关键词: 水稻, 遥感, 农情监测, 多源异构数据, WebGIS

Abstract: To address the pressing challenges associated with inadequate fertilizer and pesticide application, as well as the absence of systematic management, resulting in low yield per unit area and agricultural non‑point source pollution in rice cultivation in China, this study focuses on the integration of multi‑source data through collaborative monitoring. A frontend‑backend decoupled, cloud‑based, real‑time, field‑level rice crop monitoring platform was constructed, based on WebGIS technology and the Ant Design front‑end framework. The platform construction incorporated heterogeneous spatiotemporal geospatial data from multiple sources and utilized a distributed data storage architecture, employing techniques such as Python, HTML, JavaScript+CSS, ArcGIS Server, Mapbox Studio web services, and the PostgreSQL database. Consequently, the platform offered a myriad of functionalities, including rice growth parameter inversion, yield prediction, plot parameter query, as well as spatiotemporal data visualization and statistical analysis, among other functionalities. The system was tested in two experimental areas: Xingqiao Town and Jinggangshan National Agricultural Science and Technology Park in Jiangxi Province. The visual analysis of the platform indicates that the distribution of rice fields in Xingqiao Town in 2022 was fragmented, with higher rice yields in the northeastern region compared to the southwest. The rice yields in the town ranged between 6 750 kg/hm2 and 8 250 kg/hm2. The study also found a significant correlation between rice growth in the experimental area and the historical application of pesticides and fertilizers, indicating that application strategies have a considerable impact on rice growth. In conclusion, this platform, through the collaborative integration of multiple data sources, can effectively fulfill the accuracy and comprehensiveness requirements for rice monitoring at a large regional scale. It also enables attribution analysis of rice growth and yield to some extent. This platform serves as an effective example of fine‑grained, multi‑source monitoring of rice farming at the field scale.

Key words: rice, remote sensing, crop condition monitoring, multiple heterogeneous data, WebGIS