基于DHSVM模型的作物渍害时空分布信息提取

为了实时监测作物渍害时空分布,以湖北省监利县作为研究对象,采用分布式水文土壤植被模型-DHSVM(Distributed Hydrology Soil Vegetation Model)将监利县分成若干90 m×90 m的栅格单元,通过计算每个栅格单元的水平衡,达到模拟农田土壤水分状况的目的,结合作物轻、中、重度3种渍害的水分特征指标(地下水位埋深小于0.6 m,土壤表层相对体积含水率5 d的均值大于90%的持续期分别大于5、12、20 d)的方法,模拟监利县夏收作物不同渍害的时空分布。结果表明,模拟得到了2014年、2015年每年1—5月(夏收作物生长期间)监利县农田表层土壤水分和地下水位埋深的时空分布(时间分辨率为1 d,空间分辨率为90 m),首次实现了作物渍害的时空间分布信息提取;通过多次田间调查,证明此反演方法结果是准确的。

Calculating Spatio-spatial Distribution of Waterlogged in Jianli County Using the DHSVM Model

We calculated the spatio-spatial distribution of waterloggingin Jianli County of Hubei province using the DHSVM model (The Distributed Hydrology Soil Vegetation Model). The region in the county was divided in-to a number of grids, each being 90 m×90 m. The dynamic of soil moisture in each grid was simulated using a wa-ter balance model. We calculated the distribution of waterlogging in summer by classifying the waterlogging into three categories when the groundwater table was less than 0.6 m deep:when the duration of five-day average vol-umetric water higher than 90%lasted more than 5 days (light waterlogging), or lasted more than 12 days (moder-ate waterlogging), or lasted more than 20 days (severe waterlogging). We simulated the distribution of soil mois-ture and groundwater table in all grids from January to May in 2014 and 2015 during the growing season of sum-mer crops using a temporal resolution of one day. These results enabled us to predict the spatio-temporal distribu-tion of waterlogging in the region for the first time. This retrieval method was proven accurate after testing in sev-eral fields.