perform the iris classification. Simulation results demonstrate the superiority of the proposed BNNEC.
In addition, we optimize the observing scheme for sensing surface soil moisture (SM) from simulated brightness temperatures by the BNNEC. The frequencies of interest include 6.9 and 10.7 GHz of the Advanced Microwave Scanning Radiometer (AMSR), and 1.4 GHz (L-band) of the Soil Moisture and Ocean Salinity (SMOS) sensor. The Land Surface Process/Radiobrightness (LSP/R) model is used to provide time series of both SM and brightness temperatures at 6.9 and 10.7 GHz for AMSR''s viewing angle of 55 degrees, and at L-band for SMOS''s multiple viewing angles of 0, 10, 20, 30, 40, and 50 degrees. These multiple frequencies and viewing angles allow us to design a variety of observation modes to examine their sensitivity to SM. For example, L-band brightness temperature at any single look angle is regarded as an L-band 1D observation mode. Meanwhile, it can be combined with either the observation at other angles to become an L-band 2D mode or a multiple dimensional observation mode, or with the observation at 6.9 or 10.7 GHz to become a multiple frequency/dimensional observation mode. In this study, it is shown that the L-band 1D radiometric observation is sensitive to SM. The sensitivity can be increased by incorporating radiometric observation either from a second angle, or from multiple look angles, or from any of the two lowest AMSR channels. In addition, the advantage of an L-band 2D mode or a multiple dimensional observation mode over an L-band 1D observation mode is demonstrated.
Moreover, we investigate the best observing configuration for sensing wheat plant water content (PWC) and soil moisture content (SMC) profiles from the measured H- and V-polarized brightness temperatures at 1.4 (L-band), and 10.65 (X-band) GHz by the BNNEC. The brightness temperatures were taken by the PORTOS radiometer over wheat fields through 3 months growth cycles in 1993 (PORTOS-93) and 1996 (PORTOS-96). During both field campaigns, the radiometer was used to measure brightness temperatures at incident angles from 0 to 50 degrees at L-band and at an incident angle of 50 degrees at X-band. The SMC profiles were measured to the depths of 10 cm in 1993 and 5 cm in 1996. The wheat was sampled approximately once a week in 1993 and 1996 to obtain its dry and wet biomass (i.e., PWC). The BNNEC was trained with observations randomly chosen from the PORTOS-93 data, and evaluated by the remaining data from the same set. The trained neural network is further evaluated with data from the PORTOS-96.

Abstract	IV
Acronym	        VI
List of Figures	VIII
List of Tables	X
Chapter 1  Introduction
1.1 Microwave Remote Sensing					  1
1.2 Motivation and Background					  3
1.3 Purpose and Contribution					  6
1.4 Organization of the Dissertation				  8
Chapter 2  A Backpropagation Neural Network with Error Correlation
2.1  Introduction		                            	  9
2.2  Error Correlation Learning			          11
2.2.1  The Concept of Error Correlation Learning		  11
2.2.2  The Physical Explanation of Error Correlation Learning 12
2.2.3  The Range of Parameters in Error Correlation Learning  14
2.2.4  Backpropagation Algorithm with Error Correlation	  15
2.2.5  Conjugate Gradient Algorithm with Error Correlation    18
2.2.6  Marquardt Algorithm with Error Correlation		  20
2.3  Experiments			                          20
2.3.1  Retrieving Soil Moisture from Simulated Brightness
Temperatures                                           20
2.3.2  The Iris Data Classification	                          23
2.4  Summary						          25
Chapter 3  Retrieving Soil Moisture from Simulated Brightness
Temperatures by a Backpropagation Neural Network
with Error Correlation
3.1  Introduction                                    	          30
3.2  The LSP/R Model and the BNNEC				  33
3.2.1  The LSP/R Model			                  33
3.2.2  The BNNEC                                  		  35
3.2.3  The Training and Testing Data 	                  36
3.3  Retrieval Analysis					  39
3.3.1  Observation Modes			                  39
3.3.2  Results from Simulated Single Satellite Observations	  40
3.3.3  By Combined AMSR and L-band Observations	          43
3.4  Summary							  44
Chapter 4  Radiometric Sensing of Soil Moisture and Biomass Based
on the Field Measurements
4.1  Introduction                			          56
4.2  Retrieval Description					  58
4.2.1  The Physical System		                    	  58
4.2.2  Field Measurements	                                  59
4.2.3  Neural Network	                                  62
4.2.4  Observation Modes			                  64
4.3  Results and Discussion                                     65
4.3.1  Correlation between Brightness Temperatures and Surface
Parameters                                             65
4.3.2  Retrieval of Soil Moisture Content 		          66
4.3.3  Retrieval of Vegetation Water Content 	          68
4.4  Summary 							  69
Chapter 5  Conclusions						  78
References						          81
Author’s Information						  92
Publication List						  93

                                

[1] Battiti, R., "Accelerated backpropagation learning: Two optimization methods", Complex Systems, vol. 3, pp. 331-342, 1989.
