台灣位於亞熱帶區，農業技術進步，蔬菜產量豐盛，其中葉菜類之甘藍菜產量為最大宗。但農民常集中於同一時機進行插苗，以致於在豐收時期甘藍菜會供過於求，進而導致價格崩盤，使菜農收入受到影響。農委會希望掌握甘藍菜於各地之生長期，以方便做產銷調整決策，讓甘藍菜價保持平穩。
前人運用多光譜衛星影像對生長期進行判釋，但其必須刪除衛星中雲遮蔽(陰影)區域，導致在衛星影像雲覆高時，當期則無法進行判釋，連續雨季更會長期缺失資料。
本研究使用Google Earth Engine雲端運算服務器，取用服務器內Sentinel-1合成孔徑雷達衛星影像與Senntinel-2多光譜衛星影像，抓取各個田地坵塊回波反射率，再結合農委會於民國108年8月至11月做的現地調查，運用統計機器學習的極限梯度提升法，做機器學習模型訓練，並用測試資料判釋甘藍菜生長期。
判釋結果顯示，雷達衛星影像的加入有助於整體準確率從約六成五提升至約七成，雲機率70%以上的坵塊更是提升了10%的判釋準確率。判釋錯誤主要因素以衛星資料解析度不足導致分期特徵不夠精確為主，地表調查GPS定位錯誤雖占少部分，但其與最終判釋結果誤差較大。而極限梯度提升法為此案例中，準確率最高的統計機器學習法

Taiwan is in subtropical region, which is suitable for agricultural development. The cabbage has the largest amount of vegetable production in Taiwan, but the peasantry usually starts planting on the same timing without any planning and leads to the oversupply result. The Council of Agriculture in Taiwan decided to monitor the growth period of cabbage and warn the peasantry to cut production in time for keeping the cabbage price stable. When using multispectral image (MSI) data to complete the task, some problems are encountered. The pixels in image which shrouded in clouds would be masked out and been eliminated from the image. If a satellite image is covered by clouds, then the task can’t be completed, which is very common in the sub-tropical area like Taiwan. Missing one satellite image means another blank 7~10 days information is introduced into the system.
To maintain the temporal resolution of interpretation, the study takes the advantage of synthetic aperture radar (SAR). This study was using Sentinel-1 SAR and Sentinel-2 MSI data on Google Earth Engine, combining them with the ground truth data investigated in field, training model with the eXtreme Gradient Boosting (Xgboost) method, accomplishing interpretation of testing data eventually.
At the result, the join of SAR data uplifts the total accuracy from 65% to close to 70%. The accuracy of interpretation at the region where cloud probability exceeds 70% increases 10% moreover. The reasons of the incorrect interpretation take to the insufficient of image’s pixel resolution and unstable of GPS in ground truth data. Xgboost is the most accurate model in this study.

Belgiu, M., & Csillik, O. (2018). Sentinel-2 cropland mapping using pixel-based and object-based time-weighted dynamic time warping analysis. Remote Sensing of Environment, 204, 509-523. https://doi.org/10.1016/j.rse.2017.10.005
Bolstad, P. (2016). GIS Fundamentals.
Chen, F., Zhang, M., Tian, B., & Li, Z. (2017). Extraction of Glacial Lake Outlines in Tibet Plateau Using Landsat 8 Imagery and Google Earth Engine. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(9), 4002-4009. https://doi.org/10.1109/jstars.2017.2705718
Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. KDD '16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. https://doi.org/10.1145/2939672.2939785
Corporation, S. I. Sentinel-2A SatelliteSensor https://www.satimagingcorp.com/satellite-sensors/other-satellite-sensors/sentinel-2a/
ESA. Level-2A Algorithm. https://sentinels.copernicus.eu/web/sentinel/technical-guides/sentinel-2-msi/level-2a/algorithm
ESA. Spatial Resolution. https://sentinels.copernicus.eu/web/sentinel/user-guides/sentinel-2-msi/resolutions/spatial
ESA. (2003). Work progress of the Initial Period of the GMES Action Plan.
ESA. (2007). ASAR Product Handbook.
ESA. (2020). Sentinel2_infographic. https://sentinels.copernicus.eu/web/sentinel/missions/sentinel-2/news/-/asset_publisher/Ac0d/content/id/4181047
Freund, Y., & Schapir, R. E. (1999). A Short Introduction to Boosting.
Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., & Moore, R. (2017). Google Earth Engine: Planetary-scale geospatial analysis for everyone [Article]. Remote Sensing of Environment, 202, 18-27. https://doi.org/10.1016/j.rse.2017.06.031
Joshi, A. R., Dinerstein, E., Wikramanayake, E., Anderson, M. L., Olson, D., Jones, B. S., Seidensticker, J., Lumpkin, S., Matthew C. Hansen, Sizer, N. C., Davis, C. L., Suzanne Palminteri, & Hahn, N. R. (2016). Tracking changes and preventing loss in critical.
Khabbazan, S., Vermunt, P., Steele-Dunne, S., Ratering Arntz, L., Marinetti, C., van der Valk, D., Iannini, L., Molijn, R., Westerdijk, K., & van der Sande, C. (2019). Crop Monitoring Using Sentinel-1 Data: A Case Study from The Netherlands. Remote Sensing, 11(16). https://doi.org/10.3390/rs11161887
Nagaraju, A., Kumar reddy, M. A., Venugopal reddy, C. H., & Mohandas, R. (2021). Multifactor Analysis to Predict Best Crop using Xg-Boost Algorithm 2021 5th International Conference on Trends in Electronics and Informatics (ICOEI),
NASA, E. O. (2000). Normalized Difference Vegetation Index (NDVI).
NASA, S., & SilvaCarbon. (2019). THE SAR HANDBOOK. https://doi.org/10.25966/nr2c-s697
Nikulski, J. (2020). The Ultimate Guide to AdaBoost, random forests and XGBoost.
Prasad, A. K., Chai, L., Singh, R. P., & Kafatos, M. (2006). Crop yield estimation model for Iowa using remote sensing and surface parameters. International Journal of Applied Earth Observation and Geoinformation, 8(1), 26-33. https://doi.org/10.1016/j.jag.2005.06.002
QUINLAN, J. R. (1985). Induction of Decision Trees.
Raschka, S. (2015). Python Machine Learning.
Richter, R., & Schl¨apfer, D. Atmospheric Topographic Correction for Satellite Imagery.
Rizzoli, P., & Brautigam, B. (2014). Radar Backscatter Modeling Based on Global TanDEM-X Mission Data. IEEE Transactions on Geoscience and Remote Sensing, 52(9), 5974-5988. https://doi.org/10.1109/tgrs.2013.2294352
Schapire, R. E. (2001). Random Forests.
Souza, C. M., Z. Shimbo, J., Rosa, M. R., Parente, L. L., A. Alencar, A., Rudorff, B. F. T., Hasenack, H., Matsumoto, M., G. Ferreira, L., Souza-Filho, P. W. M., de Oliveira, S. W., Rocha, W. F., Fonseca, A. V., Marques, C. B., Diniz, C. G., Costa, D., Monteiro, D., Rosa, E. R., Vélez-Martin, E., . . . Azevedo, T. (2020). Reconstructing Three Decades of Land Use and Land Cover Changes in Brazilian Biomes with Landsat Archive and Earth Engine. Remote Sensing, 12(17). https://doi.org/10.3390/rs12172735
Veloso, A., Mermoz, S., Bouvet, A., Le Toan, T., Planells, M., Dejoux, J.-F., & Ceschia, E. (2017). Understanding the temporal behavior of crops using Sentinel-1 and Sentinel-2-like data for agricultural applications. Remote Sensing of Environment, 199, 415-426. https://doi.org/10.1016/j.rse.2017.07.015
Zamani Joharestani, M., Cao, C., Ni, X., Bashir, B., & Talebiesfandarani, S. (2019). PM2.5 Prediction Based on Random Forest, XGBoost, and Deep Learning Using Multisource Remote Sensing Data. Atmosphere, 10(7). https://doi.org/10.3390/atmos10070373
行政院農業委員會農糧署. (2020). 農產品生產量值. 農業統計年報.
李弘毅. (2018). ML Lecture 22: Ensemble https://www.youtube.com/watch?v=tH9FH1DH5n0&t=494s
農委會. (2020). 109 年度多時期衛星影像作物判釋研究計畫.
廖芳心, 洪立, & 黃涵. (1993). 甘藍及結球白菜之結球生理. 蔬菜生產與發展研討會專刊.
蕭柊瓊. (2019). 應用衛星影像光譜大數據自動判釋作物生長天數之研究.