Estimation of planting date in paddy fields by time-series MODIS data for basin-scale rice production modeling
Estimation of planting date in paddy fields by time-series MODIS data for basin-scale rice production modeling
复制标题
利用时间序列 MODIS 数据估算稻田播种日期，建立流域规模水稻生产模型
DOI：
10.1007/s10333-019-00700-x
复制
发表时间：
2019
影响因子：
2.2
通讯作者：
Koki Homma
中科院分区：
文献类型：
--
作者：
Kumiko Tsujimoto;Tetsu Ohta;Yoshihiro Hirooka;Koki Homma
通过数值模型模拟预测未来农业用水供需变化，可以作为气候变化下农村长期规划的参考资料之一。通过包含农业子模型的分布式水文模型，可以根据水文模型模拟的时间序列土壤湿度来模拟水稻生长及其产量的时间序列。虽然这种农业水文模型有助于评估气候变化对农业的影响，因为它可以在给定的降雨和气象条件下模拟水稻的潜在产量以及与水有关的损害，但通过模型模拟现实情况的挑战之一是如何在模型中给出不固定在某个日历日期且不同田地和不同年份不同的地区的种植日期。在这项研究中，我们研究了一种利用卫星数据估算稻田播种日期的算法。我们使用的数据是中分辨率成像光谱仪（MODIS）的16天复合增强植被指数（EVI）产品，它以250米的空间分辨率提供全球覆盖。首先分析每个像素处的 EVI 值的时间序列，以检测标题日期作为每年的局部最大值。考虑到水稻的生长天数，根据局部极大值的时间间隔来判断是否存在二茬和三茬稻。然后根据确定的抽穗日期和 EVI 时间序列 EVI 的局部最小值来估计种植日期。预计的种植日期根据日本的国家统计数据以及柬埔寨的实地调查数据进行了验证。他们对地区统计数据做出了合理的估计，但在个别领域的比较中仍然存在差距。
Projection of future changes in agricultural water demand and supply through numerical model simulations can be one of the reference materials used for long-term rural planning under climate change. With a distributed hydrological model incorporating an agricultural sub-model, time-series rice growth and its yield can be simulated in correspondence with time-series soil moisture simulated by a hydrological model. Whereas such an agro-hydrological model is useful for assessing the climate change impact on agriculture in that it can simulate the potential rice production and the water-related damages on it against the given rainfall and meteorological condition, one of the challenges in simulating a realistic situation by a model is how to give a planting date in the model in regions where it is not fixed to a certain calendar date and differs in different fields and in different years. In this study, we examined an algorithm to estimate planting date in paddy fields using satellite data. The data we used are the 16-day composite Enhanced Vegetation Index (EVI) product by the moderate resolution imaging spectroradiometer (MODIS), which provides global coverage with 250-m spatial resolution. The time series of the EVI values at each pixel was analyzed first to detect heading dates as the local maxima in each year. Considering the number of days for growing rice, the existence of the second and the third rice cropping was judged from the time intervals of the local maxima. The planting dates were then estimated from the determined heading dates and local minima of the EVI time-series EVI. The estimated planting dates were validated against national statistical data in Japan and also with field-surveyed data in Cambodia. They showed reasonable estimations in the areal statistics, but there remained gaps in the comparison at individual fields.
基于LAI测量的柬埔寨菩萨省农户稻田耕作环境与管理评价
影响因子：
5.8
作者：
Yoshihiro Hirooka;Koki Hommaa;Taisuke Kodo;Tatsuhiko Shiraiwa;Kim Soben;Mithona Chann;Kumiko Tsujimoto;Katsunori Tamagawa;Toshio Koike
通讯作者：
Toshio Koike