本文主要研究利用全球衛星導航系統反射訊號接收法(GNSS-R)來進行內陸水體的探測。GNSS-R是一種遙測方法，其為在衛星上安裝訊號接收儀來接收地表反射的全球導航系統(GNSS)之訊號。藉由量測地表反射之特性，可進行颱風強度演算、海面風速計算、土壤濕度、測高、水體偵測等研究。而Cyclone Global Navigation Satellite System (CYGNSS)作為使用此方法進行任務的主要衛星之一，過去許多研究透過其資料來計算地表的有效表面反射率(P_(r,eff))和訊噪比(SNR)來監測土壤濕度變化和檢測內陸水體。而因水體大多在CYGNSS產出的延遲都卜勒圖(DDM)中呈現同調性(coherency)，所以為了加快資料處理的速度和維持判斷的正確性，吾人嘗試利用卷積神經網路的方法來辨別7個月的DDM之同調性並計算其P_(r,eff)和SNR，最後與無利用卷積神經網路處理之資料計算P_(r,eff)和SNR的結果相比，並觀察兩者的特性，可以得到以下結論:第一，根據驗證統計圖表，亞馬遜地區的P_(r,eff)範圍約為 -49 至 -3 dB。加入CNN後，範圍變為 -48 至 0 dB。可視化結果顯示，亞馬遜支流在 -50 至 -45 dB 開始與地表水存在率圖(SWO) 重合，東亞馬遜的濕地和河流在 -40 至 -35 dB 重合，亞馬遜主流在 -35 至 -30 dB 重合。主要區別在於CNN會消除亞馬遜支流中的大部分異調性(non-coherent)的資料。第二，亞馬遜地區的水體 SNR 範圍約為 5 到 30，加入CNN後也是如此。可視化SNR結果顯示，亞馬遜河支流、東亞馬遜的濕地和河流在SNR 5 到 10 時開始與 SWO 重合，亞馬遜河主流在SNR 9 到 14 時重合。加入CNN後的結果與未加入時相似，主要區別與之前的結果相同。第三，根據統計圖表，泰國地區的水體P_(r,eff)範圍約為 -46 至 0 dB。加入CNN進行估算後，水體的範圍約為 -50 至 -5 dB。可視化結果顯示，南巫河中游的濕地在 -45 至 -40 dB 開始與SWO重合，南巫河和昭披耶河的上游及小河流在 -40 至 -35 dB 重合，南巫河和昭披耶河下游的三角洲在 -30 至 -20 dB 重合。加入CNN後的結果與未加入時相似，唯一的區別在於 -50 至 -35 dB 範圍內，異調性資料被徹底消除。第四，根據統計圖表，泰國地區的水體 SNR 範圍約為 7 至 30，加入CNN後也是如此。可視化SNR結果顯示，昭披耶河和南巫河的上游及南巫河中游的濕地在SNR 5 至 10 時開始與SWO重合，兩河流的三角洲在SNR 7 至 12 重合，昭披耶河中游河流在SNR 11 至 16 重合，昭披耶河和南巫河的三角洲在SNR 15 至 20 重合。加入CNN後的結果與未加入時相似，主要區別在於SNR 1 至 8 的範圍內，異調性資料被消除。

This study focuses on utilizing Global Navigation Satellite System Reflectometry (GNSS-R) for the detection of inland water bodies. GNSS-R is a remote sensing technique that involves the installation of signal receivers on satellites to capture signals reflected from the Earth's surface. By analyzing the characteristics of these reflected signals, various phenomena can be studied, including typhoon intensity, sea surface wind speed, soil moisture, altimetry, and waterbody detection. The Cyclone Global Navigation Satellite System (CYGNSS) is one of the principal satellites employing this technique. Numerous previous studies have leveraged CYGNSS data to calculate the effective surface reflectivity (Pr,eff) and signal-to-noise ratio (SNR) for monitoring soil moisture variations and detecting inland waterbodies. Given that waterbodies typically exhibit coherency in the Delay Doppler Maps (DDMs) generated by CYGNSS, this study seeks to enhance data processing speed and maintain detection accuracy by employing convolutional neural networks (CNNs). We analyze the coherency in seven months of DDMs, calculating the corresponding Pr,eff and SNR. Finally, we compare these results with those obtained without CNNs to examine the differences and characteristics of each approach. We find that first, the Pr,eff range in the Amazon region is approximately -49 to -3 dB, extending to -48 to 0 dB after adding CNN. Visualization shows Amazon tributaries aligning with the Surface Water Occurrence (SWO) at -50 to -45 dB, eastern Amazon wetlands and rivers at -40 to -35 dB, and the Amazon mainstream at -35 to -30 dB. CNN eliminates most non-coherent data in the tributaries. Second, the SNR range for Amazon water bodies is about 5 to 30, unchanged with CNN. Visualization reveals tributaries and eastern Amazon aligning with SWO at SNR 5 to 10, and the Amazon mainstream at SNR 9 to 14. The primary difference with CNN is the removal of non-coherent data. Third, in Thailand, the Pr,eff range is approximately -46 to 0 dB, changing to -50 to -5 dB with CNN. Visualization shows midstream Nam Ngum wetlands aligning with SWO at -45 to -40 dB, upstream Nam Ngum and Chao Phraya Rivers at -40 to -35 dB, and their deltas at -30 to -20 dB. CNN eliminates non-coherent data in the -50 to -35 dB range. Fourth, the SNR range for Thailand water bodies is approximately 7 to 30, unchanged with CNN. Visualization shows upstream Chao Phraya and Nam Ngum Rivers aligning with SWO at SNR 5 to 10, their deltas at SNR 7 to 12, Chao Phraya midstream at SNR 11 to 16, and deltas at SNR 15 to 20. The main difference with CNN is the elimination of non-coherent data in the SNR 1 to 8 range.