[2] Bertuzzi, P., Bruckler, L., Gabilly, Y., and Gaudu, J.-C., "Calibration and error analysis of a gamma-ray probe for the in-situ measurement of dry bulk density", Soil Sci., vol. 144, pp. 425-436, 1987.
[3] Calvet, J.?C., Wigneron, J.?P., Chanzy, A., Raju, S., and Laguerre, L., "Microwave dielectric properties of a silt?loam at high frequencies", IEEE Trans. Geosc. Remote Sensing, vol. 33, pp. 634-642, May. 1995.
[4] Demuth, H. and Beale, M., Neural Network Toolbook: For use with MATLAB, User''s Guide Version 3.0, fifth printing, pp. 742, The MathWorks Inc., Natick, MA, USA, 1998.
[5] DeVries, D. A., "Simultaneous transfer of heat and moisture in porous media", Trans. Am. Geophys. Union, vol.39, pp. 909-916, Oct. ,1958.
[6] Drucker, H., Cortes, C., Jackel, L. D., LeCun, Y. and Vapnik, V., "Boosting and other ensemble methods", Neural Computation, vol.6, pp. 1289-1301, 1994.
[7] Duda, R., & Hart, P., Pattern classification and scene analysis, NewYork, Wiley, 1973.
[8] England, A. W., "Radiobrightness of diurnally heated, freezing soils", IEEE Trans. Geosci. Remote Sensing, vol. 28, pp. 464-476,1990.
[9] England, A.-W. et al., The HYDROSTAR Mission. Full proposal, Answer to the Earth Systems Science Pathfinder (ESSP) announcement of opportunity, NASA, 1998.
[10] Ferrazzoli,P., Wigneron, J.-P., Guerriero, L., and Chanzy, A., "Multifrequency emission of whet: modeling and applications", IEEE Trans. Geosci. Remote Sensing, vol. 38, pp. 2598-2607, Jun. 2000.
[11] Gupta, R. P., Remote sensing geology, Springer-Verlag Berlin Heidelberg, 1991.
[12] Guyton, A. C., Textbook of medical physiology, Press of W. B. Saunders Company, 1986.
[13] Hagan, M. T., and Menhaj, M., "Training feedforward networks with the Marquardt algorithm", IEEE Transactions on Neural Networks, vol. 5, no. 6, pp. 989-993, 1994.
[14] Hecht-Nielsen, R., "Application of counterpropagation networks", Neural Networks, vol.1, pp. 131-139, 1988.
[15] Jackson, T. J., LeVince, D. M., Hsu, A. Oldak, Y., A., Starks, P. J., Swift,C. T., Isham, J., and Haken, M., "Soil moisture mapping at regional scales using microwave radiometry: the Southern Great Plains hydrology experiment", IEEE Trans. Geosci. Remote Sensing, vol. 37, pp.2 136-2 151,1999.
