本研究利用日本宇宙航空研究開發機構（Japan Aerospace Exploration Agency，JAXA）向日葵衛星-8/9（Himawari-8/9）的四個波段資料，通過機器學習方法訓練模型來估算海洋中的葉綠素-a 濃度。隨著全球氣候變化和人類活動對海洋生態系統的影響日益增加，準確監測海洋中葉綠素-a 濃度變得尤為重要。葉綠素-a 作為浮游植物的主要色素，其濃度變化可反映海洋初級生產力和生態系統健康狀況。現有的監測技術雖然能提供一定程度的數據支持，但受限於空間解析度和時間分辨率的不足，無法滿足精細化監測的需求。

為了提高估算的準確性，我們首先對收集的多光譜衛星數據進行了預處理和特徵提取，並使用了包括XGBoost、LightGBM 和CatBoost 在內的先進機器學習算法進行模型訓練。我們的研究方法包括對衛星數據進行噪音去除、波段間的融合以及特徵工程等，旨在最大化地利用數據信息。為了驗證模型的效果，我們將估算結果與日本 JAXA 提供的實際葉綠素-a 濃度數據進行對比，評估模型的準確性和穩定性。結果顯示，我們的模型能夠有效捕捉海洋中葉綠素-a 濃度的變化趨勢，並在不同地理區域和時間範圍內保持良好的性能。

此外，我們應用了超分辨率技術（SR）來提升原始衛星波段資料的影像解析度，將解析度從5 公里提升至2 公里。我們基於雙三次插值（Bicubic）進行影像超解析方法。這方法能有效提升影像的細節和清晰度，結果表明，提升後的高解析度影像顯著提高了葉綠素-a 濃度估算的精度和空間分辨率。這不僅有助於更精確地監測海洋生態系統的變化，還能為海洋生態環境的管理和保護提供更加詳細和可靠的數據支持。

本研究的創新之處在於結合了先進的機器學習算法和超分辨率技術，為衛星遙測數據的應用開創了新的可能性。我們展示了這些技術在提升數據解析度和準確性方面的潛力，並為未來的海洋環境監測和研究提供了新的技術路徑和方法。這些成果不僅可以應用於海洋生態系統的監測，還可擴展至其他環境監測領域，如大氣污染、水質監測等，具有廣泛的應用前景。

This study uses data from four spectral bands of the Himawari-8/9 satellites operated by the Japan Aerospace Exploration Agency (JAXA) to estimate chlorophyll-a concentrations in the ocean using machine learning methods. Accurately monitoring ocean chlorophyll-a concentration is crucial due to the impact of climate change and human activities on marine ecosystems. Chlorophyll-a, as the main pigment of phytoplankton, reflects changes in marine ecosystem health.

To enhance the accuracy of the estimation, we first preprocess the collected multispectral satellite data and perform feature extraction. Advanced machine learning algorithms, including XGBoost, LightGBM, and CatBoost, are employed for model training. The estimated results from the model are compared with actual chlorophyll-a concentration data provided by Japan JAXA to evaluate the accuracy and robustness of the model. The results show that our model effectively captures the variations in chlorophyll-a concentration in the ocean and maintains good performance across different geographical areas and time periods.

Furthermore, super-resolution (SR) techniques were applied to enhance the image resolution of the original satellite spectral data, improving the resolution from 5 kilometers to 2 kilometers. We utilized the Interpolation-Based Super-Resolution method for image super-resolution. The results indicate that the high-resolution images significantly enhance the accuracy and spatial resolution of chlorophyll-a concentration estimation.

Through this study, we demonstrate the potential of using machine learning and superresolution techniques to enhance the application of satellite remote sensing data, providing new technical pathways and methods for future marine environmental monitoring and research.

