首页> 外文期刊>ISPRS Journal of Photogrammetry and Remote Sensing >Open-air grape classification and its application in parcel-level risk assessment of late frost in the eastern Helan Mountains
【24h】

Open-air grape classification and its application in parcel-level risk assessment of late frost in the eastern Helan Mountains

机译:露天葡萄分类及其在东升山东山末霜冻风险评估的应用

获取原文
获取原文并翻译 | 示例
       

摘要

The eastern Helan Mountains are one of the most important plantings and production bases for high-quality grapes in China. In winter, grape root is prone to freezing damage owing to natural phenomenon such as early and late frost, and methods to assess disaster risk at the grape parcel level is one of the current research hotspots. However, there are two problems at present; first is the lack of grape parcel maps and second is that scale of disaster risk assessment is limited to the level of administrative divisions and grid systems. Therefore, according to the perennial characteristics of the grapes and clear texture of the open-air grape parcels based on public data sets (Sentinel-2A/B) and high-quality winery address information, combined with Google Earth engine (GEE) platform, Google Earth, and Ovitalmap, this study designs a collection scheme of open-air grape parcel samples and non-grape parcel samples, analyzes the environmental characteristics of the distribution of high-quality wineries, constructs a classification feature database in three aspects (time, terrain, and vegetation), uses the random forest classification method to classify open-air grape parcels in the eastern Helan Mountains in 2019, and conducts a late frost disaster risk assessment of the open-air grape parcel. The results show that: (1) The high-quality wineries are mainly distributed in the elevation of 1106-1240 m, with the average slope of 0.82., the organic matter content of 10.79-14.27 g/kg, the P content of 24.87-37.58 mg/kg, the K content of 116-166 mg/kg, the pH 8.38-8.55, N content of 0.65-0.98 g/kg, annual average active accumulated temperature = 10 degrees C of 3490-3590 degrees C, annual average rainfall of 189-201 mm, the annual average sunshine duration of 7.3-7.6 h, annual average air temperature of 8.8-9.4 degrees C, annual average temperature difference of 12.7-13.2 degrees C, annual average frost-free period at 263-270 d and annual average relative air Humidity of 50-54%. (2) The overall classification accuracy of open-air grape parcels reached 98.96% and the verification accuracy of ground truth parcel samples reached 87.89%. The open-air grape parcels in the eastern Helan Mountains are mainly distributed in Helan County, Xixia District, Yongning County, and Qingtongxia City. (3) The highest risk of late frost was found in Xixia District and the lowest risk in Yongning County. Helan County was higher than Qingtongxia City in the medium risk of late frost. The typical small area high risk includes Liangtian Town, the experimental field of Ningxia Academy of Agricultural and Forestry Sciences in Zhenbeibao Town, Yuhuang Winery in Ganchengzi Village, and the State-run Warm Spring Farm in Helan County. Thus, the sample collection and augmentation scheme are feasible, the open-air grape classification accuracy is high and its result is consistent with field investigation, which can provide accurate parcel-level disaster risk assessment and provide decision support for agricultural departments and farmers.Y
机译:东升山是中国高品质葡萄的最重要的种植和生产基地之一。在冬天,由于早期和晚期和晚期霜冻等自然现象,葡萄根易于冻结损坏,以及评估葡萄包裹水平的灾害风险的方法是目前的研究热点之一。但是,目前有两个问题;首先是缺乏葡萄包地图,第二是灾害风险评估规模仅限于行政划分和网格系统的水平。因此,根据葡萄的常年特征和露天葡萄包的透明质感,基于公共数据集(Sentinel-2a / b)和高质量的酿酒厂地址信息,结合谷歌地球发动机(Gee)平台,谷歌地球和ovitalMap,这项研究设计了露天葡萄包样品和非葡萄包装样本的收集方案,分析了高质量葡萄酒厂的分布的环境特征,在三个方面构建了分类功能数据库(时间,地形和植被)使用随机森林分类方法在2019年在东部亥安山山区进行分类,并对露天葡萄包裹进行霜冻灾害风险评估。结果表明:(1)高质量的葡萄酒厂主要分布在1106-1240米的升高,平均斜率为0.82。,有机质含量为10.79-14.27g / kg,p含量为24.87 -37.58 mg / kg,k含量为116-166 mg / kg,pH 8.38-8.55,n含量为0.65-0.98g / kg,年平均活动累积温度> = 10度C 3490-3590 C,年平均降雨量为189-201毫米,年平均阳光持续时间为7.3-7.6小时,年平均空气温度为8.8-9.4摄氏度,年平均温度差12.7-13.2摄氏度,年平均冻结期263 -270 D和年平均相对空气湿度为50-54%。 (2)露天葡萄包裹的整体分类准确性达到98.96%,地面真理包裹样品的验证准确性达到87.89%。东部升航山脉的露天葡萄包裹主要分布在贺兰县,Xixia区,永宁县和青通峡市。 (3)在西溪区发现了最高霜的风险以及永宁县的最低风险。贺兰县高于青东峡城市中等风险的晚期霜冻。典型的小面积高风险包括梁天镇,浙北镇宁夏农业和林业科学院的实验领域,玉煌酒厂在江晟村,贺兰县的国有温暖的春季农场。因此,样品收集和增强方案是可行的,露天葡萄分类精度高,其结果与现场调查一致,可提供准确的包裹级灾害风险评估,并为农业部门和农民提供决策支持.Y

著录项

获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号