Clarizia, M.P., Investigating the Effect of Ocean Waves on GNSS-R Microwave Remote Sensing Measurements, University of Southampton PhD Thesis, 2012
S. T. Lowe, J. L. LaBrecque, C. Zuffada, L. J. Romans, L. E. Young and G. A. Hajj, "First spaceborne observation of an Earth-reflected GPS signal," in Radio Science, vol. 37, no. 1, pp. 1-28, Jan. 2002, doi: 10.1029/2000RS002539.
M. P. Clarizia, N. Pierdicca, F. Costantini and N. Floury, "Analysis of CYGNSS Data for Soil Moisture Retrieval," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 12, no. 7, pp. 2227-2235, July 2019, doi: 10.1109/JSTARS.2019.2895510.
Chew, C.; Small, E. Description of the UCAR/CU Soil Moisture Product. Remote Sens. 2020, 12, 1558. https://doi.org/10.3390/rs12101558
Wei, H.; Yu, T.; Tu, J.; Ke, F. Detection and Evaluation of Flood Inundation Using CYGNSS Data during Extreme Precipitation in 2022 in Guangdong Province, China. Remote Sens. 2023, 15, 297. https://doi.org/10.3390/rs15020297
M. M. Al-Khaldi, J. T. Johnson, S. Gleason, E. Loria, A. J. O’Brien and Y. Yi, "An Algorithm for Detecting Coherence in Cyclone Global Navigation Satellite System Mission Level-1 Delay-Doppler Maps," in IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 5, pp. 4454-4463, May 2021, doi: 10.1109/TGRS.2020.3009784.
M. M. Al-Khaldi et al., "Inland Water Body Mapping Using CYGNSS Coherence Detection," in IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 9, pp. 7385-7394, Sept. 2021, doi: 10.1109/TGRS.2020.3047075.
Arabi, S.; Asgarimehr, M.; Kada, M.; Wickert, J. Hybrid CNN-LSTM Deep Learning for Track-Wise GNSS-R Ocean Wind Speed Retrieval. Remote Sens. 2023, 15, 4169. https://doi.org/10.3390/rs15174169
Y. Wang and Y. J. Morton, "Coherent GNSS Reflection Signal Processing for High-Precision and High-Resolution Spaceborne Applications," in IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 1, pp. 831-842, Jan. 2021, doi: 10.1109/TGRS.2020.2993804.
S. Gleason, J. Johnson, C. Ruf and C. Bussy-Virat, "Characterizing Background Signals and Noise in Spaceborne GNSS Reflection Ocean Observations," in IEEE Geoscience and Remote Sensing Letters, vol. 17, no. 4, pp. 587-591, April 2020, doi: 10.1109/LGRS.2019.2926695.
Y. Jia et al., "Temporal-Spatial Soil Moisture Estimation from CYGNSS Using Machine Learning Regression With a Preclassification Approach," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 4879-4893, 2021, doi: 10.1109/JSTARS.2021.3076470.
Pavur, G.K., Kim, H., Fang, B. et al. Spatial comparison of inland water observations from CYGNSS, MODIS, Landsat, and commercial satellite imagery. Geosci. Lett. 11, 12 (2024). https://doi.org/10.1186/s40562-024-00321-1
Rose, Razal & Gleason, Scott & Ruf, Christopher. (2014). The NASA CYGNSS mission: a pathfinder for GNSS scatterometry remote sensing applications. Proceedings of SPIE - The International Society for Optical Engineering. 9240. 10.1117/12.2068378
Jensen, K.; McDonald, K.; Podest, E.; Rodriguez-Alvarez, N.; Horna, V.; Steiner, N. Assessing L-Band GNSS-Reflectometry and Imaging Radar for Detecting Sub-Canopy Inundation Dynamics in a Tropical Wetlands Complex. Remote Sens. 2018, 10, 1431. https://doi.org/10.3390/rs10091431
E. Valencia, V. U. Zavorotny, D. M. Akos and A. Camps, "Using DDM Asymmetry Metrics for Wind Direction Retrieval From GPS Ocean-Scattered Signals in Airborne Experiments," in IEEE Transactions on Geoscience and Remote Sensing, vol. 52, no. 7, pp. 3924-3936, July 2014, doi: 10.1109/TGRS.2013.2278151.