[1] N. A. M. R. 國家海洋研究院, "近即時之高時間分辨率AI 超解析度海表面葉綠素a 產品精度驗證NAMR-113023," 2024.
[2] N. A. M. R. 國家海洋研究院. https://www.namr.gov.tw/ch/.
[3] 海 洋 委 員 會海洋保育署, and O. C. A. , Ocean Affairs Council. https://www.oca.gov.tw/ch/index.jsp.
[4] J. Meteorological Satellite Center (MSC). "Himawari User's Guide - Imager (AHI),"2015; https://www.data.jma.go.jp/mscweb/en/himawari89/space_segment/spsg_ahi.html?fbclid=IwAR2fpub2091wW05umc2EYW_EZfV3ezGshzMI8c-Z9B2HEOUgkwpEJmvHugA.
[5] K. Bessho, K. Date, M. Hayashi et al., “An Introduction to Himawari-8/9— Japan' s New-Generation Geostationary Meteorological Satellites,” Journal of the Meteorological Society of Japan. Ser. II, vol. 94, no. 2, pp. 151-183, 2016.
[6] X. Sun, S. Gnanamuthu, N. Zagade et al., “Data article: Full disk real-time Himawari-8/9 satellite AHI imagery from JAXA,” Journal of Renewable and Sustainable Energy, vol. 13, no. 6, 2021.
[7] J. E. O’Reilly, S. Maritorena, D. A. Siegel et al., “Ocean color chlorophyll a algorithms for SeaWiFS, OC2, and OC4: Version 4,” SeaWiFS postlaunch calibration and validation analyses, Part, vol. 3, pp. 9-23, 2000.
[8] M. J. Behrenfeld, R. T. O’Malley, D. A. Siegel et al., “Climate-driven trends in contemporary ocean productivity,” Nature, vol. 444, no. 7120, pp. 752-755, 2006.
[9] H. Murakami, “Ocean color estimation by Himawari-8/AHI,” Remote sensing of the oceans and inland waters: Techniques, applications, and challenges, pp. Vol. 9878, pp. 177-186, 2016.
[10] G. Panteras, and G. Cervone, “Enhancing the temporal resolution of satellite-based flood extent generation using crowdsourced data for disaster monitoring,”International journal of remote sensing, vol. 39, no. 5, pp. 1459-1474, 2018.
[11] K. Kaku, “Satellite remote sensing for disaster management support: A holistic and staged approach based on case studies in Sentinel Asia,” International Journal of Disaster Risk Reduction, vol. 33, pp. 417-432, 2019.
[12] C. A. Boulton, T. M. Lenton, and N. Boers, “Pronounced loss of Amazon rainforest resilience since the early 2000s,” Nature Climate Change, vol. 12, no. 3, pp. 271-278, 2022.
[13] A. Morel, Y. Huot, B. Gentili et al., “Examining the consistency of products derived from various ocean color sensors in open ocean (Case 1) waters in the perspective of a multi-sensor approach,” Remote Sensing of Environment, vol. 111, no. 1, pp. 69-88, 2007.
[14] R. Doerffer, and H. Schiller, “The MERIS Case 2 water algorithm,” International Journal of Remote Sensing, vol. 28, no. 3-4, pp. 517-535, 2007.
[15] E. Siswanto, J. Tang, H. Yamaguchi et al., “Empirical ocean-color algorithms to retrieve chlorophyll-a, total suspended matter, and colored dissolved organic matter absorption coefficient in the Yellow and East China Seas,” Journal of oceanography, vol. 67, pp. 627-650, 2011.
[16] C.-F. Chen, N.-T. Son, N.-B. Chang et al., “Multi-decadal mangrove forest change detection and prediction in Honduras, Central America, with Landsat imagery and a Markov chain model,” Remote Sensing, vol. 5, no. 12, pp. 6408-6426, 2013.
[17] A. Ansper, and K. Alikas, “Retrieval of chlorophyll a from Sentinel-2 MSI data for the European Union water framework directive reporting purposes,” Remote Sensing, vol. 11, no. 1, pp. 64, 2018.
[18] G. Chavula, P. Brezonik, P. Thenkabail et al., “Estimating chlorophyll concentration in Lake Malawi from MODIS satellite imagery,” Physics and Chemistry of the Earth, Parts A/B/C, vol. 34, no. 13-16, pp. 755-760, 2009.
[19] S. I. Salem, H. Higa, H. Kim et al., “Assessment of chlorophyll-a algorithms considering different trophic statuses and optimal bands,” Sensors, vol. 17, no. 8, pp. 1746, 2017.
