| 摘要: |
| 甘蔗寒冻害是蔗糖主产区主要气象灾害之一。大范围、严重的甘蔗寒冻害造成甘蔗产量和蔗糖分含量下降，给广大蔗农及企业带来巨大的经济损失。但受蔗区地形地貌复杂及气象观测台站分布稀疏等因素制约，寒冻害过程影响期间难以获取区域甘蔗实际受害信息，给甘蔗寒冻害过程监测、评估及防灾、减灾措施的实施带来了极大困难。为了实现及时、真实地反演甘蔗产区寒冻害程度，文章基于广西历年冬季寒冻害过程最低气温资料、地理信息数据以及EOS/MODIS卫星资料，采用多元回归法，分别构建基于地理信息、基于地理信息及下垫面信息耦合的最低气温空间推算模型，创建了逐年寒冻害过程最低气温格点数据库。并以2008年辐射型、2010年平流型寒冻害过程为例，将反演结果与实况进行比较，结果表明两种模型反演所得寒冻害空间分布状况均与自动站实况一致，两种模型对平流型寒冻害的模拟效果优于对辐射型寒冻害的模拟。 |
| 关键词: 甘蔗 寒冻害 反演模型 MODIS 多元回归 |
| DOI：10.7621/cjarrp.1005-9121.20160603 |
| 分类号: |
| 基金项目: |
|
| RESEARCH ON INVERSION MODEl OF SUGARCANE′S FREEZING DAMAGE DISTRIBUTION IN GUANGXI |
|
Meng Cuili1, Tan Zongkun2, Li Zitian3
|
|
1.Wuhan National Agro meteorology Station，Wuhan Hubei 430040，China;2.Guangxi Meteorological Disaster Mitigation Institute，Nanning Guangxi 530022，China;3.Nanning Meteorological Bureau，Nanning Guangxi 530029，China
|
| Abstract: |
| The freezing injury was one of the main meteorological disasters in main sugar production region.A wide spread and serious freezing injury would cause the decrease in sugarcane production and sugar content，and hence lead to the huge economic losses for farmers and business.However，the information of the injury during the freeze stage was hard to acquire due to the complex topography and the sparse distribution of the meteorological stations.This led to huge difficulties for monitoring and assessment of the freezing injury，and the disaster prevention and reduction.To timely reflect the real situation of regional freezing damage，based on the geographic information，EOS/MODIS data and minimum temperature data during freezing damage in Guangxi，this paper built up the spatial estimation models of minimum temperature of geographic information，the coupling of geographic information，and underlying surface information using the multivariate statistical regression method，and created the minimum temperature lattice database of annual freezing damage process.Then，it compared the inversion results with the in-situ data taking radiation freezing damage in 2008 and advection freezing damage in 2010 as an example.The results showed that the inverted spatial distribution of the freezing damage was agreement with the site monitoring result.In addition，the inversion modelling for advection freezing damage had higher accuracy than that for radiation freezing damage. |
| Key words: sugarcane freezing damage inversion model MODIS multivariate regression |