[16] Jackson, T. J., and Schmugge, T. J., "Passive microwave remote sensing system for soil moisture: Some supporting research", IEEE Trans. Geosci. Remote Sensing , vol. 27, pp. 225-235, 1989.
[17] Jensen, J. R., Remote sensing of the environment: an earth resource perspective, Upper Saddle River, N.J.: Prentice Hall, 2000.
[18] Kerr, Y. H. et al., "MIRAS on RAMSES: Radiometry applied to soil moisture and salinity measurements", Full proposal, Answer to the Call for Earth Explorer Opportunity Missions, ESA, 1998.
[19] Kerr, Y. et al., SMOS: Soil Moisture and Ocean Salinity, Full proposal, Answer to the Call for Earth Explorer Opportunity Missions, ESA,1998.
[20] Kerr, Y. H., and Wigneron, J. P., "Vegetation models and observations - a review", In Proc. ESA/NASA Int''l Workshop on Passive Microwave Remote Sensing Research Related to Land-Atmosphere Interactions, Saint-lary, France, Jan. 11-15, 1993.
[21] Kohonen, T., "The self-organizing map", Proc. IEEE, vol. 78, no. 9, pp. 1481 -1490, 1990.
[22] Kohonen, T.; Barna, G.; Chrisley, R., "Statistical pattern recognition with neural networks: benchmarking studies", IEEE International Conference, vol.1, pp. 61 -68, 1988.
[23] Laguerre, L., "Influence de la rugosite de surface en radiometrie micro-onde des sols nus: modelisation et inversion", PhD thesis, CESBIO and Institut National Polytechnique, Toulouse, Nov. 23, p. 161, 1995.
[24] Le Vine, D. M., "Synthetic aperture systems", IEEE Trans. MTT, vol. 47, pp. 2228-2236, Dec. 1999.
[25] Le Vine, D. M., Griffis, A. J., Swift, C. T., and Jackson, T. J., "ESTAR: A synthetic aperture microwave radiometer for remote sensing applications", IEEE Proc., vol. 82, pp. 1 787-1 801, Dec. 1994.
[26] Lillesand, T. M., and Kiefer, R. W., Remote sensing and image interpretation, New York: Wiley, 1979
[27] Liou, Y.-A., Chen, K.-S., and Wu, T.-D., "Reanalysis of L-band brightness predicted by the LSP/R model: Incorporation of rough surface scattering", IEEE Trans. Geosci. Remote Sensing, vol. 39, pp. 129-135, Jan. 2001.
[28] Liou, Y.-A., and England, A. W., "Annual temperature and radiobrightness signatures for bare soils", IEEE Trans. Geosci. Remote Sensing. vol. 34, pp. 981-990, 1996.
[29] Liou, Y.-A., and England, A. W., "A land surface process/radiobrightness model with coupled heat and moisture transport in soil", IEEE Trans. Geosci. Remote Sensing, vol. 36, pp. 273-286, Jan. 1998.
[30] Liou, Y.-A., and England, A. W., "A land surface process/radiobrightness model with coupled heat and moisture transport for freezing soils", IEEE Trans. Geosci. Remote Sensing, vol. 38, pp. 669-677, 1998.
[31] Liou, Y.-A., Galantowicz, J., and England, A. W., "A land surface process /radiobrightness with coupled heat and moisture transport for prairie grassland", IEEE Trans. Geosci. Remote Sensing, vol. 37, pp. 1848-1859, Jul. 1999.
[32] Liou, Y.-A., Kim, E. J., and England, A. W., "Radiobrightness of prairie soil and grassland during dry-down simulations", Radio Sci., vol. 33, pp. 259-265, 1998.
[33] Liou, Y.-A., Liu, S. F., Lee, J.-B., Wang,W. J,. and Wigneron, J.-P., "Retrievals of land surface parameters from measured brightness temperatures at 1.4 and 10.65 GHz by a neural network", IGARSS2001, Sydney, Australia, pp. 9-13, July 2001.