[20] J. E. O'Reilly, S. Maritorena, B. G. Mitchell et al., “Ocean color chlorophyll algorithms for SeaWiFS,” Journal of Geophysical Research: Oceans, vol. 103, no. C11, pp. 24937-24953, 1998.
[21] M. W. Matthews, S. Bernard, and K. Winter, “Remote sensing of cyanobacteriadominant algal blooms and water quality parameters in Zeekoevlei, a small hypertrophic lake, using MERIS,” Remote Sensing of Environment, vol. 114, no. 9, pp. 2070-2087, 2010.
[22] W. J. Moses, A. A. Gitelson, S. Berdnikov et al., “Estimation of chlorophyll-a concentration in case II waters using MODIS and MERIS data—successes and challenges,” Environmental research letters, vol. 4, no. 4, pp. 045005, 2009.
[23] M. Kahru, and B. G. Mitchell, “Spectral reflectance and absorption of a massive red tide off southern California,” Journal of Geophysical Research: Oceans, vol. 103, no. C10, pp. 21601-21609, 1998.
[24] A. A. Gitelson, D. Gurlin, W. J. Moses et al., “A bio-optical algorithm for the remote estimation of the chlorophyll-a concentration in case 2 waters,” Environmental Research Letters, vol. 4, no. 4, pp. 045003, 2009.
[25] M. Woźniak, K. M. Bradtke, and A. Krężel, “Comparison of satellite chlorophyll a algorithms for the Baltic Sea,” Journal of Applied Remote Sensing, vol. 8, no. 1, pp. 083605, 2014.
[26] Y. Zhang, R. Ma, H. Duan et al., “A spectral decomposition algorithm for estimating chlorophyll-a concentrations in Lake Taihu, China,” remote Sensing, vol. 6, no. 6, pp. 5090-5106, 2014.
[27] P. V. Zimba, and A. Gitelson, “Remote estimation of chlorophyll concentration in hyper-eutrophic aquatic systems: Model tuning and accuracy optimization,”Aquaculture, vol. 256, no. 1-4, pp. 272-286, 2006.
[28] S. Koponen, J. Attila, J. Pulliainen et al., “A case study of airborne and satellite remote sensing of a spring bloom event in the Gulf of Finland,” Continental Shelf Research, vol. 27, no. 2, pp. 228-244, 2007.
[29] K. Toming, T. Kutser, A. Laas et al., “First experiences in mapping lake water quality parameters with Sentinel-2 MSI imagery,” Remote Sensing, vol. 8, no. 8, pp. 640, 2016.
[30] L. Dong, C. Gong, H. Huai et al., “Retrieval of Water Quality Parameters in Dianshan Lake Based on Sentinel-2 MSI Imagery and Machine Learning: Algorithm Evaluation and Spatiotemporal Change Research,” Remote Sensing, vol. 15, no. 20,pp. 5001, 2023.
[31] E. Martinez, R. Argamosa, R. Torres et al., “Modelling the Chlorophyll-A Concentration of Laguna Lake Using HIMAWARI-8 Satellite Imagery and Machine Learning Algorithms for Near Real Time Monitoring,” The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 46, pp. 211-214, 2022.
[32] H. Su, X. Lu, Z. Chen et al., “Estimating Coastal Chlorophyll-A Concentration from Time-Series OLCI Data Based on Machine Learning,” Remote Sensing, vol. 13, no. 4, pp. 576, 2021.
[33] B. Ping, F. Su, X. Han et al., “Applications of deep learning-based super-resolution for sea surface temperature reconstruction,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 887-896, 2020.
[34] D. Lambhate, and D. N. Subramani, "Super-resolution of sea surface temperature satellite images." pp. 1-7.
[35] J. A. Parker, R. V. Kenyon, and D. E. Troxel, “Comparison of interpolating methods for image resampling,” IEEE Transactions on medical imaging, vol. 2, no. 1, pp. 31- 39, 1983.
[36] T. Chen, and C. Guestrin, "Xgboost: A scalable tree boosting system." pp. 785-794.
[37] G. Ke, Q. Meng, T. Finley et al., “Lightgbm: A highly efficient gradient boosting decision tree,” Advances in neural information processing systems, vol. 30, 2017.
[38] L. Prokhorenkova, G. Gusev, A. Vorobev et al., “CatBoost: unbiased boosting with categorical features,” Advances in neural information processing systems, vol. 31, 2018.