This study aims to detect flooded rice paddies in Indonesia using remotely sensed data from a relatively extensive flood that occurred in the Tegalluar area of Bojongsoang in the spring of 2018, which was observed by the Sentinel-1 and PlanetScope satellites. We propose an automatic thresholding method for the detection of flooded areas in rice paddy fields using Sentinel-1 C -band synthetic aperture radar (SAR) data acquisitions from before and during flooding. The flood-detection accuracy was verified using visible and near-infrared images acquired by the PlanetScope satellites. The proposed method showed that the VV (transmit V and receive V polarizations) data outperformed the VH (transmit V and receive H polarizations) data in terms of correlation ratio and discriminant accuracy. The overall classification accuracy of the nonflooded and flooded areas reached 84.7% with the VV data and 80.6% with the VH data, including the error that resulted from the time difference in the data acquired by Sentinel-1 and PlanetScope. Utilizing speckle-reducing filters with SAR data was found to improve the overall classification accuracy by 5%.

中文翻译：

---

使用 Sentinel-1 和 PlanetScope 数据检测洪水泛滥的稻田：印度尼西亚西爪哇 2018 年春季洪水的案例研究

这项研究旨在使用遥感数据来检测印度尼西亚的洪水泛滥的稻田，该数据来自 2018 年春季发生在 Bojongsoang 的 Tegalluar 地区的相对广泛的洪水，由 Sentinel-1 和 PlanetScope 卫星观测到。我们提出了一种使用 Sentinel-1 检测稻田淹水区域的自动阈值方法C 洪水前和洪水期间的波段合成孔径雷达 (SAR) 数据采集。使用 PlanetScope 卫星获取的可见光和近红外图像验证了洪水探测的准确性。所提出的方法表明，在相关比和判别精度方面，VV（发射 V 和接收 V 极化）数据优于 VH（发射 V 和接收 H 极化）数据。非淹没区和淹没区的总体分类准确率在 VV 数据下达到了 84.7%，在 VH 数据下达到了 80.6%，其中包括 Sentinel-1 和 PlanetScope 获取的数据时间差导致的误差。发现对 SAR 数据使用去斑滤波器可以将整体分类精度提高 5%。