[34] Liou, Y.-A., Liu, S. F., Wang,W. J, "Retrieving soil moisture from simulated brightness temperatures by a neural network", IEEE Trans. Geosci. Remote Sensing, vol. 39, pp. 1662-1672, Aug. 2001.
[35] Liou, Y. A., Tzeng, Y. C. and Chen, K. S., "The use of neural networks in radiometric studies of land surface parameters", Proc. NSC Part A: Physical Science and Engineering, vol. 23, pp. 511-518, 1999.
[36] Liou, Y. A., Tzeng, Y. C. and Chen, K. S., "A neural network approach to radiometric sensing of land surface parameters", IEEE Trans. Geosci. Remote Sensing, vol. 37, pp. 2 718-2 724, 1999.
[37] Liou, Y. A., Tzeng, Y. C. and Wigneron, J. -P., " Probing Soil Moisture Profiles by L-band 2-D Radiometry over a Wheat Field" , In: Remote Sensing and Hydrology 2000, M. Owe and K. Brubaker ,Eds., Wallingford, UK, 2001.
[38] Lippmann, R. P., "An introduction to computing with neural nets", IEEE ASSP Magazine,pp. 4-44, April 1987.
[39] Liu, Y. and Yao, X., "Ensemble learning via negative correlation" , Neural Networks, vol. 12, pp. 1399-1404, 1999.
[40] Magagi, R. D., Kerr,Y. H., and Meunier, J.-C., "Results of combining L- and C-band passive microwave airborne data over the Sahelian area", IEEE Trans. Geosci. Remote Sensing, vol. 38, pp. 1 997-2 008, 2000
[41] Moller, M. F., "A scaled conjugate gradient algorithm for fast supervised learning" , Neural Networks, vol. 6, pp. 525-533, 1993.
[42] Mualem, Y., "A new model for predicting the hydraulic conductivity of unsaturated porous media" , Water Resour. Res., vol. 12, pp. 513-522, Jun. 1976.
[43] Nadler, A., Dasberg, S., and Lapid, I., "Time domain reflectrometry measurements of water content and electrical conductivity of layered soil columns" , Soil Sci. Soc. Am. J., vol. 55, pp. 938-943, 1991.
[44] Njoku, E. G., and Entekhabi, D., "Passive microwave remote sensing of soil moisture," J. Hydrology, vol. 184, pp. 101-129, 1996.
[45] Njoku, E. G., and Li, L., "Retrieval of land surface parameters using passive microwave measurements at 6-18 GHz", IEEE Trans. Geosci. Remote Sensing, vol. 37, pp. 79-93, 1999
[46] Njoku, E. G., Yahya, R.-S., Sercel, J., Wilson,W. J., and Moghaddam, M., "Evaluaiton of an inflatable antenna concept for microwave sensing of soil moisture and ocean salinity", IEEE Trans. Geosci. Remote Sensing, vol. 37, pp. 63-78, 1999
[47] Perrone, M. and Cooper, L. N., "When networks disagree: ensemble methods for hybrid neural networks" , In R. J. Mammone, Neural networks for speech and image pressing, London: Chapman & Hall, 1993.
[48] Philip, J. R., and D. A. de Vries, Moisture movement in porous materials under temperature gradients,'' Trans. Am. Geophys. Union, vol. 38, pp. 222-232, Apr. 1957.
[49] Powell, M. J. D., "Restart procedures for the conjugate gradient method" , Mathematical Programming, vol. 12, pp. 241-254, 1977.
[50] Rossi, C., and Nimmo, J. R., "Modeling of soil water retention from saturation to oven dryness", Water Resour. Res., vol. 30, pp. 701-708, Mar. 1994.
[51] Rumelhart, D. E., Hinton, G. E. and Williams, R. J., "Learning Internal Representations by Error Propagation" , in D. E. Rumelhart, J. L. McClelland, and the PDP Research Group, eds. , Parallel Distributed Processing, Vols. 1:Foundations, Cambridge, MA: The M.I.T. Press, 1986.
