基于多时相植被指数的云南高原山地冬小麦识别与研究

粮食安全是最根本的民生问题，云、雾等自然因素是影响遥感种植监测的主要因素之一，因此获取精准、高效的耕地种植监测信息的对保障当地粮农安全、粮食估产及面积估算具有重要意义。在利用多时相植被指数（VI）合成模型的构建、农作物特征与耕地信息的可分离性两方面对高原山地农作物耕地面积提取的研究少。本研究基于哨兵2（Sentinel-2）数据，构建了多时相植被指数合成模型，估算了2020-2021年归一化植被指数（NDVI）、增强植被指数（EVI）和红绿叶绿素植被指数（RECI）三种植被指数的提取结果，研究了预测模型与高原山地农作物的相关性，探讨了不同植被指数模型对农作物识别精度。结果表明：①多时相NDVI模型相较EVI、RECI对冬小麦面积提取精度更高，与云南高原山地冬小麦相关性最强，用户精度约为93.28%；②利用三期NDVI组合与两期NDVI组合均可对冬小麦精准提取，但三期NDVI草型提取精度更高。因此，本研究利用多时相NDVI指数模型对冬小麦种植面积的精准预测，证明了该模型可有效适用于云南高原山地冬小麦，并为当地冬小麦面积的预测提供了数据支撑。

Identification and research of winter wheat in Yunnan Plateau based on multi-temporal vegetation index

The stability of food security is the most fundamental livelihood issue of the country, as the key factor affecting remote sensing planting monitoring, natural factors such as clouds and fog are key factors. Therefore, accurate and efficient rapid acquisition of cultivated land planting monitoring information is of great significance for ensuring food and agriculture security, grain yield estimation, and area estimation in the study area. At present, there are few studies on the construction of the multi-period vegetation index ( VI ) synthetic model and the separability of crop characteristics and cultivated land information in remote sensing cultivated land monitoring. We use Sentinel-2 data to study through a multi-temporal vegetation index synthesis model. In this study, we compared the extraction results of the Normalized Vegetation index (NDVI), Enhanced Vegetation Index (EVI), and Red Green Chlorophyll Vegetation Index (RECI), studied the correlation between the prediction model and the winter wheat on the Yunnan Plateau and discussed the crop identification accuracy of different vegetation index models.The result shows that: (1) Compared with EVI and RECI, the multi-temporal NDVI model can better identify winter wheat with higher area extraction accuracy and has the strongest correlation with winter wheat in Yunnan Plateau. The user accuracy is about 93.28%. (2) The total accuracy of the three-stage index combination and two-stage index combination is above 90%. Overall, We use the multi-temporal NDVI index model to accurately predict the winter wheat planting area, which proves that the model can be effectively applied to the mountain crops in Yunnan Plateau, and provides data support for the prediction of the local winter wheat area.