基于卫星遥感的耕地质量监测与评价研究进展

RESEARCH PROGRESS ON CULTIVATED LAND QUALITY MONITORING AND ASSESSMENT BY REMOTE SENSING

关键词:  耕地质量  遥感  评价指标  土壤有机质  耕地生产力

Key words:  cultivated land quality  remote sensing  evaluation indicator  soil organic matter  farmland productivity

摘要:

目的 耕地质量监测与评价对于耕地质量保护提升、水肥高效精准施用、粮食生产持续发展及高标准农田建设管理具有重要意义。卫星遥感技术以其覆盖范围大、探测周期短、监测成本低等优点可为耕地质量的及时准确监测提供强有力的技术支撑。方法 文章对近年来国内外耕地质量遥感监测与评价方面的研究进展进行了系统总结，梳理了卫星遥感可获取的耕地质量表征指标，分析了各指标的遥感估算方法及所需数据源，归纳了现有耕地质量综合评价使用的研究方法。结果 （1）目前国内外利用卫星遥感影像获取的耕地质量指标主要包括土壤属性、环境属性、耕地生产力和综合性指标4种类型；（2）基于卫星遥感的耕地质量指标获取方法主要有波段运算、统计模型、机器学习和神经网络等；（3）耕地质量评价方法主要为加权求和型和连乘型2种类型。结论 耕地质量遥感监测要构建利用遥感电磁波特征估算多种耕地质量监测与评价指标的深度学习模型，综合利用多源、多平台以及多时相数据提高遥感估算精度，科学确定耕地质量综合评价指标权重，实现区域耕地质量的客观评价。

Abstract:

Monitoring and evaluation of cultivated land quality are of great significance for the protection and improvement of cultivated land quality, efficient and precise application of water and fertilizer, sustainable development of grain production, and management of high-standard farmland. Satellite remote sensing technology, with its advantages of large coverage, short detection cycle, and low monitoring cost, can provide strong technical support for the timely and accurate monitoring of cultivated land quality. This research systematically summarized the research progress in the field of remote sensing monitoring and evaluation of cultivated land quality at home and abroad in recent years, sorted out the cultivated land quality characterization indicators that could be obtained by satellite remote sensing, analyzed the remote sensing estimation methods and required data sources of each indicator, and summarized the research methods used in the existing comprehensive evaluation of cultivated land quality. The results were showed as follows. (1) The cultivated land quality indicators obtained by satellite remote sensing images at home and abroad mainly included four types: soil properties, environmental properties, cultivated land productivity, and comprehensive indicators; (2)The main methods for obtaining cultivated land quality indicators based on satellite remote sensing include band operation, statistical models, machine learning, and neural network, etc.; (3) The evaluation methods of cultivated land quality mainly included two types: weighted sum type and product type. In summary, remote sensing monitoring of cultivated land quality should construct a deep learning model for estimating multiple cultivated land quality monitoring and evaluation indicators using the electromagnetic wave characteristics of remote sensing, comprehensively utilize multi-source, multi-platform, and multi-temporal data to improve the accuracy of remote sensing estimation, scientifically determine the weights of comprehensive evaluation indicators of cultivated land quality, and achieve objective evaluation of regional cultivated land quality.