[52] Sabins, F. F., Remote sensing: principles and interpretation, San Francisco: W. H. Freeman , 1978
[53] Scales, L. E., Introduction to Non-Linear Optimization, New York: Springer Verlag, 1985.
[54] Schmugge, T., O''Neill,P. E., and Wang, J. R., "Passive microwave soil moisture research", Trans. Geosci. Remote Sensing, vol. 24, pp. 12-22, 1986.
[55] Shibata, A., "AMSR/AMSR-E algorithm development and data distribution", Proc. IGARSS, Hawaii, pp. 24-28, July 2000.
[56] Specht, D. F., "Probabilistic neural networks", Neural Networks, Vol. 3, pp. 109-118, 1990.
[57] Ulaby, F. T, Moore, R. K., and Fung, A. K., Microwave Remote Sensing: Active and Passive, Vol. I: Microwave Remote Sensing Fundamentals and Radiometry. Artech House Inc, Norwood, 1981.
[58] Ulaby, F. T., and El-Rayes, M. A., "Microwave dielectric spectrum of vegetation--Part II: Dual-dispersion model", IEEE Trans. Geosci. Rem. Sens., vol. 25, pp. 550-557, Sep. 1987.
[59] Wang, J.-R., Newton, R. W., and Rouse,J. W., Jr., "Passive microwave remote sensing of soil moisture," IEEE Trans. Geosci. Remote Sensing, vol. GE-18(4), pp. 296-302, 1980.
[60] Wang, J. R., "Microwave emission from smooth bare fields and soil moisture sampling depth", IEEE Trans. Geosci. Remote Sens, Vol. 25, pp. 616-622, 1987.
[61] Watrous, R. L., "Learning algorithms for connectionist networks: Applied gradient methods of nonlinear optimization" , Proceedings IEEE 1st International Conference on Neural Networks, vol. 2, pp. 619-628, 1987.
[62] Wigneron, J.-P., Chanzy, A., Calvet, J.-C., and Bruguier, N., "A simple algorithm to retrieve soil moisture and vegetation biomass using passive microwave measurements over crop fields" , Remote Sens. Environ,vol. 51, pp. 331-341,1995.
[63] Wigneron, J.-P., Laguerre, L., and Kerr, Y. H., "A simple parameterization of the L-band microwave emission from rough agricultural soils" , IEEE Trans. Geosci. Remote Sensing, vol. 39, pp. 1697-1707, Aug. 2001
[64] Wigneron, J.-P., Kerr, Y. H., Chanzy, A., and Jin, Y. Q., "Inversion of surface parameters from passive microwave measurements over a soybean field", Remote Sens. Environ., vol. 46, pp. 61-72, 1993.
[65] Wigneron, J.-P., Schmugge,T., Chanzy,A., Calvet,J.-C., and Kerr,Y., "Use of passive microwave remote sensing to monitor soil moisture" , Agronomie, vol. 18, pp. 27-43, 1998.
[66] Wigneron, J.-P., Waldteufel P., Chanzy A., Calvet J.-C., Marloie O., Hanocq J.-F., and Kerr Y., "Retrieval capabilities of L-Band 2-D interferometric radiometry over land surfaces (SMOS Mission) ", Proceedings of 6th Specialist Meeting on Microwave Radiometry, Firenze, Italy, March 16-18, S. Paloscia and P. Pampaloni (Eds), VSP, The Netherland, 1999.
[67] Wigneron, J.-P., Waldteufel P., Chanzy A., Calvet J.-C.,, and Kerr, Y., "Two-dimensional microwave inteferometer retrieval capabilities over land surfaces (SMOS mission)", Remote Sensing Environ., vol. 73, pp. 270-282, 2000.
[68] Williams, P. J., "Experimental determination of apparent specific heats of frozen soils", Geotechnique, vol. 14, pp. 133-142, 1964.
[69] Yeh, Y. C., Neural network model applications and implementation, Fifth Edition, Press of Scholars Company, Taipei, 1999. (in Chinese)

簡易檢索 / 詳目顯示

相關論文