基于MODIS数据的泰国耕地信息提取

农业是中国和多数东盟国家的基础产业，加深对东盟国家农业生产信息的认识，可以为中国与东盟的农业合作提供科学的决策参考，促进合作的健康发展。该文选取了世界上重要的水稻种植地之一的泰国作为研究对象，应用时间序列MODIS数据以及CBERS-02B数据、实测和统计等辅助数据，进行了泰国耕地信息的提取研究。该文采用的技术流程首先根据时间序列归一化植被指数数据，采用排除法提取出研究区耕地范围；其次采用探测峰值的方法确定农作物的熟制，根据熟制分区进行作物类型识别。结果表明：水稻是泰国主要的农作物类型，集中分布在湄南河流域和呵叻高原地区；旱地作物主要分布在湄南河流域边缘的中高台地地区；一年一熟的种植方式集中分布在东北部的呵叻高原；一年两熟的种植方式集中分布在湄南河流域，又可以划分为水田多熟系统和旱地多熟系统；一年三熟的种植方式也集中分布在湄南河流域地区，种植面积很小且多为水田多熟系统。在精度验证中，将CBERS-02B数据提取结果与MODIS提取结果进行空间比对，二者具有非常相似的空间分布特征。分省统计结果与遥感反演结果也具有较好的相关性。证明了MODIS数据以及该研究采用的方法在耕地信息提取中的有效性，为耕地信息的遥感提取研究提供了参考方法。

Extraction of information of cultivated land using time-series MODIS data in Thailand

The difference and the complementariness on the agriculture decide that there is a wide-ranging cooperation on agriculture between China and ASEAN. So the research on ASEAN's agriculture has important significance. In order to acquire the cultivated land related information in Thailand, time-series MODIS data were used in this paper. Firstly, the 250 m cultivated land map was generated by using time-series NDVI data. Then the cropping intensity map was acquired by determining how many peaks were detected in the smoothed NDVI pro?le on a pixel-by-pixel basis, which then served as a mask to extract crop types in different cropping intensity regions. The results showed that the main land use type in the study area was cultivated land, most of which were dominated by rice. Double or triple rice-cropping system was commonly employed in the Chao Phraya Basin. Upland crops were mainly distributed in the high alluvial terraces of the Chao Phraya Basin. Because Khorat Plateau in the northeast of Thailand often have water shortage problem, particularly in the dry season, which can just support only one crop in a year. However, some upland areas can be cultivated twice a year with crops which have short growing seasons. The crop information extracted from MODIS data sets were assessed by CBERS data, statistic data. It was shown that MODIS derived information related to cultivated land coincided well with the statistic data at the provincial level. At the same time, information extracted by MODIS data sets and CBERS were compared with each other which also showed similar spatial patterns.